diff options
21 files changed, 9037 insertions, 0 deletions
@@ -0,0 +1,674 @@ + GNU GENERAL PUBLIC LICENSE + Version 3, 29 June 2007 + + Copyright (C) 2007 Free Software Foundation, Inc. <http://fsf.org/> + Everyone is permitted to copy and distribute verbatim copies + of this license document, but changing it is not allowed. + + Preamble + + The GNU General Public License is a free, copyleft license for +software and other kinds of works. + + The licenses for most software and other practical works are designed +to take away your freedom to share and change the works. By contrast, +the GNU General Public License is intended to guarantee your freedom to +share and change all versions of a program--to make sure it remains free +software for all its users. We, the Free Software Foundation, use the +GNU General Public License for most of our software; it applies also to +any other work released this way by its authors. You can apply it to +your programs, too. + + When we speak of free software, we are referring to freedom, not +price. Our General Public Licenses are designed to make sure that you +have the freedom to distribute copies of free software (and charge for +them if you wish), that you receive source code or can get it if you +want it, that you can change the software or use pieces of it in new +free programs, and that you know you can do these things. + + To protect your rights, we need to prevent others from denying you +these rights or asking you to surrender the rights. Therefore, you have +certain responsibilities if you distribute copies of the software, or if +you modify it: responsibilities to respect the freedom of others. + + For example, if you distribute copies of such a program, whether +gratis or for a fee, you must pass on to the recipients the same +freedoms that you received. You must make sure that they, too, receive +or can get the source code. And you must show them these terms so they +know their rights. + + Developers that use the GNU GPL protect your rights with two steps: +(1) assert copyright on the software, and (2) offer you this License +giving you legal permission to copy, distribute and/or modify it. + + For the developers' and authors' protection, the GPL clearly explains +that there is no warranty for this free software. For both users' and +authors' sake, the GPL requires that modified versions be marked as +changed, so that their problems will not be attributed erroneously to +authors of previous versions. + + Some devices are designed to deny users access to install or run +modified versions of the software inside them, although the manufacturer +can do so. This is fundamentally incompatible with the aim of +protecting users' freedom to change the software. The systematic +pattern of such abuse occurs in the area of products for individuals to +use, which is precisely where it is most unacceptable. Therefore, we +have designed this version of the GPL to prohibit the practice for those +products. If such problems arise substantially in other domains, we +stand ready to extend this provision to those domains in future versions +of the GPL, as needed to protect the freedom of users. + + Finally, every program is threatened constantly by software patents. +States should not allow patents to restrict development and use of +software on general-purpose computers, but in those that do, we wish to +avoid the special danger that patents applied to a free program could +make it effectively proprietary. To prevent this, the GPL assures that +patents cannot be used to render the program non-free. + + The precise terms and conditions for copying, distribution and +modification follow. + + TERMS AND CONDITIONS + + 0. Definitions. + + "This License" refers to version 3 of the GNU General Public License. + + "Copyright" also means copyright-like laws that apply to other kinds of +works, such as semiconductor masks. + + "The Program" refers to any copyrightable work licensed under this +License. Each licensee is addressed as "you". "Licensees" and +"recipients" may be individuals or organizations. + + To "modify" a work means to copy from or adapt all or part of the work +in a fashion requiring copyright permission, other than the making of an +exact copy. The resulting work is called a "modified version" of the +earlier work or a work "based on" the earlier work. + + A "covered work" means either the unmodified Program or a work based +on the Program. + + To "propagate" a work means to do anything with it that, without +permission, would make you directly or secondarily liable for +infringement under applicable copyright law, except executing it on a +computer or modifying a private copy. Propagation includes copying, +distribution (with or without modification), making available to the +public, and in some countries other activities as well. + + To "convey" a work means any kind of propagation that enables other +parties to make or receive copies. Mere interaction with a user through +a computer network, with no transfer of a copy, is not conveying. + + An interactive user interface displays "Appropriate Legal Notices" +to the extent that it includes a convenient and prominently visible +feature that (1) displays an appropriate copyright notice, and (2) +tells the user that there is no warranty for the work (except to the +extent that warranties are provided), that licensees may convey the +work under this License, and how to view a copy of this License. If +the interface presents a list of user commands or options, such as a +menu, a prominent item in the list meets this criterion. + + 1. Source Code. + + The "source code" for a work means the preferred form of the work +for making modifications to it. "Object code" means any non-source +form of a work. + + A "Standard Interface" means an interface that either is an official +standard defined by a recognized standards body, or, in the case of +interfaces specified for a particular programming language, one that +is widely used among developers working in that language. + + The "System Libraries" of an executable work include anything, other +than the work as a whole, that (a) is included in the normal form of +packaging a Major Component, but which is not part of that Major +Component, and (b) serves only to enable use of the work with that +Major Component, or to implement a Standard Interface for which an +implementation is available to the public in source code form. A +"Major Component", in this context, means a major essential component +(kernel, window system, and so on) of the specific operating system +(if any) on which the executable work runs, or a compiler used to +produce the work, or an object code interpreter used to run it. + + The "Corresponding Source" for a work in object code form means all +the source code needed to generate, install, and (for an executable +work) run the object code and to modify the work, including scripts to +control those activities. However, it does not include the work's +System Libraries, or general-purpose tools or generally available free +programs which are used unmodified in performing those activities but +which are not part of the work. For example, Corresponding Source +includes interface definition files associated with source files for +the work, and the source code for shared libraries and dynamically +linked subprograms that the work is specifically designed to require, +such as by intimate data communication or control flow between those +subprograms and other parts of the work. + + The Corresponding Source need not include anything that users +can regenerate automatically from other parts of the Corresponding +Source. + + The Corresponding Source for a work in source code form is that +same work. + + 2. Basic Permissions. + + All rights granted under this License are granted for the term of +copyright on the Program, and are irrevocable provided the stated +conditions are met. This License explicitly affirms your unlimited +permission to run the unmodified Program. The output from running a +covered work is covered by this License only if the output, given its +content, constitutes a covered work. This License acknowledges your +rights of fair use or other equivalent, as provided by copyright law. + + You may make, run and propagate covered works that you do not +convey, without conditions so long as your license otherwise remains +in force. You may convey covered works to others for the sole purpose +of having them make modifications exclusively for you, or provide you +with facilities for running those works, provided that you comply with +the terms of this License in conveying all material for which you do +not control copyright. Those thus making or running the covered works +for you must do so exclusively on your behalf, under your direction +and control, on terms that prohibit them from making any copies of +your copyrighted material outside their relationship with you. + + Conveying under any other circumstances is permitted solely under +the conditions stated below. Sublicensing is not allowed; section 10 +makes it unnecessary. + + 3. Protecting Users' Legal Rights From Anti-Circumvention Law. + + No covered work shall be deemed part of an effective technological +measure under any applicable law fulfilling obligations under article +11 of the WIPO copyright treaty adopted on 20 December 1996, or +similar laws prohibiting or restricting circumvention of such +measures. + + When you convey a covered work, you waive any legal power to forbid +circumvention of technological measures to the extent such circumvention +is effected by exercising rights under this License with respect to +the covered work, and you disclaim any intention to limit operation or +modification of the work as a means of enforcing, against the work's +users, your or third parties' legal rights to forbid circumvention of +technological measures. + + 4. Conveying Verbatim Copies. + + You may convey verbatim copies of the Program's source code as you +receive it, in any medium, provided that you conspicuously and +appropriately publish on each copy an appropriate copyright notice; +keep intact all notices stating that this License and any +non-permissive terms added in accord with section 7 apply to the code; +keep intact all notices of the absence of any warranty; and give all +recipients a copy of this License along with the Program. + + You may charge any price or no price for each copy that you convey, +and you may offer support or warranty protection for a fee. + + 5. Conveying Modified Source Versions. + + You may convey a work based on the Program, or the modifications to +produce it from the Program, in the form of source code under the +terms of section 4, provided that you also meet all of these conditions: + + a) The work must carry prominent notices stating that you modified + it, and giving a relevant date. + + b) The work must carry prominent notices stating that it is + released under this License and any conditions added under section + 7. This requirement modifies the requirement in section 4 to + "keep intact all notices". + + c) You must license the entire work, as a whole, under this + License to anyone who comes into possession of a copy. This + License will therefore apply, along with any applicable section 7 + additional terms, to the whole of the work, and all its parts, + regardless of how they are packaged. This License gives no + permission to license the work in any other way, but it does not + invalidate such permission if you have separately received it. + + d) If the work has interactive user interfaces, each must display + Appropriate Legal Notices; however, if the Program has interactive + interfaces that do not display Appropriate Legal Notices, your + work need not make them do so. + + A compilation of a covered work with other separate and independent +works, which are not by their nature extensions of the covered work, +and which are not combined with it such as to form a larger program, +in or on a volume of a storage or distribution medium, is called an +"aggregate" if the compilation and its resulting copyright are not +used to limit the access or legal rights of the compilation's users +beyond what the individual works permit. Inclusion of a covered work +in an aggregate does not cause this License to apply to the other +parts of the aggregate. + + 6. Conveying Non-Source Forms. + + You may convey a covered work in object code form under the terms +of sections 4 and 5, provided that you also convey the +machine-readable Corresponding Source under the terms of this License, +in one of these ways: + + a) Convey the object code in, or embodied in, a physical product + (including a physical distribution medium), accompanied by the + Corresponding Source fixed on a durable physical medium + customarily used for software interchange. + + b) Convey the object code in, or embodied in, a physical product + (including a physical distribution medium), accompanied by a + written offer, valid for at least three years and valid for as + long as you offer spare parts or customer support for that product + model, to give anyone who possesses the object code either (1) a + copy of the Corresponding Source for all the software in the + product that is covered by this License, on a durable physical + medium customarily used for software interchange, for a price no + more than your reasonable cost of physically performing this + conveying of source, or (2) access to copy the + Corresponding Source from a network server at no charge. + + c) Convey individual copies of the object code with a copy of the + written offer to provide the Corresponding Source. This + alternative is allowed only occasionally and noncommercially, and + only if you received the object code with such an offer, in accord + with subsection 6b. + + d) Convey the object code by offering access from a designated + place (gratis or for a charge), and offer equivalent access to the + Corresponding Source in the same way through the same place at no + further charge. You need not require recipients to copy the + Corresponding Source along with the object code. If the place to + copy the object code is a network server, the Corresponding Source + may be on a different server (operated by you or a third party) + that supports equivalent copying facilities, provided you maintain + clear directions next to the object code saying where to find the + Corresponding Source. Regardless of what server hosts the + Corresponding Source, you remain obligated to ensure that it is + available for as long as needed to satisfy these requirements. + + e) Convey the object code using peer-to-peer transmission, provided + you inform other peers where the object code and Corresponding + Source of the work are being offered to the general public at no + charge under subsection 6d. + + A separable portion of the object code, whose source code is excluded +from the Corresponding Source as a System Library, need not be +included in conveying the object code work. + + A "User Product" is either (1) a "consumer product", which means any +tangible personal property which is normally used for personal, family, +or household purposes, or (2) anything designed or sold for incorporation +into a dwelling. In determining whether a product is a consumer product, +doubtful cases shall be resolved in favor of coverage. For a particular +product received by a particular user, "normally used" refers to a +typical or common use of that class of product, regardless of the status +of the particular user or of the way in which the particular user +actually uses, or expects or is expected to use, the product. A product +is a consumer product regardless of whether the product has substantial +commercial, industrial or non-consumer uses, unless such uses represent +the only significant mode of use of the product. + + "Installation Information" for a User Product means any methods, +procedures, authorization keys, or other information required to install +and execute modified versions of a covered work in that User Product from +a modified version of its Corresponding Source. The information must +suffice to ensure that the continued functioning of the modified object +code is in no case prevented or interfered with solely because +modification has been made. + + If you convey an object code work under this section in, or with, or +specifically for use in, a User Product, and the conveying occurs as +part of a transaction in which the right of possession and use of the +User Product is transferred to the recipient in perpetuity or for a +fixed term (regardless of how the transaction is characterized), the +Corresponding Source conveyed under this section must be accompanied +by the Installation Information. But this requirement does not apply +if neither you nor any third party retains the ability to install +modified object code on the User Product (for example, the work has +been installed in ROM). + + The requirement to provide Installation Information does not include a +requirement to continue to provide support service, warranty, or updates +for a work that has been modified or installed by the recipient, or for +the User Product in which it has been modified or installed. Access to a +network may be denied when the modification itself materially and +adversely affects the operation of the network or violates the rules and +protocols for communication across the network. + + Corresponding Source conveyed, and Installation Information provided, +in accord with this section must be in a format that is publicly +documented (and with an implementation available to the public in +source code form), and must require no special password or key for +unpacking, reading or copying. + + 7. Additional Terms. + + "Additional permissions" are terms that supplement the terms of this +License by making exceptions from one or more of its conditions. +Additional permissions that are applicable to the entire Program shall +be treated as though they were included in this License, to the extent +that they are valid under applicable law. If additional permissions +apply only to part of the Program, that part may be used separately +under those permissions, but the entire Program remains governed by +this License without regard to the additional permissions. + + When you convey a copy of a covered work, you may at your option +remove any additional permissions from that copy, or from any part of +it. (Additional permissions may be written to require their own +removal in certain cases when you modify the work.) You may place +additional permissions on material, added by you to a covered work, +for which you have or can give appropriate copyright permission. + + Notwithstanding any other provision of this License, for material you +add to a covered work, you may (if authorized by the copyright holders of +that material) supplement the terms of this License with terms: + + a) Disclaiming warranty or limiting liability differently from the + terms of sections 15 and 16 of this License; or + + b) Requiring preservation of specified reasonable legal notices or + author attributions in that material or in the Appropriate Legal + Notices displayed by works containing it; or + + c) Prohibiting misrepresentation of the origin of that material, or + requiring that modified versions of such material be marked in + reasonable ways as different from the original version; or + + d) Limiting the use for publicity purposes of names of licensors or + authors of the material; or + + e) Declining to grant rights under trademark law for use of some + trade names, trademarks, or service marks; or + + f) Requiring indemnification of licensors and authors of that + material by anyone who conveys the material (or modified versions of + it) with contractual assumptions of liability to the recipient, for + any liability that these contractual assumptions directly impose on + those licensors and authors. + + All other non-permissive additional terms are considered "further +restrictions" within the meaning of section 10. If the Program as you +received it, or any part of it, contains a notice stating that it is +governed by this License along with a term that is a further +restriction, you may remove that term. If a license document contains +a further restriction but permits relicensing or conveying under this +License, you may add to a covered work material governed by the terms +of that license document, provided that the further restriction does +not survive such relicensing or conveying. + + If you add terms to a covered work in accord with this section, you +must place, in the relevant source files, a statement of the +additional terms that apply to those files, or a notice indicating +where to find the applicable terms. + + Additional terms, permissive or non-permissive, may be stated in the +form of a separately written license, or stated as exceptions; +the above requirements apply either way. + + 8. Termination. + + You may not propagate or modify a covered work except as expressly +provided under this License. Any attempt otherwise to propagate or +modify it is void, and will automatically terminate your rights under +this License (including any patent licenses granted under the third +paragraph of section 11). + + However, if you cease all violation of this License, then your +license from a particular copyright holder is reinstated (a) +provisionally, unless and until the copyright holder explicitly and +finally terminates your license, and (b) permanently, if the copyright +holder fails to notify you of the violation by some reasonable means +prior to 60 days after the cessation. + + Moreover, your license from a particular copyright holder is +reinstated permanently if the copyright holder notifies you of the +violation by some reasonable means, this is the first time you have +received notice of violation of this License (for any work) from that +copyright holder, and you cure the violation prior to 30 days after +your receipt of the notice. + + Termination of your rights under this section does not terminate the +licenses of parties who have received copies or rights from you under +this License. If your rights have been terminated and not permanently +reinstated, you do not qualify to receive new licenses for the same +material under section 10. + + 9. Acceptance Not Required for Having Copies. + + You are not required to accept this License in order to receive or +run a copy of the Program. Ancillary propagation of a covered work +occurring solely as a consequence of using peer-to-peer transmission +to receive a copy likewise does not require acceptance. However, +nothing other than this License grants you permission to propagate or +modify any covered work. These actions infringe copyright if you do +not accept this License. Therefore, by modifying or propagating a +covered work, you indicate your acceptance of this License to do so. + + 10. Automatic Licensing of Downstream Recipients. + + Each time you convey a covered work, the recipient automatically +receives a license from the original licensors, to run, modify and +propagate that work, subject to this License. You are not responsible +for enforcing compliance by third parties with this License. + + An "entity transaction" is a transaction transferring control of an +organization, or substantially all assets of one, or subdividing an +organization, or merging organizations. If propagation of a covered +work results from an entity transaction, each party to that +transaction who receives a copy of the work also receives whatever +licenses to the work the party's predecessor in interest had or could +give under the previous paragraph, plus a right to possession of the +Corresponding Source of the work from the predecessor in interest, if +the predecessor has it or can get it with reasonable efforts. + + You may not impose any further restrictions on the exercise of the +rights granted or affirmed under this License. For example, you may +not impose a license fee, royalty, or other charge for exercise of +rights granted under this License, and you may not initiate litigation +(including a cross-claim or counterclaim in a lawsuit) alleging that +any patent claim is infringed by making, using, selling, offering for +sale, or importing the Program or any portion of it. + + 11. Patents. + + A "contributor" is a copyright holder who authorizes use under this +License of the Program or a work on which the Program is based. The +work thus licensed is called the contributor's "contributor version". + + A contributor's "essential patent claims" are all patent claims +owned or controlled by the contributor, whether already acquired or +hereafter acquired, that would be infringed by some manner, permitted +by this License, of making, using, or selling its contributor version, +but do not include claims that would be infringed only as a +consequence of further modification of the contributor version. For +purposes of this definition, "control" includes the right to grant +patent sublicenses in a manner consistent with the requirements of +this License. + + Each contributor grants you a non-exclusive, worldwide, royalty-free +patent license under the contributor's essential patent claims, to +make, use, sell, offer for sale, import and otherwise run, modify and +propagate the contents of its contributor version. + + In the following three paragraphs, a "patent license" is any express +agreement or commitment, however denominated, not to enforce a patent +(such as an express permission to practice a patent or covenant not to +sue for patent infringement). To "grant" such a patent license to a +party means to make such an agreement or commitment not to enforce a +patent against the party. + + If you convey a covered work, knowingly relying on a patent license, +and the Corresponding Source of the work is not available for anyone +to copy, free of charge and under the terms of this License, through a +publicly available network server or other readily accessible means, +then you must either (1) cause the Corresponding Source to be so +available, or (2) arrange to deprive yourself of the benefit of the +patent license for this particular work, or (3) arrange, in a manner +consistent with the requirements of this License, to extend the patent +license to downstream recipients. "Knowingly relying" means you have +actual knowledge that, but for the patent license, your conveying the +covered work in a country, or your recipient's use of the covered work +in a country, would infringe one or more identifiable patents in that +country that you have reason to believe are valid. + + If, pursuant to or in connection with a single transaction or +arrangement, you convey, or propagate by procuring conveyance of, a +covered work, and grant a patent license to some of the parties +receiving the covered work authorizing them to use, propagate, modify +or convey a specific copy of the covered work, then the patent license +you grant is automatically extended to all recipients of the covered +work and works based on it. + + A patent license is "discriminatory" if it does not include within +the scope of its coverage, prohibits the exercise of, or is +conditioned on the non-exercise of one or more of the rights that are +specifically granted under this License. You may not convey a covered +work if you are a party to an arrangement with a third party that is +in the business of distributing software, under which you make payment +to the third party based on the extent of your activity of conveying +the work, and under which the third party grants, to any of the +parties who would receive the covered work from you, a discriminatory +patent license (a) in connection with copies of the covered work +conveyed by you (or copies made from those copies), or (b) primarily +for and in connection with specific products or compilations that +contain the covered work, unless you entered into that arrangement, +or that patent license was granted, prior to 28 March 2007. + + Nothing in this License shall be construed as excluding or limiting +any implied license or other defenses to infringement that may +otherwise be available to you under applicable patent law. + + 12. No Surrender of Others' Freedom. + + If conditions are imposed on you (whether by court order, agreement or +otherwise) that contradict the conditions of this License, they do not +excuse you from the conditions of this License. If you cannot convey a +covered work so as to satisfy simultaneously your obligations under this +License and any other pertinent obligations, then as a consequence you may +not convey it at all. For example, if you agree to terms that obligate you +to collect a royalty for further conveying from those to whom you convey +the Program, the only way you could satisfy both those terms and this +License would be to refrain entirely from conveying the Program. + + 13. Use with the GNU Affero General Public License. + + Notwithstanding any other provision of this License, you have +permission to link or combine any covered work with a work licensed +under version 3 of the GNU Affero General Public License into a single +combined work, and to convey the resulting work. The terms of this +License will continue to apply to the part which is the covered work, +but the special requirements of the GNU Affero General Public License, +section 13, concerning interaction through a network will apply to the +combination as such. + + 14. Revised Versions of this License. + + The Free Software Foundation may publish revised and/or new versions of +the GNU General Public License from time to time. Such new versions will +be similar in spirit to the present version, but may differ in detail to +address new problems or concerns. + + Each version is given a distinguishing version number. If the +Program specifies that a certain numbered version of the GNU General +Public License "or any later version" applies to it, you have the +option of following the terms and conditions either of that numbered +version or of any later version published by the Free Software +Foundation. If the Program does not specify a version number of the +GNU General Public License, you may choose any version ever published +by the Free Software Foundation. + + If the Program specifies that a proxy can decide which future +versions of the GNU General Public License can be used, that proxy's +public statement of acceptance of a version permanently authorizes you +to choose that version for the Program. + + Later license versions may give you additional or different +permissions. However, no additional obligations are imposed on any +author or copyright holder as a result of your choosing to follow a +later version. + + 15. Disclaimer of Warranty. + + THERE IS NO WARRANTY FOR THE PROGRAM, TO THE EXTENT PERMITTED BY +APPLICABLE LAW. EXCEPT WHEN OTHERWISE STATED IN WRITING THE COPYRIGHT +HOLDERS AND/OR OTHER PARTIES PROVIDE THE PROGRAM "AS IS" WITHOUT WARRANTY +OF ANY KIND, EITHER EXPRESSED OR IMPLIED, INCLUDING, BUT NOT LIMITED TO, +THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR +PURPOSE. THE ENTIRE RISK AS TO THE QUALITY AND PERFORMANCE OF THE PROGRAM +IS WITH YOU. SHOULD THE PROGRAM PROVE DEFECTIVE, YOU ASSUME THE COST OF +ALL NECESSARY SERVICING, REPAIR OR CORRECTION. + + 16. Limitation of Liability. + + IN NO EVENT UNLESS REQUIRED BY APPLICABLE LAW OR AGREED TO IN WRITING +WILL ANY COPYRIGHT HOLDER, OR ANY OTHER PARTY WHO MODIFIES AND/OR CONVEYS +THE PROGRAM AS PERMITTED ABOVE, BE LIABLE TO YOU FOR DAMAGES, INCLUDING ANY +GENERAL, SPECIAL, INCIDENTAL OR CONSEQUENTIAL DAMAGES ARISING OUT OF THE +USE OR INABILITY TO USE THE PROGRAM (INCLUDING BUT NOT LIMITED TO LOSS OF +DATA OR DATA BEING RENDERED INACCURATE OR LOSSES SUSTAINED BY YOU OR THIRD +PARTIES OR A FAILURE OF THE PROGRAM TO OPERATE WITH ANY OTHER PROGRAMS), +EVEN IF SUCH HOLDER OR OTHER PARTY HAS BEEN ADVISED OF THE POSSIBILITY OF +SUCH DAMAGES. + + 17. Interpretation of Sections 15 and 16. + + If the disclaimer of warranty and limitation of liability provided +above cannot be given local legal effect according to their terms, +reviewing courts shall apply local law that most closely approximates +an absolute waiver of all civil liability in connection with the +Program, unless a warranty or assumption of liability accompanies a +copy of the Program in return for a fee. + + END OF TERMS AND CONDITIONS + + How to Apply These Terms to Your New Programs + + If you develop a new program, and you want it to be of the greatest +possible use to the public, the best way to achieve this is to make it +free software which everyone can redistribute and change under these terms. + + To do so, attach the following notices to the program. It is safest +to attach them to the start of each source file to most effectively +state the exclusion of warranty; and each file should have at least +the "copyright" line and a pointer to where the full notice is found. + + {one line to give the program's name and a brief idea of what it does.} + Copyright (C) {year} {name of author} + + This program is free software: you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation, either version 3 of the License, or + (at your option) any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program. If not, see <http://www.gnu.org/licenses/>. + +Also add information on how to contact you by electronic and paper mail. + + If the program does terminal interaction, make it output a short +notice like this when it starts in an interactive mode: + + {project} Copyright (C) {year} {fullname} + This program comes with ABSOLUTELY NO WARRANTY; for details type `show w'. + This is free software, and you are welcome to redistribute it + under certain conditions; type `show c' for details. + +The hypothetical commands `show w' and `show c' should show the appropriate +parts of the General Public License. Of course, your program's commands +might be different; for a GUI interface, you would use an "about box". + + You should also get your employer (if you work as a programmer) or school, +if any, to sign a "copyright disclaimer" for the program, if necessary. +For more information on this, and how to apply and follow the GNU GPL, see +<http://www.gnu.org/licenses/>. + + The GNU General Public License does not permit incorporating your program +into proprietary programs. If your program is a subroutine library, you +may consider it more useful to permit linking proprietary applications with +the library. If this is what you want to do, use the GNU Lesser General +Public License instead of this License. But first, please read +<http://www.gnu.org/philosophy/why-not-lgpl.html>. diff --git a/README.md b/README.md new file mode 100644 index 0000000..a9d166f --- /dev/null +++ b/README.md @@ -0,0 +1,309 @@ +# neural-style-tf + +This is a TensorFlow implementation of several techniques described in the papers: +* [Image Style Transfer Using Convolutional Neural Networks](http://www.cv-foundation.org/openaccess/content_cvpr_2016/papers/Gatys_Image_Style_Transfer_CVPR_2016_paper.pdf) +by Leon A. Gatys, Alexander S. Ecker, Matthias Bethge +* [Artistic style transfer for videos](https://arxiv.org/abs/1604.08610) +by Manuel Ruder, Alexey Dosovitskiy, Thomas Brox +* [Preserving Color in Neural Artistic Style Transfer](https://arxiv.org/abs/1606.05897) +by Leon A. Gatys, Matthias Bethge, Aaron Hertzmann, Eli Shechtman + +Additionally, techniques are presented for semantic segmentation and multiple style transfer. + +The first paper presents an algorithm for combining the content of one image with the style of another image using convolutional neural networks. Below is an example of transferring the artistic style of [The Starry Night](https://en.wikipedia.org/wiki/The_Starry_Night) onto a photograph of an African lion: + +<p align="center"> +<img src="https://dl.dropboxusercontent.com/u/63267778/examples/lions/42_output.png" width="512"/> +<img src="https://dl.dropboxusercontent.com/u/63267778/examples/content_style.png" width="290"/> +</p> + +Transfering the style of various artworks to the same content image produces qualitatively convincing results: +<p align="center"> +<img src="https://dl.dropboxusercontent.com/u/63267778/examples/lions/32_output.png" width="192"> +<img src="https://dl.dropboxusercontent.com/u/63267778/examples/my_styles/matisse_crop.jpg" width="192"/> +<img src="https://dl.dropboxusercontent.com/u/63267778/examples/lions/33_output.png" width="192"/> +<img src="https://dl.dropboxusercontent.com/u/63267778/examples/my_styles/water_lilies_crop.jpg" width="192"/> + +<img src="https://dl.dropboxusercontent.com/u/63267778/examples/lions/kandinsky_output.png" width="192"/> +<img src="https://dl.dropboxusercontent.com/u/63267778/examples/my_styles/kandinsky_crop.jpg" width="192"/> +<img src="https://dl.dropboxusercontent.com/u/63267778/examples/lions/basquiat_output.png" width="192"/> +<img src="https://dl.dropboxusercontent.com/u/63267778/examples/my_styles/basquiat_crop.jpg" width="192"/> +</p> + +Here we reproduce Figure 2 from the first paper, which renders a photograph of the Tubingen in Germany in the style of 5 different iconic paintings [The Shipwreck of the Minotaur](http://www.artble.com/artists/joseph_mallord_william_turner/paintings/the_shipwreck_of_the_minotaur), [The Starry Night](https://www.wikiart.org/en/vincent-van-gogh/the-starry-night-1889), [Composition VII](https://www.wikiart.org/en/wassily-kandinsky/composition-vii-1913), [The Scream](https://www.wikiart.org/en/edvard-munch/the-scream-1893), [Seated Nude](http://www.pablopicasso.org/seated-nude.jsp): +<p align="center"> +<img src="https://dl.dropboxusercontent.com/u/63267778/examples/gatys_figure/tubingen.png" height="192px"> +<img src="https://dl.dropboxusercontent.com/u/63267778/examples/gatys_figure/tubingen_shipwreck.png" height="192px"> +<img src="https://dl.dropboxusercontent.com/u/63267778/examples/gatys_figure/tubingen_starry_night.png" height="192px"> +<img src="https://dl.dropboxusercontent.com/u/63267778/examples/gatys_figure/tubingen_picasso.png" height="192px"> +<img src="https://dl.dropboxusercontent.com/u/63267778/examples/gatys_figure/1_output.png" height="192px"> +<img src="https://dl.dropboxusercontent.com/u/63267778/examples/gatys_figure/tubingen_kandinsky.png" height="192px"> +</p> + +### Content / Style Tradeoff + +The algorithm allows the user to trade-off the relative weight of the style and content reconstruction terms. + +Here we render with an increasing style weight applied to [Red Canna](http://www.georgiaokeeffe.net/red-canna.jsp): +<p align="center"> +<img src="https://dl.dropboxusercontent.com/u/63267778/examples/style_and_content_tradeoff/okeffe.jpg" height="160px"> +<img src="https://dl.dropboxusercontent.com/u/63267778/examples/style_and_content_tradeoff/okeffe_10.png" width="160px"> +<img src="https://dl.dropboxusercontent.com/u/63267778/examples/style_and_content_tradeoff/okeffe_100.png" width="160px"> +<img src="https://dl.dropboxusercontent.com/u/63267778/examples/style_and_content_tradeoff/okeffe_10000.png" width="160px"> +<img src="https://dl.dropboxusercontent.com/u/63267778/examples/style_and_content_tradeoff/output_1000000.png" width="160px"> +</p> + +### Multiple Style Images +More than one style image can be used to blend multiple artistic styles. + +*Top row (left to right)*: [The Starry Night](https://www.wikiart.org/en/vincent-van-gogh/the-starry-night-1889) + [The Scream](https://www.wikiart.org/en/edvard-munch/the-scream-1893), [The Scream](https://www.wikiart.org/en/edvard-munch/the-scream-1893) + [Composition VII](https://www.wikiart.org/en/wassily-kandinsky/composition-vii-1913), [Seated Nude](http://www.pablopicasso.org/seated-nude.jsp) + [Composition VII](https://www.wikiart.org/en/wassily-kandinsky/composition-vii-1913) +*Bottom row (left to right)*: [Seated Nude](http://www.pablopicasso.org/seated-nude.jsp) + [The Starry Night](https://www.wikiart.org/en/vincent-van-gogh/the-starry-night-1889), [Oversoul](http://alexgrey.com/art/paintings/soul/oversoul/) + [Freshness of Cold](https://afremov.com/FRESHNESS-OF-COLD-PALETTE-KNIFE-Oil-Painting-On-Canvas-By-Leonid-Afremov-Size-30-x40.html), [David Bowie](http://www.francoise-nielly.com/index.php/galerie/index/56) + [Skull](https://www.wikiart.org/en/jean-michel-basquiat/head) + +<p align="center"> +<img src="https://dl.dropboxusercontent.com/u/63267778/examples/multiple_styles/tubingen_starry_scream.png" height="192px"> +<img src="https://dl.dropboxusercontent.com/u/63267778/examples/multiple_styles/tubingen_scream_kandinsky.png" height="192px"> +<img src="https://dl.dropboxusercontent.com/u/63267778/examples/multiple_styles/tubingen_starry_seated.png" height="192px"> +<img src="https://dl.dropboxusercontent.com/u/63267778/examples/multiple_styles/tubingen_seated_kandinsky.png.png" height="192px"> +<img src="https://dl.dropboxusercontent.com/u/63267778/examples/multiple_styles/output_tubingen_afremov_grey.png" height="192px"> +<img src="https://dl.dropboxusercontent.com/u/63267778/examples/multiple_styles/output_basquiat_nielly.png" height="192px"> +</p> + +### Style Interpolation +When using multiple style images, the degree to which they are blended can be controlled. +<p align="center"> +<img src="https://dl.dropboxusercontent.com/u/63267778/examples/style_interpolation/golden_gate_scream_7_starry_3.png" height="192px"> +<img src="https://dl.dropboxusercontent.com/u/63267778/examples/style_interpolation/golden_gate_scream_5_starry_5.png" height="192px"> +<img src="https://dl.dropboxusercontent.com/u/63267778/examples/style_interpolation/golden_gate_scream_3_starry_7.png" height="192px"> +</p> + +### Transfer style but not color +By including the flag `--original_colors` the output image will retain the colors of the original image. + +*Left to right*: content image, stylized image, stylized image with the original colors of the content image +<p align="center"> +<img src="https://dl.dropboxusercontent.com/u/63267778/examples/original_colors/new_york.png" height="165px"> +<img src="https://dl.dropboxusercontent.com/u/63267778/examples/original_colors/stylized.png" height="165px"> +<img src="https://dl.dropboxusercontent.com/u/63267778/examples/original_colors/stylized_original_colors.png" height="165px"> +</p> + +### Textures +The algorithm is not constrained to artistic painting styles. It can also be applied to photographic textures to create [pareidolic](https://en.wikipedia.org/wiki/Pareidolia) images. +<p align="center"> +<img src="https://dl.dropboxusercontent.com/u/63267778/examples/pareidolic/flowers_output.png" width="192px"> +<img src="https://dl.dropboxusercontent.com/u/63267778/examples/my_styles/flowers_crop.jpg" width="192px"/> +<img src="https://dl.dropboxusercontent.com/u/63267778/examples/pareidolic/oil_output.png" width="192px"> +<img src="https://dl.dropboxusercontent.com/u/63267778/examples/my_styles/oil_crop.jpg" width="192px"> + +<img src="https://dl.dropboxusercontent.com/u/63267778/examples/pareidolic/dark_matter_output.png" width="192px"> +<img src="https://dl.dropboxusercontent.com/u/63267778/examples/my_styles/dark_matter_bw.png" width="192px"> +<img src="https://dl.dropboxusercontent.com/u/63267778/examples/pareidolic/ben_giles_output.png" width="192px"> +<img src="https://dl.dropboxusercontent.com/u/63267778/examples/my_styles/ben_giles.png" width="192px"> +</p> + +### Segmentation +Style can be transferred to semantic segmentations in the content image. +<p align="center"> +<img src="https://dl.dropboxusercontent.com/u/63267778/examples/segmentation/00110.jpg" height="192px"> +<img src="https://dl.dropboxusercontent.com/u/63267778/examples/segmentation/00110_mask.png" height="192px"> +<img src="https://dl.dropboxusercontent.com/u/63267778/examples/segmentation/00110_output.png" height="192px"> +<img src="https://dl.dropboxusercontent.com/u/63267778/examples/segmentation/00017.jpg" height="192px"> +<img src="https://dl.dropboxusercontent.com/u/63267778/examples/segmentation/00017_mask.png" height="192px"> +<img src="https://dl.dropboxusercontent.com/u/63267778/examples/segmentation/output_nielly.png" height="192px"> +<img src="https://dl.dropboxusercontent.com/u/63267778/examples/segmentation/00768.jpg" height="192px"> +<img src="https://dl.dropboxusercontent.com/u/63267778/examples/segmentation/00768_mask.png" height="192px"> +<img src="https://dl.dropboxusercontent.com/u/63267778/examples/segmentation/00768_output.png" height="192px"> +<img src="https://dl.dropboxusercontent.com/u/63267778/examples/segmentation/02630.png" height="192px"> +<img src="https://dl.dropboxusercontent.com/u/63267778/examples/segmentation/02630_mask.png" height="192px"> +<img src="https://dl.dropboxusercontent.com/u/63267778/examples/segmentation/02630_output.png" height="192px"> +</p> + +Multiple styles can be transferred to the foreground and background of the content image. + +*Left to right*: content image, foreground style, background style, foreground mask, background mask, stylized image +<p align="center"> +<img src="https://dl.dropboxusercontent.com/u/63267778/examples/segmentation/02390.jpg" height="192px"> +<img src="https://dl.dropboxusercontent.com/u/63267778/examples/segmentation/basquiat.png" height="192px"> +<img src="https://dl.dropboxusercontent.com/u/63267778/examples/segmentation/frida.png" height="192px"> +<img src="https://dl.dropboxusercontent.com/u/63267778/examples/segmentation/02390_mask.png" height="192px"> +<img src="https://dl.dropboxusercontent.com/u/63267778/examples/segmentation/02390_mask_inv.png" height="192px"> +<img src="https://dl.dropboxusercontent.com/u/63267778/examples/segmentation/02390_output.png" height="192px"> +<img src="https://dl.dropboxusercontent.com/u/63267778/examples/segmentation/02270.jpg" height="192px"> +<img src="https://dl.dropboxusercontent.com/u/63267778/examples/segmentation/okeffe_crop.png" height="192px"> +<img src="https://dl.dropboxusercontent.com/u/63267778/examples/segmentation/okeffe_iris.png" height="192px"> +<img src="https://dl.dropboxusercontent.com/u/63267778/examples/segmentation/02270_mask_face.png" height="192px"> +<img src="https://dl.dropboxusercontent.com/u/63267778/examples/segmentation/02270_mask_face_inv.png" height="192px"> +<img src="https://dl.dropboxusercontent.com/u/63267778/examples/segmentation/02270_output.png" height="192px"> +</p> + +### Video +Demo videos coming soon... + +## Setup +#### Dependencies: +* [tensorflow](https://github.com/tensorflow/tensorflow) +* [opencv](http://opencv.org/downloads.html) + +#### Optional (but recommended) dependencies: +* [CUDA](https://developer.nvidia.com/cuda-downloads) 7.5+ +* [cuDNN](https://developer.nvidia.com/cudnn) 5.0+ + +#### After installing the dependencies: +* Download the [VGG-19 model weights](http://www.vlfeat.org/matconvnet/pretrained/) (see the "VGG-VD models from the *Very Deep Convolutional Networks for Large-Scale Visual Recognition* project" section). More info about the VGG-19 network can be found [here](http://www.robots.ox.ac.uk/~vgg/research/very_deep/). +* After downloading, copy the weights file `imagenet-vgg-verydeep-19.mat` to the project directory. + + +## Usage +### Basic Usage + +#### Single Image +1. Copy 1 content image (`.png`, `.jpg`, `.ppm`, `.pgm`) to the default content directory `./content` +2. Copy 1 or more style images (`.png`, `.jpg`, `.ppm`, `.pgm`) to the default style directory `./styles` +3. Run the command: +``` +bash stylize_image.sh <path_to_content_image> <path_to_style_image> +``` +*Example*: +``` +bash stylize_image.sh ./content/lion.jpg ./styles/starry-night.jpg +``` + +*Note*: Paths to images should not contain the `~` character to represent your home directory; you should instead use a relative path or the absolute path. + +#### Video Frames +1. Copy 1 content video (`.mp4`, `.mov`) to the default content directory `./video_input` +2. Copy 1 or more style images to the default style directory `./styles` +3. Run the command: +``` +bash stylize_video.sh <path_to_video> <path_to_style_image> +``` +*Example*: +``` +bash stylize_video.sh ./video_input/video.mp4 ./styles/starry-night.jpg +``` + +### Advanced Usage + +Run the command with specific arguments: +``` +python neural_style.py <arguments> +``` +*Example*: +``` +python neural_style.py --content_img lion.jpg --style_imgs starry-night.jpg --device /cpu:0 +``` + +To use multiple style images, pass a *space-separated* list of the image names and image weights like this: + +`--style_imgs starry_night.jpg the_scream.jpg --style_imgs_weights 0.5 0.5` + +#### Arguments +* `--style_imgs`: Filenames of the style images. To use multiple style images, pass a *space-separated* list. *Example*: `--style_imgs starry-night.jpg` +* `--style_imgs_weights`: The blending weights for each style image. *Default*: `1.0` (assumes only 1 style image) +* `--content_img`: Filename of the content image. *Example*: `lion.jpg` +* `--style_imgs_dir`: Relative or absolute directory path to the style images. *Default*: `./styles` +* `--content_img_dir`: Relative or absolute directory path to the content image. *Default*: `./content` +* `--init_img_type`: Image used to initialize the network. *Choices*: `content`, `random`, `style`. *Default*: `content` +* `--max_size`: Maximum width or height of the input images. *Default*: `512` +* `--content_weight`: Weight for the content loss function. *Default*: `5e0` +* `--style_weight`: Weight for the style loss function. *Default*: `1e4` +* `--tv_weight`: Weight for the total variational loss function. *Default*: `0` +* `--temporal_weight`: Weight for the temporal loss function. *Default*: `2e2` +* `--content_layers`: *Space-separated* VGG19 layer names used for the content image. *Default*: `conv4_2` +* `--style_layers`: *Space-separated* VGG19 layer names used for the style image. *Default*: `relu1_1 relu2_1 relu3_1 relu4_1 relu5_1` +* `--content_layer_weights`: Space-separated weights of each content layer to the content loss. *Default*: `1.0` +* `--style_layer_weights`: Space-separated weights of each style layer to loss. *Default*: `0.2 0.2 0.2 0.2 0.2` +* `--style_scale`: Scale of the style image. Not currently implemented. +* `--original_colors`: Boolean *flag* indicating if the style is transferred but not the colors. +* `--noise_ratio`: Interpolation value between the content image and noise image if network is initialized with `random`. *Default*: `1.0` +* `--seed`: Seed for the random number generator. *Default*: `0` +* `--model_weights`: Weights of the VGG-19 network. Download [here](http://www.vlfeat.org/matconvnet/pretrained/). *Default*:`imagenet-vgg-verydeep-19.mat` +* `--pooling_type`: Type of pooling in convolutional neural network. *Choices*: `avg`, `max`. *Default*: `avg` +* `--device`: GPU or CPU device. GPU mode highly recommended but requires NVIDIA CUDA. *Choices*: `/gpu:0` `/cpu:0`. *Default*: `/gpu:0`. +* `--image_output_dir`: Directory to write output to. *Default*: `./image_output` + +#### Optimization Arguments +* `--optimizer`: Loss minimization optimizer. L-BFGS gives better results. Adam uses less memory. *Choices*: `lbfgs`, `adam`. *Default*: `lbfgs` +* `--learning_rate`: Learning-rate parameter for the Adam optimizer. *Default*: `1e1` +* `--max_iterations`: Max number of iterations for the Adam or L-BFGS optimizer. *Default*: `1e3` +* `--print_iterations`: Number of iterations to print. *Default*: `50` +* `--loss_threshold`: . *Default*: `1e-2` + +#### Video Frame Arguments +* `--video`: Boolean *flag* indicating if the user is creating a video. *Default*: `False` +* `--start_frame`: First frame number. *Default*: `1` +* `--end_frame`: Last frame number. *Default*: `1` +* `--first_frame_type`: Image used to initialize the network during the rendering of the first frame. *Choices*: `content`, `random`, `style`. *Default*: `random` +* `--init_frame_type`: Image used to initialize the network during the every rendering after the first frame. *Choices*: `prev_warped`, `prev`, `content`, `random`, `style`. *Default*: `prev_warped` +* `--video_input_dir`: Relative or absolute directory path to input frames. *Default*: `./video_input` +* `--video_output_dir`: Relative or absolute directory path to write output frames to. *Default*: `./video_output` +* `--content_frame_frmt`: Format string of input frames. *Default*: `frame_{}.png` +* `--backward_optical_flow_frmt`: Format string of backward optical flow files. *Default*: `backward_{}_{}.flo` +* `--forward_optical_flow_frmt`: Format string of forward optical flow files. *Default*: `forward_{}_{}.flo` +* `--content_weights_frmt`: Format string of optical flow consistency files. *Default*: `reliable_{}_{}.txt` +* `--prev_frame_indices`: Previous frames to consider for longterm temporal consistency. *Default*: `1` + +## Questions and Errata + +Send questions or issues to: cysmith1010@gmail.com + +## Memory +By default, `neural-style-tf` uses the NVIDIA cuDNN GPU backend for convolutions and L-BFGS for optimization. +These produce better and faster results, but can consume a lot of memory. You can reduce memory usage with the following: + +* **Use Adam**: Add the flag `--optimizer adam` to use Adam instead of L-BFGS. This should significantly + reduce memory usage, but will require tuning of other parameters for good results; in particular you should + experiment with different values of `--learning_rate`, `--content_weight`, `--style_weight` +* **Reduce image size**: You can reduce the size of the generated image with the `--max_size` argument. + +## Implementation Details +All images were rendered on a machine with: +* **CPU:** Intel Core i7-6800K @ 3.40GHz × 12 +* **GPU:** NVIDIA GeForce GTX 1080/PCIe/SSE2 +* **OS:** Linux Ubuntu 16.04.1 LTS 64-bit + +## Acknowledgements + +The implementation is based on the projects: +* Torch (Lua) implementation 'neural-style' by [jcjohnson](https://github.com/jcjohnson) +* Torch (Lua) implementation 'artistic-videos' by [manuelruder](https://github.com/manuelruder) + +Source video frames were obtained from: +* [MPI Sintel Flow Dataset](http://sintel.is.tue.mpg.de/) + +Souce images and corresponding segmentation masks were obtained from: +* [Automatic Portrait Segmentation for Image Stylization](http://xiaoyongshen.me/webpage_portrait/index.html) + +Artistic images were created by the modern artists: +* [Alex Grey](http://alexgrey.com/) +* [Minjae Lee](http://www.grenomj.com/) +* [Leonid Afremov](https://afremov.com/) +* [Françoise Nielly](http://www.francoise-nielly.com/) +* [James Jean](http://www.jamesjean.com/) +* [Ben Giles](https://benlewisgiles.format.com/) +* [Voka](http://www.voka.at/) + +Artistic images were created by the popular historical artists: +* [Vincent Van Gough](https://www.wikiart.org/en/vincent-van-gogh) +* [Wassily Kandinsky](https://www.wikiart.org/en/wassily-kandinsky) +* [Georgia O'Keeffe](http://www.georgiaokeeffe.net/) +* [Jean-Michel Basquiat](http://basquiat.com/) +* [Édouard Manet](http://www.manet.org/) +* [Pablo Picasso](https://www.wikiart.org/en/pablo-picasso) +* [Joseph Mallord William Turner](https://en.wikipedia.org/wiki/J._M._W._Turner) + +Several Bash shell scripts for testing were created by my brother [Sheldon Smith](http://www.imdb.com/name/nm4328496/). + +## Citation + +If you find this code useful for your research, please cite: + +``` +@misc{Smith2016, + author = {Smith, Cameron}, + title = {neural-style-tf}, + year = {2016}, + publisher = {GitHub}, + journal = {GitHub repository}, + howpublished = {\url{https://github.com/cysmith/neural-style-tf}}, +} +```
\ No newline at end of file diff --git a/neural_style.py b/neural_style.py new file mode 100644 index 0000000..994e2dc --- /dev/null +++ b/neural_style.py @@ -0,0 +1,809 @@ +import matplotlib.pyplot as plt +import tensorflow.python +import tensorflow as tf +import numpy as np +import scipy.io +import argparse +import struct +import time +import cv2 +import csv +import os + +''' + parsing and configuration +''' +def parse_args(): + + desc = "TensorFlow implementation of 'A Neural Algorithm for Artisitc Style'" + parser = argparse.ArgumentParser(description=desc) + + # options for single image + parser.add_argument('--img_name', type=str, + default="result", + help="Basename of output file.") + + parser.add_argument('--style_imgs', nargs='+', type=str, + help='Filenames of the style images (example: starry-night.jpg)', + required=True) + + parser.add_argument('--style_imgs_weights', nargs='+', type=float, + default=[1.0], + help='Interpolation weights of each of the style images. (example: 0.5 0.5)') + + parser.add_argument('--content_img', type=str, + help='Filename of the content image (example: lion.jpg)') + + parser.add_argument('--style_imgs_dir', type=str, + default='./input/styles', + help='Directory path to the style images. (default: %(default)s)') + + parser.add_argument('--content_img_dir', type=str, + default='./input/content', + help='Directory path to the content image. (default: %(default)s)') + + parser.add_argument('--init_img_type', type=str, + default='content', + choices=['random', 'content', 'style'], + help='Image used to initialize the network. (default: %(default)s)') + + parser.add_argument('--max_size', type=int, + default=512, + help='Maximum width or height of the input images. (default: %(default)s)') + + parser.add_argument('--content_weight', type=float, + default=5e0, + help='Weight for the content loss function. (default: %(default)s)') + + parser.add_argument('--style_weight', type=float, + default=1e3, + help='Weight for the style loss function. (default: %(default)s)') + + parser.add_argument('--tv_weight', type=float, + default=0, + help='Weight for the transvariational loss function. Set small (e.g. 1e-3). (default: %(default)s)') + + parser.add_argument('--temporal_weight', type=float, + default=2e2, + help='Weight for the temporal loss function. (default: %(default)s)') + + parser.add_argument('--content_loss_function', type=int, + default=1, + choices=[1, 2, 3], + help='A few different constants for the content layer loss functions have been presented. (default: %(default)s)') + + parser.add_argument('--content_layers', type=str, + default=['conv4_2'], + help='VGG19 layers used for the content image. (default: %(default)s)') + + parser.add_argument('--style_layers', nargs='+', type=str, + default=['relu1_1', 'relu2_1', 'relu3_1', 'relu4_1', 'relu5_1'], + help='VGG19 layers used for the style image. (default: %(default)s)') + + parser.add_argument('--content_layer_weights', type=float, + default=[1.0], + help='Contributions (weights) of each content layer to loss. (default: %(default)s)') + + parser.add_argument('--style_layer_weights', nargs='+', type=float, + default=[0.2, 0.2, 0.2, 0.2, 0.2], + help='Contributions (weights) of each style layer to loss. (default: %(default)s)') + + parser.add_argument('--style_scale', type=float, default=1.0) + + parser.add_argument('--is_original_colors', type=bool, + default=False, + help='Transfer the style but not the colors. (default: %(default)s)') + + parser.add_argument('--has_style_mask', type=bool, + default=False, + help='Transfer the style to masked regions.') + + parser.add_argument('--style_mask_imgs', nargs='+', type=str, + default=None, + help='Filenames of the style mask images (example: face_mask.png)') + + parser.add_argument('--noise_ratio', type=float, default=1.0) + + parser.add_argument('--seed', type=int, + default=0, + help='Seed for the random number generator. (default: %(default)s)') + + parser.add_argument('--model_weights', type=str, + default='imagenet-vgg-verydeep-19.mat') + + parser.add_argument('--pooling_type', type=str, + default='avg', + choices=['avg', 'max'], + help="Type of pooling in convolutional neural network. (default: %(default)s)") + + parser.add_argument('--device', type=str, + default='/gpu:0', + choices=['/gpu:0', '/cpu:0'], + help='GPU or CPU mode. GPU mode requires NVIDIA CUDA. (default|recommended: %(default)s)') + + parser.add_argument('--image_output_dir', type=str, + default='./image_output', + help='Relative or absolute directory path to output image and data.') + + # optimizations + parser.add_argument('--optimizer', type=str, + default='lbfgs', + choices=['lbfgs', 'adam'], + help='Loss minimization optimizer. L-BFGS gives better results. Adam uses less memory. (default|recommended: %(default)s)') + + parser.add_argument('--learning_rate', type=float, + default=1e1, + help='Learning rate parameter for the Adam optimizer. (default: %(default)s)') + + parser.add_argument('--max_iterations', type=int, + default=1e3, + help='Max number of iterations for the Adam or L-BFGS optimizer. (default: %(default)s)') + + parser.add_argument('--verbose', action='store_true', + help="Boolean flag indicating if print statements should be included during execution.") + + # options for video frames + parser.add_argument('--is_video', action='store_true', + help='Boolean flag indicating if the user is generating a video. (default=%(default)s)') + + parser.add_argument('--start_frame', type=int, default=1, + help='First frame number.') + + parser.add_argument('--end_frame', type=int, default=1, + help='Last frame number.') + + parser.add_argument('--first_frame_type', type=str, + choices=['random', 'content', 'style'], + default='content', + help='Image used to initialize the network during the rendering of the first frame.') + + parser.add_argument('--init_frame_type', type=str, + choices=['prev_warped', 'prev', 'random', 'content', 'style'], + default='prev_warped', + help='Image used to initialize the network during the every rendering after the first frame.') + + parser.add_argument('--video_input_dir', type=str, + default='./video_input', + help='Relative or absolute directory path to input frames.') + + parser.add_argument('--video_output_dir', type=str, + default='./video_output', + help='Relative or absolute directory path to output frames.') + + parser.add_argument('--content_frame_frmt', type=str, + default='frame_{}.ppm', + help='Filename format of the input content frames.') + + parser.add_argument('--backward_optical_flow_frmt', type=str, + default='backward_{}_{}.flo', + help='Filename format of the backward optical flow files.') + + parser.add_argument('--forward_optical_flow_frmt', type=str, + default='forward_{}_{}.flo', + help='Filename format of the forward optical flow files') + + parser.add_argument('--content_weights_frmt', type=str, + default='reliable_{}_{}.txt', + help='Filename format of the optical flow consistency files.') + + parser.add_argument('--prev_frame_indices', nargs='+', type=int, + default=[1], + help='Previous frames to consider for longterm temporal consistency.') + + args = parser.parse_args() + + # create directories for output + if args.is_video: + maybe_make_directory(args.video_output_dir) + else: + maybe_make_directory(args.image_output_dir) + + return args + +''' + pre-trained vgg19 convolutional neural network + + remark: layers are manually initialized for clarity. +''' +vgg19_mean = np.array([123.68, 116.779, 103.939]).reshape((1,1,1,3)) + +def build_vgg19(input_img): + if args.verbose: print("\nBUILDING VGG-19 NETWORK") + net = {} + _, h, w, d = input_img.shape + + if args.verbose: print('loading model weights...') + vgg_rawnet = scipy.io.loadmat(args.model_weights) + vgg_layers = vgg_rawnet['layers'][0] + if args.verbose: print('constructing layers...') + net['input'] = tf.Variable(np.zeros((1, h, w, d), dtype=np.float32)) + + if args.verbose: print('LAYER GROUP 1') + net['conv1_1'] = conv_layer('conv1_1', net['input'], W=get_weights(vgg_layers, 0)) + net['relu1_1'] = relu_layer('relu1_1', net['conv1_1'], b=get_bias(vgg_layers, 0)) + + net['conv1_2'] = conv_layer('conv1_2', net['relu1_1'], W=get_weights(vgg_layers, 2)) + net['relu1_2'] = relu_layer('relu1_2', net['conv1_2'], b=get_bias(vgg_layers, 2)) + + net['pool1'] = pool_layer('pool1', net['relu1_2']) + + if args.verbose: print('LAYER GROUP 2') + net['conv2_1'] = conv_layer('conv2_1', net['pool1'], W=get_weights(vgg_layers, 5)) + net['relu2_1'] = relu_layer('relu2_1', net['conv2_1'], b=get_bias(vgg_layers, 5)) + + net['conv2_2'] = conv_layer('conv2_2', net['relu2_1'], W=get_weights(vgg_layers, 7)) + net['relu2_2'] = relu_layer('relu2_2', net['conv2_2'], b=get_bias(vgg_layers, 7)) + + net['pool2'] = pool_layer('pool2', net['relu2_2']) + + if args.verbose: print('LAYER GROUP 3') + net['conv3_1'] = conv_layer('conv3_1', net['pool2'], W=get_weights(vgg_layers, 10)) + net['relu3_1'] = relu_layer('relu3_1', net['conv3_1'], b=get_bias(vgg_layers, 10)) + + net['conv3_2'] = conv_layer('conv3_2', net['relu3_1'], W=get_weights(vgg_layers, 12)) + net['relu3_2'] = relu_layer('relu3_2', net['conv3_2'], b=get_bias(vgg_layers, 12)) + + net['conv3_3'] = conv_layer('conv3_3', net['relu3_2'], W=get_weights(vgg_layers, 14)) + net['relu3_3'] = relu_layer('relu3_3', net['conv3_3'], b=get_bias(vgg_layers, 14)) + + net['conv3_4'] = conv_layer('conv3_4', net['relu3_3'], W=get_weights(vgg_layers, 16)) + net['relu3_4'] = relu_layer('relu3_4', net['conv3_4'], b=get_bias(vgg_layers, 16)) + + net['pool3'] = pool_layer('pool3', net['relu3_4']) + + if args.verbose: print('LAYER GROUP 4') + net['conv4_1'] = conv_layer('conv4_1', net['pool3'], W=get_weights(vgg_layers, 19)) + net['relu4_1'] = relu_layer('relu4_1', net['conv4_1'], b=get_bias(vgg_layers, 19)) + + net['conv4_2'] = conv_layer('conv4_2', net['relu4_1'], W=get_weights(vgg_layers, 21)) + net['relu4_2'] = relu_layer('relu4_2', net['conv4_2'], b=get_bias(vgg_layers, 21)) + + net['conv4_3'] = conv_layer('conv4_3', net['relu4_2'], W=get_weights(vgg_layers, 23)) + net['relu4_3'] = relu_layer('relu4_3', net['conv4_3'], b=get_bias(vgg_layers, 23)) + + net['conv4_4'] = conv_layer('conv4_4', net['relu4_3'], W=get_weights(vgg_layers, 25)) + net['relu4_4'] = relu_layer('relu4_4', net['conv4_4'], b=get_bias(vgg_layers, 25)) + + net['pool4'] = pool_layer('pool4', net['relu4_4']) + + if args.verbose: print('LAYER GROUP 5') + net['conv5_1'] = conv_layer('conv5_1', net['pool4'], W=get_weights(vgg_layers, 28)) + net['relu5_1'] = relu_layer('relu5_1', net['conv5_1'], b=get_bias(vgg_layers, 28)) + + net['conv5_2'] = conv_layer('conv5_2', net['relu5_1'], W=get_weights(vgg_layers, 30)) + net['relu5_2'] = relu_layer('relu5_2', net['conv5_2'], b=get_bias(vgg_layers, 30)) + + net['conv5_3'] = conv_layer('conv5_3', net['relu5_2'], W=get_weights(vgg_layers, 32)) + net['relu5_3'] = relu_layer('relu5_3', net['conv5_3'], b=get_bias(vgg_layers, 32)) + + net['conv5_4'] = conv_layer('conv5_4', net['relu5_3'], W=get_weights(vgg_layers, 34)) + net['relu5_4'] = relu_layer('relu5_4', net['conv5_4'], b=get_bias(vgg_layers, 34)) + + net['pool5'] = pool_layer('pool5', net['relu5_4']) + + return net + +def conv_layer(layer_name, layer_input, W): + conv = tf.nn.conv2d(layer_input, W, strides=[1, 1, 1, 1], padding='SAME') + if args.verbose: print('--{} | shape={} | weights_shape={}'.format(layer_name, + conv.get_shape(), W.get_shape())) + return conv + +def relu_layer(layer_name, layer_input, b): + relu = tf.nn.relu(layer_input + b) + if args.verbose: + print('--{} | shape={} | bias_shape={}'.format(layer_name, relu.get_shape(), + b.get_shape())) + return relu + +def pool_layer(layer_name, layer_input): + if args.pooling_type == 'avg': + pool = tf.nn.avg_pool(layer_input, ksize=[1, 2, 2, 1], + strides=[1, 2, 2, 1], padding='SAME') + elif args.pooling_type == 'max': + pool = tf.nn.max_pool(layer_input, ksize=[1, 2, 2, 1], + strides=[1, 2, 2, 1], padding='SAME') + if args.verbose: + print('--{} | shape={}'.format(layer_name, pool.get_shape())) + return pool + +def get_weights(vgg_layers, i): + weights = vgg_layers[i][0][0][2][0][0] + W = tf.constant(weights) + return W + +def get_bias(vgg_layers, i): + bias = vgg_layers[i][0][0][2][0][1] + b = tf.constant(np.reshape(bias, (bias.size))) + return b + +''' + 'a neural algorithm for artistic style' loss functions +''' +def content_layer_loss(p, x): + _, h, w, d = p.get_shape() + M = h.value * w.value + N = d.value + loss = (1./(2 * N**0.5 * M**0.5 )) * tf.reduce_sum(tf.pow((x - p), 2)) + #loss = (1./2.) * tf.reduce_sum(tf.pow((x - p), 2)) + #loss = (1./(N * M)) * tf.reduce_sum(tf.pow((x - p), 2)) + return loss + +def gram_matrix(x, area, depth): + F = tf.reshape(x[0], (area, depth)) + G = tf.matmul(tf.transpose(F), F) + return G + +def style_layer_loss(a, x): + _, h, w, d = a.get_shape() + M = h.value * w.value + N = d.value + A = gram_matrix(a, M, N) + G = gram_matrix(x, M, N) + loss = (1./(4 * N**2 * M**2)) * tf.reduce_sum(tf.pow((G - A), 2)) + return loss + +def mask_style_layer(a, x, mask_img): + _, h, w, d = a.get_shape() + mask = get_mask_image(mask_img, w.value, h.value) + mask = tf.convert_to_tensor(mask) + tensors = [] + for _ in range(d.value): + tensors.append(mask) + mask = tf.pack(tensors, axis=2) + mask = tf.pack(mask, axis=0) + mask = tf.expand_dims(mask, 0) + a = tf.mul(a, mask) + x = tf.mul(x, mask) + return a, x + +def sum_masked_style_losses(sess, net, style_imgs): + total_style_loss = 0. + weights = args.style_imgs_weights + masks = args.style_mask_imgs + for img, img_weight, img_mask in zip(style_imgs, weights, masks): + sess.run(net['input'].assign(img)) + style_loss = 0. + for layer, weight in zip(args.style_layers, args.style_layer_weights): + a = sess.run(net[layer]) + x = net[layer] + a = tf.convert_to_tensor(a) + a, x = mask_style_layer(a, x, img_mask) + style_loss += style_layer_loss(a, x) * weight + style_loss /= float(len(args.style_layers)) + total_style_loss += (style_loss * img_weight) + total_style_loss /= float(len(style_imgs)) + return total_style_loss + +def sum_style_losses(sess, net, style_imgs): + total_style_loss = 0. + weights = args.style_imgs_weights + for img, img_weight in zip(style_imgs, weights): + sess.run(net['input'].assign(img)) + style_loss = 0. + for layer, weight in zip(args.style_layers, args.style_layer_weights): + a = sess.run(net[layer]) + x = net[layer] + a = tf.convert_to_tensor(a) + style_loss += style_layer_loss(a, x) * weight + style_loss /= float(len(args.style_layers)) + total_style_loss += (style_loss * img_weight) + total_style_loss /= float(len(style_imgs)) + return total_style_loss + +def sum_content_losses(sess, net, content_img): + sess.run(net['input'].assign(content_img)) + content_loss = 0. + for layer, weight in zip(args.content_layers, args.content_layer_weights): + p = sess.run(net[layer]) + x = net[layer] + p = tf.convert_to_tensor(p) + x = tf.convert_to_tensor(x) + content_loss += content_layer_loss(p, x) * weight + content_loss /= float(len(args.content_layers)) + return content_loss + +''' + 'artistic style transfer for videos' loss functions +''' +def temporal_loss(x, w, c): + c = c[np.newaxis,:,:,:] + D = float(x.size) + loss = (1. / D) * tf.reduce_sum(c * tf.nn.l2_loss(x - w)) + loss = tf.cast(loss, tf.float32) + return loss + +def get_longterm_weights(i, j): + c_sum = 0. + for k in range(args.prev_frame_indices): + if i - k > i - j: + c_sum += get_content_weights(i, i - k) + c = get_content_weights(i, i - j) + c_max = tf.maximum(c - c_sum, 0.) + return c_max + +def sum_longterm_temporal_losses(net, frame, x): + loss = 0. + for j in range(args.prev_frame_indices): + prev_frame = frame - j + w = get_prev_warped_frame(frame) + c = get_longterm_weights(frame, prev_frame) + loss += temporal_loss(x, w, c) + return loss + +def sum_shortterm_temporal_losses(net, frame, x): + prev_frame = frame - 1 + w = get_prev_warped_frame(frame) + c = get_content_weights(frame, prev_frame) + loss = temporal_loss(x, w, c) + return loss + +''' + denoising loss function + + remark: not convinced this does anything significant. +''' +def sum_total_variation_losses(x): + b, h, w, d = x.shape + tv_y_size = b * (h-1) * w * d + tv_x_size = b * h * (w-1) * d + loss_y = tf.nn.l2_loss(x[:,1:,:,:] - x[:,:h-1,:,:]) + loss_y /= tv_y_size + loss_x = tf.nn.l2_loss(x[:,:,1:,:] - x[:,:,:w-1,:]) + loss_x /= tv_x_size + loss = 2 * (loss_y + loss_x) + loss = tf.cast(loss, tf.float32) + return loss + +''' + utilities and i/o +''' +def read_image(path): + # BGR image + img = cv2.imread(path, cv2.IMREAD_COLOR).astype('float') + img = preprocess(img, vgg19_mean) + return img + +def write_image(path, img): + img = postprocess(img, vgg19_mean) + cv2.imwrite(path, img) + +def preprocess(img, mean): + # BGR to RGB + img = img[...,::-1] + # shape (H, W, D) to (1, H, W, D) + img = img[np.newaxis,:,:,:] + # subtract mean + img -= mean + return img + +def postprocess(img, mean): + # add mean + img += mean + # shape (1, H, W, D) to (H, W, D) + img = img[0] + img = np.clip(img, 0, 255).astype('uint8') + # RGB to BGR + img = img[...,::-1] + return img + +def read_flow_file(path): + with open(path, "rb") as f: + # 4 bytes header + header = struct.unpack('4s', f.read(4))[0] + # 4 bytes width, height + w = struct.unpack('i', f.read(4))[0] + h = struct.unpack('i', f.read(4))[0] + flow = np.ndarray((2, h, w), dtype=np.float32) + for y in range(h): + for x in range(w): + flow[1,y,x] = struct.unpack('f', f.read(4))[0] + flow[0,y,x] = struct.unpack('f', f.read(4))[0] + return flow + +def read_weights_file(path): + lines = open(path).read().splitlines() + header = map(int, lines[0].split(' ')) + w = header[0] + h = header[1] + vals = np.zeros((h, w), dtype=np.float32) + for i in range(1, len(lines)): + line = lines[i].rstrip().split(' ') + vals[i-1] = np.array(map(np.float32, line)) + vals[i-1] = map(lambda x: 0. if x < 255. else 1., vals[i-1]) + # expand to 3 channels + weights = np.dstack([vals.astype(np.float32)] * 3) + return weights + +def maybe_make_directory(dir_path): + if not os.path.exists(dir_path): + os.makedirs(dir_path) + +''' + rendering -- where the magic happens +''' +def stylize(content_img, style_imgs, init_img, frame=None): + with tf.device(args.device), tf.Session() as sess: + # setup network + net = build_vgg19(content_img) + + # style loss + if args.has_style_mask: + L_style = sum_masked_style_losses(sess, net, style_imgs) + else: + L_style = sum_style_losses(sess, net, style_imgs) + + # content loss + L_content = sum_content_losses(sess, net, content_img) + + # denoising loss + L_tv = sum_total_variation_losses(init_img) + + # loss weights + alpha = args.content_weight + beta = args.style_weight + theta = args.tv_weight + + # total loss + L_total = alpha * L_content + L_total += beta * L_style + L_total += theta * L_tv + + if args.is_video and frame > 1: + gamma = args.temporal_weight + L_temporal = sum_shortterm_temporal_losses(sess, frame, init_img) + L_total += gamma * L_temporal + + # optimization algorithm + optimizer = get_optimizer(L_total) + + if args.optimizer == 'adam': + minimize_with_adam(sess, net, optimizer, init_img) + elif args.optimizer == 'lbfgs': + minimize_with_lbfgs(sess, net, optimizer, init_img) + + output_img = sess.run(net['input']) + + if args.is_original_colors: + output_img = convert_to_original_colors(np.copy(content_img), np.copy(output_img)) + + if args.is_video: + write_video_output(frame, output_img) + else: + write_image_output(output_img, content_img, style_imgs, init_img) + +def minimize_with_lbfgs(sess, net, optimizer, init_img): + if args.verbose: print('MINIMIZING LOSS USING: L-BFGS OPTIMIZER') + init_op = tf.initialize_all_variables() + sess.run(init_op) + sess.run(net['input'].assign(init_img)) + optimizer.minimize(sess) + +def minimize_with_adam(sess, net, optimizer, init_img): + if args.verbose: print('MINIMIZING LOSS USING: ADAM OPTIMIZER') + train_op = optimizer.minimize(L_total) + init_op = tf.initialize_all_variables() + sess.run(init_op) + sess.run(net['input'].assign(init_img)) + iterations = 0 + while (iterations < args.max_iterations): + sess.run(train_op) + iterations += 1 + +def get_optimizer(loss): + if args.optimizer == 'lbfgs': + optimizer = tf.contrib.opt.ScipyOptimizerInterface( + loss, + method='L-BFGS-B', + options={'maxiter': args.max_iterations}) + elif args.optimizer == 'adam': + optimizer = tf.train.AdamOptimizer(args.learning_rate, epsilon=1.0) + return optimizer + +def write_video_output(frame, output_img): + output_frame_fn = args.content_frame_frmt.format(str(frame).zfill(4)) + output_frame_path = os.path.join(args.video_output_dir, output_frame_fn) + write_image(output_frame_path, output_img) + +def write_image_output(output_img, content_img, style_imgs, init_img): + out_dir = os.path.join(args.image_output_dir, args.img_name) + maybe_make_directory(out_dir) + img_path = os.path.join(out_dir, "output.png") + content_path = os.path.join(out_dir, "content.png") + init_path = os.path.join(out_dir, "init.png") + + write_image(img_path, output_img) + write_image(content_path, content_img) + write_image(init_path, init_img) + index = 0 + for style_img in style_imgs: + path = os.path.join(out_dir, str(index)+"_style.png") + write_image(path, style_img) + index += 1 + + # save the configuration settings + out_file = os.path.join(out_dir, "meta_data.txt") + f = open(out_file, "w") + f.write("image name: {}\n".format(args.img_name)) + f.write("content: {}\n".format(args.content_img)) + index = 0 + for style_img, weight in zip(args.style_imgs, args.style_imgs_weights): + f.write("styles ["+str(index)+"]: {} * {}\n".format(weight, style_img)) + index = 0 + if args.style_mask_imgs is not None: + for mask in args.style_mask_imgs: + f.write("style masks ["+str(index)+"]: {}\n".format(mask)) + f.write("init_type: {}\n".format(args.init_img_type)) + f.write("content_weight: {}\n".format(args.content_weight)) + f.write("style_weight: {}\n".format(args.style_weight)) + f.write("tv_weight: {}\n".format(args.tv_weight)) + f.write("content_layers: {}\n".format(args.content_layers)) + f.write("style_layers: {}\n".format(args.style_layers)) + f.write("optimizer_type: {}\n".format(args.optimizer)) + f.write("max_iterations: {}\n".format(args.max_iterations)) + f.write("max_image_size: {}\n".format(args.max_size)) + f.close() + +''' + image loading and processing +''' +def get_init_image(init_type, content_img, style_img, frame=None): + if init_type == 'content': + return content_img + elif init_type == 'style': + return style_img + elif init_type == 'random': + init_img = get_noise_image(args.noise_ratio, content_img) + return init_img + # only for video frames + elif init_type == 'prev': + init_img = get_prev_frame(frame) + return init_img + elif init_type == 'prev_warped': + init_img = get_prev_warped_frame(frame) + return init_img + +def get_content_frame(frame): + content_fn = args.content_frame_frmt.format(str(frame).zfill(4)) + content_path = os.path.join(args.video_input_dir, content_fn) + img = read_image(content_path) + return img + +def get_content_image(content_img): + # BGR image + path = os.path.join(args.content_img_dir, content_img) + img = cv2.imread(path, cv2.IMREAD_COLOR).astype('float') + h, w, d = img.shape + mx = args.max_size + # resize if > max size + if h > w and h > mx: + w = (float(mx) / float(h)) * w + img = cv2.resize(img, dsize=(int(w), mx), interpolation=cv2.INTER_CUBIC) + if w > mx: + h = (float(mx) / float(w)) * h + img = cv2.resize(img, dsize=(mx, int(h)), interpolation=cv2.INTER_CUBIC) + img = preprocess(img, vgg19_mean) + return img + +def get_style_images(content_img, scale): + style_imgs = [] + for style_fn in args.style_imgs: + path = os.path.join(args.style_imgs_dir, style_fn) + # BGR image + img = cv2.imread(path, cv2.IMREAD_COLOR).astype(np.float32) + _, h, w, d = content_img.shape + img = cv2.resize(img, dsize=(int(w*scale), int(h*scale))) + img = preprocess(img, vgg19_mean) + style_imgs.append(img) + return style_imgs + +def get_noise_image(noise_ratio, content_img): + np.random.seed(args.seed) + noise_img = np.random.uniform(-20., 20., content_img.shape).astype(np.float32) + img = noise_ratio * noise_img + (1.-noise_ratio) * content_img + return img + +def get_mask_image(mask_img, width, height): + path = os.path.join(args.content_img_dir, mask_img) + img = cv2.imread(path, cv2.IMREAD_GRAYSCALE) + img = cv2.resize(img, dsize=(width, height)).astype(np.float32) + mx = np.amax(img) + img /= mx + return img + +def get_prev_frame(frame): + # previously stylized frame + prev_frame = frame - 1 + prev_frame_fn = args.content_frame_frmt.format(str(prev_frame).zfill(4)) + prev_frame_path = os.path.join(args.video_output_dir, prev_frame_fn) + img = cv2.imread(prev_frame_path, cv2.IMREAD_COLOR) + return img + +def get_prev_warped_frame(frame): + prev_img = get_prev_frame(frame) + prev_frame = frame - 1 + # backwards flow: current frame -> previous frame + flow_fn = args.backward_optical_flow_frmt.format(str(frame), str(prev_frame)) + flow_path = os.path.join(args.video_input_dir, flow_fn) + flow = read_flow_file(flow_path) + warped_img = warp_image(prev_img, flow).astype('float32') + img = preprocess(warped_img, vgg19_mean) + return img + +def get_content_weights(frame, prev_frame): + forward_fn = args.content_weights_frmt.format(str(prev_frame), str(frame)) + backward_fn = args.content_weights_frmt.format(str(frame), str(prev_frame)) + forward_path = os.path.join(args.video_input_dir, forward_fn) + backward_path = os.path.join(args.video_input_dir, backward_fn) + forward_weights = read_weights_file(forward_path) + backward_weights = read_weights_file(backward_path) + forward_weights = np.clip(forward_weights, 0, 255).astype('uint8') + backward_weights = np.clip(backward_weights, 0, 255).astype('uint8') + return forward_weights #, backward_weights + +def warp_image(src, flow): + _, h, w = flow.shape + flow_map = np.zeros(flow.shape, dtype=np.float32) + for y in range(h): + flow_map[1,y,:] = float(y) + flow[1,y,:] + for x in range(w): + flow_map[0,:,x] = float(x) + flow[0,:,x] + # remap pixels to optical flow + dst = cv2.remap( + src, flow_map[0], flow_map[1], + interpolation=cv2.INTER_CUBIC, + borderMode=cv2.BORDER_TRANSPARENT) + return dst + +def convert_to_original_colors(content_img, stylized_img): + content_img = postprocess(content_img, vgg19_mean) + stylized_img = postprocess(stylized_img, vgg19_mean) + content_yuv = cv2.cvtColor(content_img, cv2.COLOR_BGR2YUV) + stylized_yuv = cv2.cvtColor(stylized_img, cv2.COLOR_BGR2YUV) + y, _, _ = cv2.split(stylized_yuv) + _, u, v = cv2.split(content_yuv) + merged = cv2.merge((y, u, v)) + dst = cv2.cvtColor(merged, cv2.COLOR_YUV2BGR).astype('float') + dst = preprocess(dst, vgg19_mean) + return dst + +def render_single_image(): + content_img = get_content_image(args.content_img) + style_imgs = get_style_images(content_img, args.style_scale) + with tf.Graph().as_default(): + print('\n---- RENDERING SINGLE IMAGE ----\n') + init_img = get_init_image(args.init_img_type, content_img, style_imgs) + tick = time.time() + stylize(content_img, style_imgs, init_img) + tock = time.time() + print('Single image elapsed time: {}'.format(tock - tick)) + +def render_video(): + for frame in range(args.start_frame, args.end_frame+1): + with tf.Graph().as_default(): + print('\n---- RENDERING VIDEO FRAME: {}/{} ----\n'.format(frame, args.end_frame)) + if frame == 1: + content_frame = get_content_frame(frame) + style_imgs = get_style_images(content_frame, args.style_scale) + init_img = get_init_image(args.first_frame_type, content_frame, style_imgs, frame) + tick = time.time() + stylize(content_frame, style_imgs, init_img, frame) + tock = time.time() + print('Frame {} elapsed time: {}'.format(frame, tock - tick)) + else: + content_frame = get_content_frame(frame) + style_imgs = get_style_images(content_frame, args.style_scale) + init_img = get_init_image(args.init_frame_type, content_frame, style_imgs, frame) + tick = time.time() + stylize(content_frame, style_imgs, init_img, frame) + tock = time.time() + print('Frame {} elapsed time: {}'.format(frame, tock - tick)) + +def main(): + global args + args = parse_args() + if args.is_video: render_video() + else: render_single_image() + +if __name__ == '__main__': + main()
\ No newline at end of file diff --git a/stylize_image.sh b/stylize_image.sh new file mode 100644 index 0000000..7c6f021 --- /dev/null +++ b/stylize_image.sh @@ -0,0 +1,43 @@ +set -e +# Get a carriage return into `cr` +cr=`echo $'\n.'` +cr=${cr%.} + +if [ "$#" -le 1 ]; then + echo "Usage: bash stylize_image.sh <path_to_content_image> <path_to_style_image>" + exit 1 +fi + +echo "" +read -p "Did you install the required dependencies? [y/n] $cr > " dependencies + +if [ "$dependencies" != "y" ]; then + echo "Error: Requires dependencies: tensorflow, opencv2 (python), scipy" + exit 1; +fi + +echo "" +read -p "Do you have a CUDA enabled GPU? [y/n] $cr > " cuda + +if [ "$cuda" != "y" ]; then + device='/cpu:0' +else + device='/gpu:0' +fi + +# Parse arguments +content_image="$1" +content_dir=$(dirname "$content_image") +content_filename=$(basename "$content_image") + +style_image="$2" +style_dir=$(dirname "$style_image" ) +style_filename=$(basename "$style_image") + +echo "Rendering stylized image. This may take a while..." +python neural_style.py \ +--content_img "${content_filename}" \ +--content_img_dir "${content_dir}" \ +--style_imgs "${style_filename}" \ +--style_imgs_dir "${style_dir}" \ +--device "${device}";
\ No newline at end of file diff --git a/stylize_video.sh b/stylize_video.sh new file mode 100644 index 0000000..cb8698c --- /dev/null +++ b/stylize_video.sh @@ -0,0 +1,84 @@ +set -e +# Get a carriage return into `cr` +cr=`echo $'\n.'` +cr=${cr%.} + +# Find out whether ffmpeg or avconv is installed on the system +FFMPEG=ffmpeg +command -v $FFMPEG >/dev/null 2>&1 || { + FFMPEG=avconv + command -v $FFMPEG >/dev/null 2>&1 || { + echo >&2 "This script requires either ffmpeg or avconv installed. Aborting."; exit 1; + } +} + +if [ "$#" -le 1 ]; then + echo "Usage: bash stylize_video.sh <path_to_video> <path_to_style_image>" + exit 1 +fi + +echo "" +read -p "Did you install the required dependencies? [y/n] $cr > " dependencies + +if [ "$dependencies" != "y" ]; then + echo "Error: Requires dependencies: tensorflow, opencv2 (python), scipy" + exit 1; +fi + +echo "" +read -p "Do you have a CUDA enabled GPU? [y/n] $cr > " cuda + +if [ "$cuda" != "y" ]; then + echo "Error: GPU required to render videos in a feasible amount of time." + exit 1; +fi + +# Parse arguments +content_video="$1" +content_dir=$(dirname "$content_video") +content_filename=$(basename "$content_video") +extension="${content_filename##*.}" +content_filename="${content_filename%.*}" +content_filename=${content_filename//[%]/x} + +style_image="$2" +style_dir=$(dirname "$style_image") +style_filename=$(basename "$style_image") + +temp_dir="./video_input/${content_filename}" + +# Create output folder +mkdir -p "$temp_dir" + +# Save frames of the video as individual image files +$FFMPEG -v quiet -i "$1" "${temp_dir}/frame_%04d.ppm" +eval $(ffprobe -v error -of flat=s=_ -select_streams v:0 -show_entries stream=width,height "$1") +width="${streams_stream_0_width}" +height="${streams_stream_0_height}" +if [ "$width" -gt "$height" ]; then + max_size="$width" +else + max_size="$height" +fi +num_frames=$(find "$temp_dir" -iname "*.ppm" | wc -l) + +echo "Computing optical flow. This will take a while..." +cd ./video_input +bash make-opt-flow.sh ${content_filename}/frame_%04d.ppm ${content_filename} +cd .. + +echo "Rendering stylized video frames. This will take a while..." +python neural_style.py --is_video \ +--video_input_dir "${temp_dir}" \ +--style_imgs_dir "${style_dir}" \ +--style_imgs "${style_filename}" \ +--end_frame "$num_frames" \ +--max_size "${max_size}" \ +--verbose; + +# Create video from output images. +echo "Converting image sequence to video. This should be quick..." +$FFMPEG -v quiet -i ./video_output/frame_%04d.ppm ./video_output/${content_filename}-stylized.$extension + +# Clean up garbage +rm -rf "${temp_dir}"
\ No newline at end of file diff --git a/video_input/consistencyChecker/CFilter.h b/video_input/consistencyChecker/CFilter.h new file mode 100644 index 0000000..29bef9e --- /dev/null +++ b/video_input/consistencyChecker/CFilter.h @@ -0,0 +1,2210 @@ +// - Classes for 1D and 2D convolution stencils
+// - Pre-defined convolution stencils for binomial filters
+// - Pre-defined convolution stencils for 1st, 2nd, 3rd and 4th derivatives up to order 10
+// - Functions for convolution
+//
+// Author: Thomas Brox
+
+#ifndef CFILTER
+#define CFILTER
+
+#include <math.h>
+#include <NMath.h>
+#include <CVector.h>
+#include <CMatrix.h>
+#include <CTensor.h>
+#include <CTensor4D.h>
+
+// CFilter is an extention of CVector. It has an additional property Delta
+// which shifts the data to the left (a vector always begins with index 0).
+// This enables a filter's range to go from A to B where A can also
+// be less than zero.
+//
+// Example:
+// CFilter<double> filter(3,1);
+// filter = 1.0;
+// cout << filter(-1) << ", " << filter(0) << ", " << filter(1) << endl;
+//
+// CFilter2D behaves the same way as CFilter but is an extension of CMatrix
+
+template <class T>
+class CFilter : public CVector<T> {
+public:
+ // constructor
+ inline CFilter(const int aSize, const int aDelta = 0);
+ // copy constructor
+ CFilter(const CFilter<T>& aCopyFrom);
+ // constructor initialized by a vector
+ CFilter(const CVector<T>& aCopyFrom, const int aDelta = 0);
+
+ // Access to the filter's values
+ inline T& operator()(const int aIndex) const;
+ inline T& operator[](const int aIndex) const;
+ // Copies a filter into this filter
+ CFilter<T>& operator=(const CFilter<T>& aCopyFrom);
+
+ // Access to the filter's delta
+ inline int delta() const;
+ // Access to the filter's range A<=i<B
+ inline int A() const;
+ inline int B() const;
+ // Returns the sum of all filter co-efficients (absolutes)
+ T sum() const;
+ // Shifts the filter
+ inline void shift(int aDelta);
+protected:
+ int mDelta;
+};
+
+template <class T>
+class CFilter2D : public CMatrix<T> {
+public:
+ // constructor
+ inline CFilter2D();
+ inline CFilter2D(const int aXSize, const int aYSize, const int aXDelta = 0, const int aYDelta = 0);
+ // copy contructor
+ CFilter2D(const CFilter2D<T>& aCopyFrom);
+ // constructor initialized by a matrix
+ CFilter2D(const CMatrix<T>& aCopyFrom, const int aXDelta = 0, const int aYDelta = 0);
+ // Normalize sum of values to 1.0
+ void normalizeSum();
+ // Moves the filter's center
+ void shift(int aXDelta, int aYDelta);
+
+ // Access to filter's values
+ inline T& operator()(const int ax, const int ay) const;
+ // Copies a filter into this filter
+ CFilter2D<T>& operator=(const CFilter2D<T>& aCopyFrom);
+
+ // Access to the filter's delta
+ inline int deltaX() const;
+ inline int deltaY() const;
+ // Access to the filter's range A<=i<B
+ inline int AX() const;
+ inline int BX() const;
+ inline int AY() const;
+ inline int BY() const;
+ // Returns the sum of all filter co-efficients (absolutes)
+ T sum() const;
+protected:
+ int mDeltaX;
+ int mDeltaY;
+};
+
+namespace NFilter {
+
+ // Linear 1D filtering
+
+ // Convolution of the vector aVector with aFilter
+ // The result will be written into aVector, so its initial values will get lost
+ template <class T> inline void filter(CVector<T>& aVector, const CFilter<T>& aFilter);
+ // Convolution of the vector aVector with aFilter, the initial values of aVector will persist.
+ template <class T> void filter(const CVector<T>& aVector, CVector<T>& aResult, const CFilter<T>& aFilter);
+
+ // Convolution with a rectangle -> approximation of Gaussian
+ template <class T> inline void boxFilter(CVector<T>& aVector, int aWidth);
+ template <class T> void boxFilter(const CVector<T>& aVector, CVector<T>& aResult, int aWidth);
+
+ // Linear 2D filtering
+
+ // Convolution of the matrix aMatrix with aFilter, aFilter should be a separable filter
+ // The result will be written into aMatrix, so its initial values will get lost
+ template <class T> inline void filter(CMatrix<T>& aMatrix, const CFilter<T>& aFilterX, const CFilter<T>& aFilterY);
+ template <class T> inline void filterMin(CMatrix<T>& aMatrix, const CFilter<T>& aFilterX, const CFilter<T>& aFilterY);
+ // Convolution of the matrix aMatrix with aFilter, aFilter must be separable
+ // The initial values of aMatrix will persist.
+ template <class T> inline void filter(const CMatrix<T>& aMatrix, CMatrix<T>& aResult, const CFilter<T>& aFilterX, const CFilter<T>& aFilterY);
+ template <class T> inline void filterMin(const CMatrix<T>& aMatrix, CMatrix<T>& aResult, const CFilter<T>& aFilterX, const CFilter<T>& aFilterY);
+
+ // Convolution of the matrix aMatrix with aFilter only in x-direction, aDummy can be set to 1
+ // The result will be written into aMatrix, so its initial values will get lost
+ template <class T> inline void filter(CMatrix<T>& aMatrix, const CFilter<T>& aFilter, const int aDummy);
+ template <class T> inline void filterMin(CMatrix<T>& aMatrix, const CFilter<T>& aFilter, const int aDummy);
+ // Convolution of the matrix aMatrix with aFilter only in x-direction, aDummy can be set to 1
+ // The initial values of aMatrix will persist.
+ template <class T> void filter(const CMatrix<T>& aMatrix, CMatrix<T>& aResult, const CFilter<T>& aFilter, const int aDummy);
+ template <class T> void filterMin(const CMatrix<T>& aMatrix, const CMatrix<T>& aOrig, CMatrix<T>& aResult, const CFilter<T>& aFilter, const int aDummy);
+ // Convolution of the matrix aMatrix with aFilter only in y-direction, aDummy can be set to 1
+ // The result will be written into aMatrix, so its initial values will get lost
+ template <class T> inline void filter(CMatrix<T>& aMatrix, const int aDummy, const CFilter<T>& aFilter);
+ template <class T> inline void filterMin(CMatrix<T>& aMatrix, const int aDummy, const CFilter<T>& aFilter);
+ // Convolution of the matrix aMatrix with aFilter only in y-direction, aDummy can be set to 1
+ // The initial values of aMatrix will persist.
+ template <class T> void filter(const CMatrix<T>& aMatrix, CMatrix<T>& aResult, const int aDummy, const CFilter<T>& aFilter);
+ template <class T> void filterMin(const CMatrix<T>& aMatrix, const CMatrix<T>& aOrig, CMatrix<T>& aResult, const int aDummy, const CFilter<T>& aFilter);
+
+ // Convolution of the matrix aMatrix with aFilter
+ // The result will be written to aMatrix, so its initial values will get lost
+ template <class T> inline void filter(CMatrix<T>& aMatrix, const CFilter2D<T>& aFilter);
+ // Convolution of the matrix aMatrix with aFilter, the initial values of aMatrix will persist
+ template <class T> void filter(const CMatrix<T>& aMatrix, CMatrix<T>& aResult, const CFilter2D<T>& aFilter);
+
+ // Convolution with a rectangle -> approximation of Gaussian
+ template <class T> inline void boxFilterX(CMatrix<T>& aMatrix, int aWidth);
+ template <class T> void boxFilterX(const CMatrix<T>& aMatrix, CMatrix<T>& aResult, int aWidth);
+ template <class T> inline void boxFilterY(CMatrix<T>& aMatrix, int aWidth);
+ template <class T> void boxFilterY(const CMatrix<T>& aMatrix, CMatrix<T>& aResult, int aWidth);
+
+ // Recursive filter -> approximation of Gaussian
+ template <class T> void recursiveSmoothX(CMatrix<T>& aMatrix, float aSigma);
+ template <class T> void recursiveSmoothY(CMatrix<T>& aMatrix, float aSigma);
+ template <class T> inline void recursiveSmooth(CMatrix<T>& aMatrix, float aSigma);
+
+ // Linear 3D filtering
+
+ // Convolution of the 3D Tensor aTensor with aFilter, aFilter must be separable
+ // The result will be written back to aTensor so its initial values will get lost
+ template <class T> inline void filter(CTensor<T>& aTensor, const CFilter<T>& aFilterX, const CFilter<T>& aFilterY, const CFilter<T>& aFilterZ);
+ // Convolution of the 3D Tensor aTensor with aFilter, aFilter must be separable
+ // The initial values of aTensor will persist
+ template <class T> inline void filter(const CTensor<T>& aTensor, CTensor<T>& aResult, const CFilter<T>& aFilterX, const CFilter<T>& aFilterY, const CFilter<T>& aFilterZ);
+
+ // Convolution of the 3D Tensor aTensor with aFilter only in x-Direction
+ template <class T> inline void filter(CTensor<T>& aTensor, const CFilter<T>& aFilter, const int aDummy1, const int aDummy2);
+ template <class T> void filter(const CTensor<T>& aTensor, CTensor<T>& aResult, const CFilter<T>& aFilter, const int aDummy1, const int aDummy2);
+ // Convolution of the 3D Tensor aTensor with aFilter only in y-Direction
+ template <class T> inline void filter(CTensor<T>& aTensor, const int aDummy1, const CFilter<T>& aFilter, const int aDummy2);
+ template <class T> void filter(const CTensor<T>& aTensor, CTensor<T>& aResult, const int aDummy1, const CFilter<T>& aFilter, const int aDummy2);
+ // Convolution of the 3D Tensor aTensor with aFilter only in z-Direction
+ template <class T> inline void filter(CTensor<T>& aTensor, const int aDummy1, const int aDummy2, const CFilter<T>& aFilter);
+ template <class T> void filter(const CTensor<T>& aTensor, CTensor<T>& aResult, const int aDummy1, const int aDummy2, const CFilter<T>& aFilter);
+
+ // Convolution with a rectangle -> approximation of Gaussian
+ template <class T> inline void boxFilterX(CTensor<T>& aTensor, int aWidth);
+ template <class T> void boxFilterX(const CTensor<T>& aTensor, CTensor<T>& aResult, int aWidth);
+ template <class T> inline void boxFilterY(CTensor<T>& aTensor, int aWidth);
+ template <class T> void boxFilterY(const CTensor<T>& aTensor, CTensor<T>& aResult, int aWidth);
+ template <class T> inline void boxFilterZ(CTensor<T>& aTensor, int aWidth);
+ template <class T> void boxFilterZ(const CTensor<T>& aTensor, CTensor<T>& aResult, int aWidth);
+
+ // Recursive filter -> approximation of Gaussian
+ template <class T> void recursiveSmoothX(CTensor<T>& aTensor, float aSigma);
+ template <class T> void recursiveSmoothY(CTensor<T>& aTensor, float aSigma);
+ template <class T> void recursiveSmoothZ(CTensor<T>& aTensor, float aSigma);
+
+ // Linear 4D filtering
+
+ // Convolution of the 4D Tensor aTensor with aFilter, aFilter must be separable
+ // The result will be written back to aTensor so its initial values will get lost
+ template <class T> inline void filter(CTensor4D<T>& aTensor, const CFilter<T>& aFilterX, const CFilter<T>& aFilterY, const CFilter<T>& aFilterZ, const CFilter<T>& aFilterA);
+
+ // Convolution of the 4D Tensor aTensor with aFilter only in x-Direction
+ template <class T> inline void filter(CTensor4D<T>& aTensor, const CFilter<T>& aFilter, const int aDummy1, const int aDummy2, const int aDummy3);
+ template <class T> void filter(const CTensor4D<T>& aTensor, CTensor4D<T>& aResult, const CFilter<T>& aFilter, const int aDummy1, const int aDummy2, const int aDummy3);
+ // Convolution of the 4D Tensor aTensor with aFilter only in y-Direction
+ template <class T> inline void filter(CTensor4D<T>& aTensor, const int aDummy1, const CFilter<T>& aFilter, const int aDummy2, const int aDummy3);
+ template <class T> void filter(const CTensor4D<T>& aTensor, CTensor4D<T>& aResult, const int aDummy1, const CFilter<T>& aFilter, const int aDummy2, const int aDummy3);
+ // Convolution of the 4D Tensor aTensor with aFilter only in z-Direction
+ template <class T> inline void filter(CTensor4D<T>& aTensor, const int aDummy1, const int aDummy2, const CFilter<T>& aFilter, const int aDummy3);
+ template <class T> void filter(const CTensor4D<T>& aTensor, CTensor4D<T>& aResult, const int aDummy1, const int aDummy2, const CFilter<T>& aFilter, const int aDummy3);
+ // Convolution of the 4D Tensor aTensor with aFilter only in a-Direction
+ template <class T> inline void filter(CTensor4D<T>& aTensor, const int aDummy1, const int aDummy2, const int aDummy3, const CFilter<T>& aFilter);
+ template <class T> void filter(const CTensor4D<T>& aTensor, CTensor4D<T>& aResult, const int aDummy1, const int aDummy2, const int aDummy3, const CFilter<T>& aFilter);
+
+ // Recursive filter -> approximation of Gaussian
+ template <class T> void recursiveSmoothX(CTensor4D<T>& aTensor, float aSigma);
+ template <class T> void recursiveSmoothY(CTensor4D<T>& aTensor, float aSigma);
+ template <class T> void recursiveSmoothZ(CTensor4D<T>& aTensor, float aSigma);
+ template <class T> void recursiveSmoothA(CTensor4D<T>& aTensor, float aSigma);
+
+ // Nonlinear filtering: Osher shock filter
+ template <class T> void osher(CMatrix<T>& aData, int aIterations = 20);
+ template <class T> inline void osher(const CMatrix<T>& aData, CMatrix<T>& aResult, int aIterations = 20);
+}
+
+// Common filters
+
+template <class T>
+class CGauss : public CFilter<T> {
+public:
+ CGauss(const int aSize, const int aDegreeOfDerivative);
+};
+
+template <class T>
+class CSmooth : public CFilter<T> {
+public:
+ CSmooth(float aSigma, float aPrecision);
+};
+
+template <class T>
+class CGaussianFirstDerivative : public CFilter<T> {
+public:
+ CGaussianFirstDerivative(float aSigma, float aPrecision);
+};
+
+template <class T>
+class CGaussianSecondDerivative : public CFilter<T> {
+public:
+ CGaussianSecondDerivative(float aSigma, float aPrecision);
+};
+
+template <class T>
+class CDerivative : public CFilter<T> {
+public:
+ CDerivative(const int aSize);
+};
+
+template <class T>
+class CHighOrderDerivative : public CFilter<T> {
+public:
+ CHighOrderDerivative(int aOrder, int aSize);
+};
+
+template <class T>
+class CGaborReal : public CFilter2D<T> {
+public:
+ CGaborReal(float aFrequency, float aAngle, float aSigma1 = 3.0, float aSigma2 = 3.0);
+};
+
+template <class T>
+class CGaborImaginary : public CFilter2D<T> {
+public:
+ CGaborImaginary(float aFrequency, float aAngle, float aSigma1 = 3.0, float aSigma2 = 3.0);
+};
+
+// Exceptions -----------------------------------------------------------------
+
+// Thrown if one tries to access an element of a filter which is out of the filter's bounds
+struct EFilterRangeOverflow {
+ EFilterRangeOverflow(const int aIndex, const int aA, const int aB) {
+ using namespace std;
+ cerr << "Exception EFilterRangeOverflow: i = " << aIndex;
+ cerr << " Allowed Range: " << aA << " <= i < " << aB << endl;
+ }
+ EFilterRangeOverflow(const int ax, const int ay, const int aAX, const int aBX, const int aAY, const int aBY) {
+ using namespace std;
+ cerr << "Exception EFilterRangeOverflow: (x,y) = (" << ax << "," << ay << ") ";
+ cerr << "Allowed Range: " << aAX << " <= x < " << aBX << " " << aAY << " <= y < " << aBY << endl;
+ }
+};
+
+// Thrown if the resulting container has not the same size as the initial container
+struct EFilterIncompatibleSize {
+ EFilterIncompatibleSize(const int aSize1, const int aSize2) {
+ using namespace std;
+ cerr << "Exception EFilterIncompatibleSize: Initial container size: " << aSize1;
+ cerr << " Resulting container size: " << aSize2 << endl;
+ }
+};
+
+// Thrown if the demanded filter is not available
+struct EFilterNotAvailable {
+ EFilterNotAvailable(int aSize, int aOrder) {
+ using namespace std;
+ cerr << "Exception EFilterNotAvailable: Mask size: " << aSize;
+ if (aOrder >= 0) cerr << " Derivative order: " << aOrder;
+ cerr << endl;
+ }
+};
+
+// I M P L E M E N T A T I O N ------------------------------------------------
+//
+// You might wonder why there is implementation code in a header file.
+// The reason is that not all C++ compilers yet manage separate compilation
+// of templates. Inline functions cannot be compiled separately anyway.
+// So in this case the whole implementation code is added to the header
+// file.
+// Users should ignore everything that's beyond this line :)
+// ----------------------------------------------------------------------------
+
+// C F I L T E R --------------------------------------------------------------
+// P U B L I C ----------------------------------------------------------------
+// constructor
+template <class T>
+inline CFilter<T>::CFilter(const int aSize, const int aDelta)
+ : CVector<T>(aSize),mDelta(aDelta) {
+}
+
+// copy constructor
+template <class T>
+CFilter<T>::CFilter(const CFilter<T>& aCopyFrom)
+ : CVector<T>(aCopyFrom.mSize),mDelta(aCopyFrom.mDelta) {
+ for (register int i = 0; i < this->mSize; i++)
+ this->mData[i] = aCopyFrom.mData[i];
+}
+
+// constructor initialized by a vector
+template <class T>
+CFilter<T>::CFilter(const CVector<T>& aCopyFrom, const int aDelta)
+ : CVector<T>(aCopyFrom.size()),mDelta(aDelta) {
+ for (register int i = 0; i < this->mSize; i++)
+ this->mData[i] = aCopyFrom(i);
+}
+
+// operator()
+template <class T>
+inline T& CFilter<T>::operator()(const int aIndex) const {
+ #ifdef DEBUG
+ if (aIndex < A() || aIndex >= B())
+ throw EFilterRangeOverflow(aIndex,A(),B());
+ #endif
+ return this->mData[aIndex+mDelta];
+}
+
+// operator[]
+template <class T>
+inline T& CFilter<T>::operator[](const int aIndex) const {
+ return operator()(aIndex);
+}
+
+// operator=
+template <class T>
+CFilter<T>& CFilter<T>::operator=(const CFilter<T>& aCopyFrom) {
+ if (this != &aCopyFrom) {
+ delete[] this->mData;
+ this->mSize = aCopyFrom.mSize;
+ mDelta = aCopyFrom.mDelta;
+ this->mData = new T[this->mSize];
+ for (register int i = 0; i < this->mSize; i++)
+ this->mData[i] = aCopyFrom.mData[i];
+ }
+ return *this;
+}
+
+// delta
+template <class T>
+inline int CFilter<T>::delta() const {
+ return mDelta;
+}
+
+// A
+template <class T>
+inline int CFilter<T>::A() const {
+ return -mDelta;
+}
+
+// B
+template <class T>
+inline int CFilter<T>::B() const {
+ return this->mSize-mDelta;
+}
+
+// sum
+template <class T>
+T CFilter<T>::sum() const {
+ T aResult = 0;
+ for (int i = 0; i < this->mSize; i++)
+ aResult += fabs(this->mData[i]);
+ return aResult;
+}
+
+// shift
+template <class T>
+inline void CFilter<T>::shift(int aDelta) {
+ mDelta += aDelta;
+}
+
+// C F I L T E R 2 D -----------------------------------------------------------
+// P U B L I C ----------------------------------------------------------------
+// constructor
+template <class T>
+inline CFilter2D<T>::CFilter2D()
+ : CMatrix<T>(),mDeltaX(0),mDeltaY(0) {
+}
+
+template <class T>
+inline CFilter2D<T>::CFilter2D(const int aXSize, const int aYSize, const int aDeltaX, const int aDeltaY)
+ : CMatrix<T>(aXSize,aYSize),mDeltaX(aDeltaX),mDeltaY(aDeltaY) {
+}
+
+// copy constructor
+template <class T>
+CFilter2D<T>::CFilter2D(const CFilter2D<T>& aCopyFrom)
+ : CMatrix<T>(aCopyFrom.mXSize,aCopyFrom.mYSize),mDeltaX(aCopyFrom.mDeltaX,aCopyFrom.mDeltaY) {
+ for (int i = 0; i < this->mXSize*this->mYSize; i++)
+ this->mData[i] = aCopyFrom.mData[i];
+}
+
+// constructor initialized by a matrix
+template <class T>
+CFilter2D<T>::CFilter2D(const CMatrix<T>& aCopyFrom, const int aDeltaX, const int aDeltaY)
+ : CMatrix<T>(aCopyFrom.xSize(),aCopyFrom.ySize()),mDeltaX(aDeltaX),mDeltaY(aDeltaY) {
+ for (register int i = 0; i < this->mXSize*this->mYSize; i++)
+ this->mData[i] = aCopyFrom.data()[i];
+}
+
+// normalizeSum
+template <class T>
+void CFilter2D<T>::normalizeSum() {
+ int aSize = this->size();
+ T aSum = 0;
+ for (int i = 0; i < aSize; i++)
+ aSum += this->mData[i];
+ T invSum = 1.0/aSum;
+ for (int i = 0; i < aSize; i++)
+ this->mData[i] *= invSum;
+}
+
+// shift
+template <class T>
+void CFilter2D<T>::shift(int aXDelta, int aYDelta) {
+ mDeltaX = aXDelta;
+ mDeltaY = aYDelta;
+}
+
+// operator()
+template <class T>
+inline T& CFilter2D<T>::operator()(const int ax, const int ay) const {
+ #ifdef DEBUG
+ if (ax < AX() || ax >= BX() || ay < AY() || ay >= BY)
+ throw EFilterRangeOverflow(ax,ay,AX(),BX(),AY(),BY());
+ #endif
+ return this->mData[(ay+mDeltaY)*this->mXSize+ax+mDeltaX];
+}
+
+// operator=
+template <class T>
+CFilter2D<T>& CFilter2D<T>::operator=(const CFilter2D<T>& aCopyFrom) {
+ if (this != &aCopyFrom) {
+ delete[] this->mData;
+ this->mXSize = aCopyFrom.mXSize;
+ this->mYSize = aCopyFrom.mYSize;
+ mDeltaX = aCopyFrom.mDeltaX;
+ mDeltaY = aCopyFrom.mDeltaY;
+ this->mData = new T[this->mXSize*this->mYSize];
+ for (register int i = 0; i < this->mXSize*this->mYSize; i++)
+ this->mData[i] = aCopyFrom.mData[i];
+ }
+ return *this;
+}
+
+// deltaX
+template <class T>
+inline int CFilter2D<T>::deltaX() const {
+ return mDeltaX;
+}
+
+// deltaY
+template <class T>
+inline int CFilter2D<T>::deltaY() const {
+ return mDeltaY;
+}
+
+// AX
+template <class T>
+inline int CFilter2D<T>::AX() const {
+ return -mDeltaX;
+}
+
+// AY
+template <class T>
+inline int CFilter2D<T>::AY() const {
+ return -mDeltaY;
+}
+
+// BX
+template <class T>
+inline int CFilter2D<T>::BX() const {
+ return this->mXSize-mDeltaX;
+}
+
+// BY
+template <class T>
+inline int CFilter2D<T>::BY() const {
+ return this->mYSize-mDeltaY;
+}
+
+// sum
+template <class T>
+T CFilter2D<T>::sum() const {
+ T aResult = 0;
+ for (int i = 0; i < this->mXSize*this->mYSize; i++)
+ aResult += abs(this->mData[i]);
+ return aResult;
+}
+
+// C G A U S S -----------------------------------------------------------------
+template <class T>
+CGauss<T>::CGauss(const int aSize, const int aDegreeOfDerivative)
+ : CFilter<T>(aSize,aSize >> 1) {
+ CVector<int> *oldData;
+ CVector<int> *newData;
+ CVector<int> *temp;
+ oldData = new CVector<int>(aSize);
+ newData = new CVector<int>(aSize);
+
+ (*oldData)(0) = 1;
+ (*oldData)(1) = 1;
+
+ for (int i = 2; i < aSize-aDegreeOfDerivative; i++) {
+ (*newData)(0) = 1;
+ for (int j = 1; j < i; j++)
+ (*newData)(j) = (*oldData)(j)+(*oldData)(j-1);
+ (*newData)(i) = 1;
+ temp = oldData;
+ oldData = newData;
+ newData = temp;
+ }
+ for (int i = aSize-aDegreeOfDerivative; i < aSize; i++) {
+ (*newData)(0) = 1;
+ for (int j = 1; j < i; j++)
+ (*newData)(j) = (*oldData)(j)-(*oldData)(j-1);
+ (*newData)(i) = -(*oldData)(i-1);
+ temp = oldData;
+ oldData = newData;
+ newData = temp;
+ }
+
+ int aSum = 0;
+ for (int i = 0; i < aSize; i++)
+ aSum += abs((*oldData)(i));
+ double aInvSum = 1.0/aSum;
+ for (int i = 0; i < aSize; i++)
+ this->mData[aSize-1-i] = (*oldData)(i)*aInvSum;
+
+ delete newData;
+ delete oldData;
+}
+
+// C S M O O T H ---------------------------------------------------------------
+template <class T>
+CSmooth<T>::CSmooth(float aSigma, float aPrecision)
+ : CFilter<T>(2*(int)ceil(aPrecision*aSigma)+1,(int)ceil(aPrecision*aSigma)) {
+ float aSqrSigma = aSigma*aSigma;
+ for (int i = 0; i <= (this->mSize >> 1); i++) {
+ T aTemp = exp(i*i/(-2.0*aSqrSigma))/(aSigma*sqrt(2.0*NMath::Pi));
+ this->operator()(i) = aTemp;
+ this->operator()(-i) = aTemp;
+ }
+ T invSum = 1.0/this->sum();
+ for (int i = 0; i < this->mSize; i++)
+ this->mData[i] *= invSum;
+}
+
+template <class T>
+CGaussianFirstDerivative<T>::CGaussianFirstDerivative(float aSigma, float aPrecision)
+ : CFilter<T>(2*(int)ceil(aPrecision*aSigma)+1,(int)ceil(aPrecision*aSigma)) {
+ float aSqrSigma = aSigma*aSigma;
+ float aPreFactor = 1.0/(aSqrSigma*aSigma*sqrt(2.0*NMath::Pi));
+ for (int i = 0; i <= (this->mSize >> 1); i++) {
+ T aTemp = exp(i*i/(-2.0*aSqrSigma))*i*aPreFactor;
+ this->operator()(i) = aTemp;
+ this->operator()(-i) = -aTemp;
+ }
+}
+
+template <class T>
+CGaussianSecondDerivative<T>::CGaussianSecondDerivative(float aSigma, float aPrecision)
+ : CFilter<T>(2*(int)ceil(aPrecision*aSigma)+1,(int)ceil(aPrecision*aSigma)) {
+ float aSqrSigma = aSigma*aSigma;
+ float aPreFactor = 1.0/(aSqrSigma*aSigma*sqrt(2.0*NMath::Pi));
+ for (int i = 0; i <= (this->mSize >> 1); i++) {
+ T aTemp = exp(i*i/(-2.0*aSqrSigma))*(i*i/aSqrSigma-1.0)*aPreFactor;
+ this->operator()(i) = aTemp;
+ this->operator()(-i) = aTemp;
+ }
+}
+
+// C D E R I V A T I V E -------------------------------------------------------
+template <class T>
+CDerivative<T>::CDerivative(const int aSize)
+ : CFilter<T>(aSize,(aSize-1) >> 1) {
+ switch (aSize) {
+ case 2:
+ this->mData[0] = -1;
+ this->mData[1] = 1;
+ break;
+ case 3:
+ this->mData[0] = -0.5;
+ this->mData[1] = 0;
+ this->mData[2] = 0.5;
+ break;
+ case 4:
+ this->mData[0] = 0.041666666666666666666666666666667;
+ this->mData[1] = -1.125;
+ this->mData[2] = 1.125;
+ this->mData[3] = -0.041666666666666666666666666666667;
+ break;
+ case 5:
+ this->mData[0] = 0.083333333333;
+ this->mData[1] = -0.66666666666;
+ this->mData[2] = 0;
+ this->mData[3] = 0.66666666666;
+ this->mData[4] = -0.083333333333;
+ break;
+ case 6:
+ this->mData[0] = -0.0046875;
+ this->mData[1] = 0.0651041666666666666666666666666667;
+ this->mData[2] = -1.171875;
+ this->mData[3] = 1.171875;
+ this->mData[4] = -0.0651041666666666666666666666666667;
+ this->mData[5] = 0.0046875;
+ break;
+ case 7:
+ this->mData[0] = -0.016666666666666666666666666666667;
+ this->mData[1] = 0.15;
+ this->mData[2] = -0.75;
+ this->mData[3] = 0;
+ this->mData[4] = 0.75;
+ this->mData[5] = -0.15;
+ this->mData[6] = 0.016666666666666666666666666666667;
+ break;
+ case 8:
+ this->mData[0] = 6.9754464285714285714285714285714e-4;
+ this->mData[1] = -0.0095703125;
+ this->mData[2] = 0.079752604166666666666666666666667;
+ this->mData[3] = -1.1962890625;
+ this->mData[4] = 1.1962890625;
+ this->mData[5] = -0.079752604166666666666666666666667;
+ this->mData[6] = 0.0095703125;
+ this->mData[7] = -6.9754464285714285714285714285714e-4;
+ break;
+ case 9:
+ this->mData[0] = 0.0035714285714285714285714285714286;
+ this->mData[1] = -0.038095238095238095238095238095238;
+ this->mData[2] = 0.2;
+ this->mData[3] = -0.8;
+ this->mData[4] = 0;
+ this->mData[5] = 0.8;
+ this->mData[6] = -0.2;
+ this->mData[7] = 0.038095238095238095238095238095238;
+
+ this->mData[8] = -0.0035714285714285714285714285714286;
+ break;
+ case 10:
+ this->mData[0] = -1.1867947048611111111111111111111e-4;
+ this->mData[1] = 0.0017656598772321428571428571428571;
+ this->mData[2] = -0.0138427734375;
+ this->mData[3] = 0.0897216796875;
+ this->mData[4] = -1.21124267578125;
+ this->mData[5] = 1.21124267578125;
+ this->mData[6] = -0.0897216796875;
+ this->mData[7] = 0.0138427734375;
+ this->mData[8] = -0.0017656598772321428571428571428571;
+ this->mData[9] = 1.1867947048611111111111111111111e-4;
+ break;
+ default:
+ throw EFilterNotAvailable(aSize,-1);
+ }
+}
+
+// C H I G H O R D E R D E R I V A T I V E -------------------------------------
+template <class T>
+CHighOrderDerivative<T>::CHighOrderDerivative(int aOrder, int aSize)
+ : CFilter<T>(aSize,(aSize-1) >> 1) {
+ switch (aSize) {
+ case 3:
+ switch (aOrder) {
+ case 2:
+ this->mData[0] = 1;
+ this->mData[1] = -2;
+ this->mData[2] = 1;
+ break;
+ default:
+ throw EFilterNotAvailable(aSize,aOrder);
+ }
+ break;
+ case 4:
+ switch (aOrder) {
+ case 2:
+ this->mData[0] = 0.25;
+ this->mData[1] = -0.25;
+ this->mData[2] = -0.25;
+ this->mData[3] = 0.25;
+ break;
+ case 3:
+ this->mData[0] = -0.25;
+ this->mData[1] = 0.75;
+ this->mData[2] = -0.75;
+ this->mData[3] = 0.25;
+ break;
+ default:
+ throw EFilterNotAvailable(aSize,aOrder);
+ }
+ break;
+ case 5:
+ switch (aOrder) {
+ case 2:
+ this->mData[0] = -0.083333333333333333333333333333333;
+ this->mData[1] = 1.3333333333333333333333333333333;
+ this->mData[2] = -2.5;
+ this->mData[3] = 1.3333333333333333333333333333333;
+ this->mData[4] = -0.083333333333333333333333333333333;
+ break;
+ case 3:
+ this->mData[0] = -0.5;
+ this->mData[1] = 1;
+ this->mData[2] = 0;
+ this->mData[3] = -1;
+ this->mData[4] = 0.5;
+ break;
+ case 4:
+ this->mData[0] = 1;
+ this->mData[1] = -4;
+ this->mData[2] = 6;
+ this->mData[3] = -4;
+ this->mData[4] = 1;
+ break;
+ default:
+ throw EFilterNotAvailable(aSize,aOrder);
+ }
+ break;
+ case 6:
+ switch (aOrder) {
+ case 2:
+ this->mData[0] = -0.052083333333333333333333333333333;
+ this->mData[1] = 0.40625;
+ this->mData[2] = -0.35416666666666666666666666666667;
+ this->mData[3] = -0.35416666666666666666666666666667;
+ this->mData[4] = 0.40625;
+ this->mData[5] = -0.052083333333333333333333333333333;
+ break;
+ case 3:
+ this->mData[0] = 0.03125;
+ this->mData[1] = -0.40625;
+ this->mData[2] = 1.0625;
+ this->mData[3] = -1.0625;
+ this->mData[4] = 0.40625;
+ this->mData[5] = -0.03125;
+ break;
+ case 4:
+ this->mData[0] = 0.0625;
+ this->mData[1] = -0.1875;
+ this->mData[2] = 0.125;
+ this->mData[3] = 0.125;
+ this->mData[4] = -0.1875;
+ this->mData[5] = 0.0625;
+ break;
+ default:
+ throw EFilterNotAvailable(aSize,aOrder);
+ }
+ break;
+ case 7:
+ switch (aOrder) {
+ case 2:
+ this->mData[0] = 0.011111111111111111111111111111111;
+ this->mData[1] = -0.15;
+ this->mData[2] = 1.5;
+ this->mData[3] = -2.6666666666666666666666666666667;
+ this->mData[4] = 1.5;
+ this->mData[5] = -0.15;
+ this->mData[6] = 0.011111111111111111111111111111111;
+ break;
+ case 3:
+ this->mData[0] = 0.125;
+ this->mData[1] = -1;
+ this->mData[2] = 1.625;
+ this->mData[3] = 0;
+ this->mData[4] = -1.625;
+ this->mData[5] = 1;
+ this->mData[6] = -0.125;
+ break;
+ case 4:
+ this->mData[0] = -0.16666666666666666666666666666667;
+ this->mData[1] = 2;
+ this->mData[2] = -6.5;
+ this->mData[3] = 9.3333333333333333333333333333333;
+ this->mData[4] = -6.5;
+ this->mData[5] = 2;
+ this->mData[6] = -0.16666666666666666666666666666667;
+ break;
+ default:
+ throw EFilterNotAvailable(aSize,aOrder);
+ }
+ break;
+ case 8:
+ switch (aOrder) {
+ case 2:
+ this->mData[0] = 0.011241319444444444444444444444444;
+ this->mData[1] = -0.10828993055555555555555555555556;
+ this->mData[2] = 0.507421875;
+ this->mData[3] = -0.41037326388888888888888888888889;
+ this->mData[4] = -0.41037326388888888888888888888889;
+ this->mData[5] = 0.507421875;
+ this->mData[6] = -0.10828993055555555555555555555556;
+ this->mData[7] = 0.011241319444444444444444444444444;
+ break;
+ case 3:
+ this->mData[0] = -0.0048177083333333333333333333333333;
+ this->mData[1] = 0.064973958333333333333333333333333;
+ this->mData[2] = -0.507421875;
+ this->mData[3] = 1.2311197916666666666666666666667;
+ this->mData[4] = -1.2311197916666666666666666666667;
+ this->mData[5] = 0.507421875;
+ this->mData[6] = -0.064973958333333333333333333333333;
+ this->mData[7] = 0.0048177083333333333333333333333333;
+ break;
+ case 4:
+ this->mData[0] = -0.018229166666666666666666666666667;
+ this->mData[1] = 0.15364583333333333333333333333333;
+ this->mData[2] = -0.3515625;
+ this->mData[3] = 0.21614583333333333333333333333333;
+ this->mData[4] = 0.21614583333333333333333333333333;
+ this->mData[5] = -0.3515625;
+ this->mData[6] = 0.15364583333333333333333333333333;
+ this->mData[7] = -0.018229166666666666666666666666667;
+ break;
+ default:
+ throw EFilterNotAvailable(aSize,aOrder);
+ }
+ break;
+ case 9:
+ switch (aOrder) {
+ case 2:
+ this->mData[0] = -0.0017857142857142857142857142857143;
+ this->mData[1] = 0.025396825396825396825396825396825;
+ this->mData[2] = -0.2;
+ this->mData[3] = 1.6;
+ this->mData[4] = -2.8472222222222222222222222222222;
+ this->mData[5] = 1.6;
+ this->mData[6] = -0.2;
+ this->mData[7] = 0.025396825396825396825396825396825;
+ this->mData[8] = -0.0017857142857142857142857142857143;
+ break;
+ case 3:
+ this->mData[0] = -0.029166666666666666666666666666667;
+ this->mData[1] = 0.3;
+ this->mData[2] = -1.4083333333333333333333333333333;
+ this->mData[3] = 2.0333333333333333333333333333333;
+ this->mData[4] = 0;
+ this->mData[5] = -2.0333333333333333333333333333333;
+ this->mData[6] = 1.4083333333333333333333333333333;
+ this->mData[7] = -0.3;
+ this->mData[8] = 0.029166666666666666666666666666667;
+ break;
+ case 4:
+ this->mData[0] = 0.029166666666666666666666666666667;
+ this->mData[1] = -0.4;
+ this->mData[2] = 2.8166666666666666666666666666667;
+ this->mData[3] = -8.1333333333333333333333333333333;
+ this->mData[4] = 11.375;
+ this->mData[5] = -8.1333333333333333333333333333333;
+ this->mData[6] = 2.8166666666666666666666666666667;
+ this->mData[7] = -0.4;
+ this->mData[8] = 0.029166666666666666666666666666667;
+ break;
+ default:
+ throw EFilterNotAvailable(aSize,aOrder);
+ }
+ break;
+ case 10:
+ switch (aOrder) {
+ case 2:
+ this->mData[0] = -0.0025026351686507936507936507936508;
+ this->mData[1] = 0.028759765625;
+ this->mData[2] = -0.15834263392857142857142857142857;
+ this->mData[3] = 0.57749565972222222222222222222222;
+ this->mData[4] = -0.44541015625;
+ this->mData[5] = -0.44541015625;
+ this->mData[6] = 0.57749565972222222222222222222222;
+ this->mData[7] = -0.15834263392857142857142857142857;
+ this->mData[8] = 0.028759765625;
+ this->mData[9] = -0.0025026351686507936507936507936508;
+ break;
+ case 3:
+ this->mData[0] = 0.0008342117228835978835978835978836;
+ this->mData[1] = -0.012325613839285714285714285714286;
+ this->mData[2] = 0.095005580357142857142857142857143;
+ this->mData[3] = -0.57749565972222222222222222222222;
+ this->mData[4] = 1.33623046875;
+ this->mData[5] = -1.33623046875;
+ this->mData[6] = 0.57749565972222222222222222222222;
+ this->mData[7] = -0.095005580357142857142857142857143;
+ this->mData[8] = 0.012325613839285714285714285714286;
+ this->mData[9] = -0.0008342117228835978835978835978836;
+ break;
+ case 4:
+ this->mData[0] = 0.00458984375;
+ this->mData[1] = -0.050358072916666666666666666666667;
+ this->mData[2] = 0.24544270833333333333333333333333;
+ this->mData[3] = -0.480078125;
+ this->mData[4] = 0.28040364583333333333333333333333;
+ this->mData[5] = 0.28040364583333333333333333333333;
+ this->mData[6] = -0.480078125;
+ this->mData[7] = 0.24544270833333333333333333333333;
+ this->mData[8] = -0.050358072916666666666666666666667;
+ this->mData[9] = 0.00458984375;
+ break;
+ default:
+ throw EFilterNotAvailable(aSize,aOrder);
+ }
+ break;
+ default:
+ throw EFilterNotAvailable(aSize,aOrder);
+ }
+}
+
+// C G A B O R -----------------------------------------------------------------
+template <class T>
+CGaborReal<T>::CGaborReal(float aFrequency, float aAngle, float aSigma1, float aSigma2)
+ : CFilter2D<T>() {
+ // sqrt(2.0*log(2.0))/(2.0*NMath::Pi) = 0.18739
+ float sigma1Sqr2 = aSigma1*0.18739/aFrequency;
+ sigma1Sqr2 = 0.5/(sigma1Sqr2*sigma1Sqr2);
+ float sigma2Sqr2 = aSigma2*0.18739/aFrequency;
+ sigma2Sqr2 = 0.5/(sigma2Sqr2*sigma2Sqr2);
+ float aCos = cos(aAngle);
+ float aSin = sin(aAngle);
+ float a = 0.6*aSigma1/aFrequency;
+ float b = 0.6*aSigma2/aFrequency;
+ float aXSize = fabs(a*aCos)+fabs(b*aSin);
+ float aYSize = fabs(b*aCos)+fabs(a*aSin);
+ this->setSize(1+2.0*floor(aXSize),1+2.0*floor(aYSize));
+ this->shift(floor(aXSize),floor(aYSize));
+ for (int y = this->AY(); y < this->BY(); y++)
+ for (int x = this->AX(); x < this->BX(); x++) {
+ float a = x*aCos+y*aSin;
+ float b = y*aCos-x*aSin;
+ float aGauss = exp(-sigma1Sqr2*a*a-sigma2Sqr2*b*b);
+ float aHelp = 2.0*NMath::Pi*aFrequency*(x*aCos+y*aSin);
+ this->operator()(x,y) = aGauss*cos(aHelp);
+ }
+}
+
+template <class T>
+CGaborImaginary<T>::CGaborImaginary(float aFrequency, float aAngle, float aSigma1, float aSigma2)
+ : CFilter2D<T>() {
+ // sqrt(2.0*log(2.0))/(2.0*NMath::Pi) = 0.18739
+ float sigma1Sqr2 = aSigma1*0.18739/aFrequency;
+ sigma1Sqr2 = 0.5/(sigma1Sqr2*sigma1Sqr2);
+ float sigma2Sqr2 = aSigma2*0.18739/aFrequency;
+ sigma2Sqr2 = 0.5/(sigma2Sqr2*sigma2Sqr2);
+ float aCos = cos(aAngle);
+ float aSin = sin(aAngle);
+ float a = 0.6*aSigma1/aFrequency;
+ float b = 0.6*aSigma2/aFrequency;
+ float aXSize = fabs(a*aCos)+fabs(b*aSin);
+ float aYSize = fabs(b*aCos)+fabs(a*aSin);
+ this->setSize(1+2.0*floor(aXSize),1+2.0*floor(aYSize));
+ this->shift(floor(aXSize),floor(aYSize));
+ for (int y = this->AY(); y < this->BY(); y++)
+ for (int x = this->AX(); x < this->BX(); x++) {
+ float a = x*aCos+y*aSin;
+ float b = y*aCos-x*aSin;
+ float aGauss = exp(-sigma1Sqr2*a*a-sigma2Sqr2*b*b);
+ float aHelp = 2.0*NMath::Pi*aFrequency*(x*aCos+y*aSin);
+ this->operator()(x,y) = aGauss*sin(aHelp);
+ }
+}
+
+// F I L T E R -----------------------------------------------------------------
+
+namespace NFilter {
+
+// 1D linear filtering ---------------------------------------------------------
+
+template <class T>
+inline void filter(CVector<T>& aVector, const CFilter<T>& aFilter) {
+ CVector<T> oldVector(aVector);
+ filter(oldVector,aVector,aFilter);
+}
+
+template <class T>
+void filter(const CVector<T>& aVector, CVector<T>& aResult, const CFilter<T>& aFilter) {
+ if (aResult.size() != aVector.size()) throw EFilterIncompatibleSize(aVector.size(),aResult.size());
+ int x1 = -aFilter.A();
+ int x2 = aVector.size()-aFilter.B();
+ int a2Size = 2*aVector.size()-1;
+ // Left rim
+ for (int i = 0; i < x1; i++) {
+ aResult[i] = 0;
+ for (int j = aFilter.A(); j < aFilter.B(); j++)
+ if (j+i < 0) aResult(i) += aFilter(j)*aVector(-1-j-i);
+ else aResult(i) += aFilter(j)*aVector(j+i);
+ }
+ // Middle
+ for (int i = x1; i < x2; i++) {
+ aResult[i] = 0;
+ for (int j = aFilter.A(); j < aFilter.B(); j++)
+ aResult(i) += aFilter(j)*aVector(j+i);
+ }
+ // Right rim
+ for (int i = x2; i < aResult.size(); i++) {
+ aResult[i] = 0;
+ for (int j = aFilter.A(); j < aFilter.B(); j++)
+ if (j+i >= aVector.size()) aResult(i) += aFilter(j)*aVector(a2Size-j-i);
+ else aResult(i) += aFilter(j)*aVector(j+i);
+ }
+}
+
+// boxfilter
+template <class T>
+inline void boxFilter(CVector<T>& aVector, int aWidth) {
+ CVector<T> aTemp(aVector);
+ boxFilter(aTemp,aVector,aWidth);
+}
+
+template <class T>
+void boxFilter(const CVector<T>& aVector, CVector<T>& aResult, int aWidth) {
+ if (aWidth % 2 == 0) aWidth += 1;
+ T* invWidth = new T[aWidth+1];
+ invWidth[0] = 1.0f;
+ for (int i = 1; i <= aWidth; i++)
+ invWidth[i] = 1.0/i;
+ int halfWidth = (aWidth >> 1);
+ int aRight = halfWidth;
+ if (aRight >= aVector.size()) aRight = aVector.size()-1;
+ // Initialize
+ T aSum = 0.0f;
+ for (int x = 0; x <= aRight; x++)
+ aSum += aVector(x);
+ int aNum = aRight+1;
+ // Shift
+ for (int x = 0; x < aVector.size(); x++) {
+ aResult(x) = aSum*invWidth[aNum];
+ if (x-halfWidth >= 0) {
+ aSum -= aVector(x-halfWidth); aNum--;
+ }
+ if (x+halfWidth+1 < aVector.size()) {
+ aSum += aVector(x+halfWidth+1); aNum++;
+ }
+ }
+ delete[] invWidth;
+}
+
+// 2D linear filtering ---------------------------------------------------------
+
+template <class T>
+inline void filter(CMatrix<T>& aMatrix, const CFilter<T>& aFilterX, const CFilter<T>& aFilterY) {
+ CMatrix<T> tempMatrix(aMatrix.xSize(),aMatrix.ySize());
+ filter(aMatrix,tempMatrix,aFilterX,1);
+ filter(tempMatrix,aMatrix,1,aFilterY);
+}
+
+template <class T>
+inline void filter(const CMatrix<T>& aMatrix, CMatrix<T>& aResult, const CFilter<T>& aFilterX, const CFilter<T>& aFilterY) {
+ CMatrix<T> tempMatrix(aMatrix.xSize(),aMatrix.ySize());
+ filter(aMatrix,tempMatrix,aFilterX,1);
+ filter(tempMatrix,aResult,1,aFilterY);
+}
+
+template <class T>
+inline void filter(CMatrix<T>& aMatrix, const CFilter<T>& aFilter, const int aDummy) {
+ CMatrix<T> tempMatrix(aMatrix.xSize(),aMatrix.ySize());
+ filter(aMatrix,tempMatrix,aFilter,1);
+ aMatrix = tempMatrix;
+}
+
+template <class T>
+void filter(const CMatrix<T>& aMatrix, CMatrix<T>& aResult, const CFilter<T>& aFilter, const int aDummy) {
+ if (aResult.xSize() != aMatrix.xSize() || aResult.ySize() != aMatrix.ySize())
+ throw EFilterIncompatibleSize(aMatrix.xSize()*aMatrix.ySize(),aResult.xSize()*aResult.ySize());
+ int x1 = -aFilter.A();
+ int x2 = aMatrix.xSize()-aFilter.B();
+ int a2Size = 2*aMatrix.xSize()-1;
+ aResult = 0;
+ for (int y = 0; y < aMatrix.ySize(); y++) {
+ int aOffset = y*aMatrix.xSize();
+ // Left rim
+ for (int x = 0; x < x1; x++)
+ for (int i = aFilter.A(); i < aFilter.B(); i++) {
+ if (x+i < 0) aResult.data()[aOffset+x] += aFilter[i]*aMatrix.data()[aOffset-1-x-i];
+ else if (x+i >= aMatrix.xSize()) aResult.data()[aOffset+x] += aFilter[i]*aMatrix.data()[aOffset+a2Size-x-i];
+ else aResult.data()[aOffset+x] += aFilter[i]*aMatrix.data()[aOffset+x+i];
+ }
+ // Center
+ for (int x = x1; x < x2; x++)
+ for (int i = aFilter.A(); i < aFilter.B(); i++)
+ aResult.data()[aOffset+x] += aFilter[i]*aMatrix.data()[aOffset+x+i];
+ // Right rim
+ for (int x = x2; x < aMatrix.xSize(); x++)
+ for (int i = aFilter.A(); i < aFilter.B(); i++) {
+ if (x+i < 0) aResult.data()[aOffset+x] += aFilter[i]*aMatrix.data()[aOffset-1-x-i];
+ else if (x+i >= aMatrix.xSize()) aResult.data()[aOffset+x] += aFilter[i]*aMatrix.data()[aOffset+a2Size-x-i];
+ else aResult.data()[aOffset+x] += aFilter[i]*aMatrix.data()[aOffset+x+i];
+ }
+ }
+}
+
+template <class T>
+inline void filter(CMatrix<T>& aMatrix, const int aDummy, const CFilter<T>& aFilter) {
+ CMatrix<T> tempMatrix(aMatrix.xSize(),aMatrix.ySize());
+ filter(aMatrix,tempMatrix,1,aFilter);
+ aMatrix = tempMatrix;
+}
+
+template <class T>
+void filter(const CMatrix<T>& aMatrix, CMatrix<T>& aResult, const int aDummy, const CFilter<T>& aFilter) {
+ if (aResult.xSize() != aMatrix.xSize() || aResult.ySize() != aMatrix.ySize())
+ throw EFilterIncompatibleSize(aMatrix.xSize()*aMatrix.ySize(),aResult.xSize()*aResult.ySize());
+ int y1 = -aFilter.A();
+ int y2 = aMatrix.ySize()-aFilter.B();
+ int a2Size = 2*aMatrix.ySize()-1;
+ // Upper rim
+ for (int y = 0; y < y1; y++)
+ for (int x = 0; x < aMatrix.xSize(); x++) {
+ aResult(x,y) = 0;
+ for (int j = aFilter.A(); j < aFilter.B(); j++) {
+ if (y+j < 0) aResult(x,y) += aFilter[j]*aMatrix(x,-1-y-j);
+ else if (y+j >= aMatrix.ySize()) aResult(x,y) += aFilter[j]*aMatrix(x,a2Size-y-j);
+ else aResult(x,y) += aFilter[j]*aMatrix(x,y+j);
+ }
+ }
+ // Lower rim
+ for (int y = y2; y < aMatrix.ySize(); y++)
+ for (int x = 0; x < aMatrix.xSize(); x++) {
+ aResult(x,y) = 0;
+ for (int j = aFilter.A(); j < aFilter.B(); j++) {
+ if (y+j < 0) aResult(x,y) += aFilter[j]*aMatrix(x,-1-y-j);
+ else if (y+j >= aMatrix.ySize()) aResult(x,y) += aFilter[j]*aMatrix(x,a2Size-y-j);
+ else aResult(x,y) += aFilter[j]*aMatrix(x,y+j);
+ }
+ }
+ // Center
+ for (int y = y1; y < y2; y++)
+ for (int x = 0; x < aMatrix.xSize(); x++) {
+ aResult(x,y) = 0;
+ for (int j = aFilter.A(); j < aFilter.B(); j++)
+ aResult(x,y) += aFilter[j]*aMatrix(x,y+j);
+ }
+}
+
+template <class T>
+inline void filter(CMatrix<T>& aMatrix, const CFilter2D<T>& aFilter) {
+ CMatrix<T> tempMatrix(aMatrix.xSize(),aMatrix.ySize());
+ filter(aMatrix,tempMatrix,aFilter);
+ aMatrix = tempMatrix;
+}
+
+template <class T>
+void filter(const CMatrix<T>& aMatrix, CMatrix<T>& aResult, const CFilter2D<T>& aFilter) {
+ if (aResult.xSize() != aMatrix.xSize() || aResult.ySize() != aMatrix.ySize())
+ throw EFilterIncompatibleSize(aMatrix.xSize()*aMatrix.ySize(),aResult.xSize()*aResult.ySize());
+ int x1 = -aFilter.AX();
+ int y1 = -aFilter.AY();
+ int x2 = aMatrix.xSize()-aFilter.BX();
+ int y2 = aMatrix.ySize()-aFilter.BY();
+ int a2XSize = 2*aMatrix.xSize()-1;
+ int a2YSize = 2*aMatrix.ySize()-1;
+ // Upper rim
+ for (int y = 0; y < y1; y++)
+ for (int x = 0; x < aMatrix.xSize(); x++) {
+ aResult(x,y) = 0;
+ for (int j = aFilter.AY(); j < aFilter.BY(); j++) {
+ int tempY;
+ if (y+j < 0) tempY = -1-y-j;
+ else if (y+j >= aMatrix.ySize()) tempY = a2YSize-y-j;
+ else tempY = y+j;
+ for (int i = aFilter.AX(); i < aFilter.BX(); i++) {
+ if (x+i < 0) aResult(x,y) += aFilter(i,j)*aMatrix(-1-x-i,tempY);
+ else if (x+i >= aMatrix.xSize()) aResult(x,y) += aFilter(i,j)*aMatrix(a2XSize-x-i,tempY);
+ else aResult(x,y) += aFilter(i,j)*aMatrix(x+i,tempY);
+ }
+ }
+ }
+ // Lower rim
+ for (int y = y2; y < aMatrix.ySize(); y++)
+ for (int x = 0; x < aMatrix.xSize(); x++) {
+ aResult(x,y) = 0;
+ for (int j = aFilter.AY(); j < aFilter.BY(); j++) {
+ int tempY;
+ if (y+j < 0) tempY = -1-y-j;
+ else if (y+j >= aMatrix.ySize()) tempY = a2YSize-y-j;
+ else tempY = y+j;
+ for (int i = aFilter.AX(); i < aFilter.BX(); i++) {
+ if (x+i < 0) aResult(x,y) += aFilter(i,j)*aMatrix(-1-x-i,tempY);
+ else if (x+i >= aMatrix.xSize()) aResult(x,y) += aFilter(i,j)*aMatrix(a2XSize-x-i,tempY);
+ else aResult(x,y) += aFilter(i,j)*aMatrix(x+i,tempY);
+ }
+ }
+ }
+ for (int y = y1; y < y2; y++) {
+ // Left rim
+ for (int x = 0; x < x1; x++) {
+ aResult(x,y) = 0;
+ for (int j = aFilter.AY(); j < aFilter.BY(); j++) {
+ for (int i = aFilter.AX(); i < aFilter.BX(); i++) {
+ if (x+i < 0) aResult(x,y) += aFilter(i,j)*aMatrix(-1-x-i,y+j);
+ else if (x+i >= aMatrix.xSize()) aResult(x,y) += aFilter(i,j)*aMatrix(a2XSize-x-i,y+j);
+ else aResult(x,y) += aFilter(i,j)*aMatrix(x+i,y+j);
+ }
+ }
+ }
+ // Right rim
+ for (int x = x2; x < aMatrix.xSize(); x++) {
+ aResult(x,y) = 0;
+ for (int j = aFilter.AY(); j < aFilter.BY(); j++) {
+ for (int i = aFilter.AX(); i < aFilter.BX(); i++) {
+ if (x+i < 0) aResult(x,y) += aFilter(i,j)*aMatrix(-1-x-i,y+j);
+ else if (x+i >= aMatrix.xSize()) aResult(x,y) += aFilter(i,j)*aMatrix(a2XSize-x-i,y+j);
+ else aResult(x,y) += aFilter(i,j)*aMatrix(x+i,y+j);
+ }
+ }
+ }
+ }
+ // Center
+ for (int y = y1; y < y2; y++)
+ for (int x = x1; x < x2; x++) {
+ aResult(x,y) = 0;
+ for (int j = aFilter.AY(); j < aFilter.BY(); j++)
+ for (int i = aFilter.AX(); i < aFilter.BX(); i++)
+ aResult(x,y) += aFilter(i,j)*aMatrix(x+i,y+j);
+ }
+}
+
+
+
+template <class T>
+inline void filterMin(CMatrix<T>& aMatrix, const CFilter<T>& aFilterX, const CFilter<T>& aFilterY) {
+ CMatrix<T> tempMatrix(aMatrix.xSize(),aMatrix.ySize());
+ filterMin(aMatrix,aMatrix,tempMatrix,aFilterX,1);
+ CMatrix<T> tmp(aMatrix);
+ filterMin(tempMatrix,tmp,aMatrix,1,aFilterY);
+}
+
+template <class T>
+inline void filterMin(const CMatrix<T>& aMatrix, CMatrix<T>& aResult, const CFilter<T>& aFilterX, const CFilter<T>& aFilterY) {
+ CMatrix<T> tempMatrix(aMatrix.xSize(),aMatrix.ySize());
+ filterMin(aMatrix,aMatrix,tempMatrix,aFilterX,1);
+ filterMin(tempMatrix,aMatrix,aResult,1,aFilterY);
+}
+
+template <class T>
+inline void filterMin(CMatrix<T>& aMatrix, const CFilter<T>& aFilter, const int aDummy) {
+ CMatrix<T> tempMatrix(aMatrix.xSize(),aMatrix.ySize());
+ filterMin(aMatrix,aMatrix,tempMatrix,aFilter,1);
+ aMatrix = tempMatrix;
+}
+
+template <class T>
+void filterMin(const CMatrix<T>& aMatrix, const CMatrix<T>& aOrig, CMatrix<T>& aResult, const CFilter<T>& aFilter, const int aDummy) {
+ if (aResult.xSize() != aMatrix.xSize() || aResult.ySize() != aMatrix.ySize())
+ throw EFilterIncompatibleSize(aMatrix.xSize()*aMatrix.ySize(),aResult.xSize()*aResult.ySize());
+ int x1 = -aFilter.A();
+ int x2 = aMatrix.xSize()-aFilter.B();
+ int a2Size = 2*aMatrix.xSize()-1;
+ aResult = 0;
+ for (int y = 0; y < aMatrix.ySize(); y++) {
+ int aOffset = y*aMatrix.xSize();
+ // Left rim
+ for (int x = 0; x < x1; x++)
+ for (int i = aFilter.A(); i < aFilter.B(); i++) {
+ int matrixIdx;
+ if (x+i < 0) matrixIdx = aOffset-1-x-i;
+ else if (x+i >= aMatrix.xSize()) matrixIdx = aOffset+a2Size-x-i;
+ else matrixIdx = aOffset+x+i;
+ if (matrixIdx == aOffset+x || aOrig.data()[matrixIdx] - 1e-5 <= aOrig.data()[aOffset+x])
+ aResult.data()[aOffset+x] += aFilter[i]*aMatrix.data()[matrixIdx];
+ }
+ // Center
+ for (int x = x1; x < x2; x++)
+ for (int i = aFilter.A(); i < aFilter.B(); i++)
+ if (i == 0 || aOrig.data()[aOffset+x+i] - 1e-5 <= aOrig.data()[aOffset+x])
+ aResult.data()[aOffset+x] += aFilter[i]*aMatrix.data()[aOffset+x+i];
+ // Right rim
+ for (int x = x2; x < aMatrix.xSize(); x++)
+ for (int i = aFilter.A(); i < aFilter.B(); i++) {
+ int matrixIdx;
+ if (x+i < 0) matrixIdx = aOffset-1-x-i;
+ else if (x+i >= aMatrix.xSize()) matrixIdx = aOffset+a2Size-x-i;
+ else matrixIdx = aOffset+x+i;
+ if (matrixIdx == aOffset+x || aOrig.data()[matrixIdx] - 1e-5 <= aOrig.data()[aOffset+x])
+ aResult.data()[aOffset+x] += aFilter[i]*aMatrix.data()[matrixIdx];
+ }
+ }
+}
+
+template <class T>
+inline void filterMin(CMatrix<T>& aMatrix, const int aDummy, const CFilter<T>& aFilter) {
+ CMatrix<T> tempMatrix(aMatrix.xSize(),aMatrix.ySize());
+ filterMin(aMatrix, aMatrix,tempMatrix,1,aFilter);
+ aMatrix = tempMatrix;
+}
+
+template <class T>
+void filterMin(const CMatrix<T>& aMatrix, const CMatrix<T>& aOrig, CMatrix<T>& aResult, const int aDummy, const CFilter<T>& aFilter) {
+ if (aResult.xSize() != aMatrix.xSize() || aResult.ySize() != aMatrix.ySize())
+ throw EFilterIncompatibleSize(aMatrix.xSize()*aMatrix.ySize(),aResult.xSize()*aResult.ySize());
+ int y1 = -aFilter.A();
+ int y2 = aMatrix.ySize()-aFilter.B();
+ int a2Size = 2*aMatrix.ySize()-1;
+ // Upper rim
+ for (int y = 0; y < y1; y++)
+ for (int x = 0; x < aMatrix.xSize(); x++) {
+ aResult(x,y) = 0;
+ for (int j = aFilter.A(); j < aFilter.B(); j++) {
+ int matrixIdx;
+ if (y+j < 0) matrixIdx = -1-y-j;
+ else if (y+j >= aMatrix.ySize()) matrixIdx = a2Size-y-j;
+ else matrixIdx = y+j;
+ if (matrixIdx == y || aOrig(x, matrixIdx) - 1e-5 <= aOrig(x, y))
+ aResult(x,y) += aFilter[j]*aMatrix(x,matrixIdx);
+ }
+ }
+ // Lower rim
+ for (int y = y2; y < aMatrix.ySize(); y++)
+ for (int x = 0; x < aMatrix.xSize(); x++) {
+ aResult(x,y) = 0;
+ for (int j = aFilter.A(); j < aFilter.B(); j++) {
+ int matrixIdx;
+ if (y+j < 0) matrixIdx = -1-y-j;
+ else if (y+j >= aMatrix.ySize()) matrixIdx = a2Size-y-j;
+ else matrixIdx = y+j;
+ if (matrixIdx == y || aOrig(x, matrixIdx) - 1e-5 <= aOrig(x, y))
+ aResult(x,y) += aFilter[j]*aMatrix(x,matrixIdx);
+ }
+ }
+ // Center
+ for (int y = y1; y < y2; y++)
+ for (int x = 0; x < aMatrix.xSize(); x++) {
+ aResult(x,y) = 0;
+ for (int j = aFilter.A(); j < aFilter.B(); j++)
+ if (j == 0 || aOrig(x,y+j) - 1e-5 <= aOrig(x,y))
+ aResult(x,y) += aFilter[j]*aMatrix(x,y+j);
+ }
+}
+
+
+
+// boxfilterX
+template <class T>
+inline void boxFilterX(CMatrix<T>& aMatrix, int aWidth) {
+ CMatrix<T> aTemp(aMatrix);
+ boxFilterX(aTemp,aMatrix,aWidth);
+}
+
+template <class T>
+void boxFilterX(const CMatrix<T>& aMatrix, CMatrix<T>& aResult, int aWidth) {
+ if (aWidth & 1 == 0) aWidth += 1;
+ T invWidth = 1.0/aWidth;
+ int halfWidth = (aWidth >> 1);
+ int aRight = halfWidth;
+ if (aRight >= aMatrix.xSize()) aRight = aMatrix.xSize()-1;
+ for (int y = 0; y < aMatrix.ySize(); y++) {
+ int aOffset = y*aMatrix.xSize();
+ // Initialize
+ T aSum = 0.0f;
+ for (int x = aRight-aWidth+1; x <= aRight; x++)
+ if (x < 0) aSum += aMatrix.data()[aOffset-x-1];
+ else aSum += aMatrix.data()[aOffset+x];
+ // Shift
+ int xm = -halfWidth;
+ int xp = halfWidth+1;
+ for (int x = 0; x < aMatrix.xSize(); x++,xm++,xp++) {
+ aResult.data()[aOffset+x] = aSum*invWidth;
+ if (xm < 0) aSum -= aMatrix.data()[aOffset-xm-1];
+ else aSum -= aMatrix.data()[aOffset+xm];
+ if (xp >= aMatrix.xSize()) aSum += aMatrix.data()[aOffset+2*aMatrix.xSize()-1-xp];
+ else aSum += aMatrix.data()[aOffset+xp];
+ }
+ }
+}
+
+// boxfilterY
+template <class T>
+inline void boxFilterY(CMatrix<T>& aMatrix, int aWidth) {
+ CMatrix<T> aTemp(aMatrix);
+ boxFilterY(aTemp,aMatrix,aWidth);
+}
+
+template <class T>
+void boxFilterY(const CMatrix<T>& aMatrix, CMatrix<T>& aResult, int aWidth) {
+ if (aWidth & 1 == 0) aWidth += 1;
+ T invWidth = 1.0/aWidth;
+ int halfWidth = (aWidth >> 1);
+ int aBottom = halfWidth;
+ if (aBottom >= aMatrix.ySize()) aBottom = aMatrix.xSize()-1;
+ for (int x = 0; x < aMatrix.xSize(); x++) {
+ // Initialize
+ T aSum = 0.0f;
+ for (int y = aBottom-aWidth+1; y <= aBottom; y++)
+ if (y < 0) aSum += aMatrix(x,-1-y);
+ else aSum += aMatrix(x,y);
+ // Shift
+ int ym = -halfWidth;
+ int yp = halfWidth+1;
+ for (int y = 0; y < aMatrix.ySize(); y++,ym++,yp++) {
+ aResult(x,y) = aSum*invWidth;
+ if (ym < 0) aSum -= aMatrix(x,-1-ym);
+ else aSum -= aMatrix(x,ym);
+ if (yp >= aMatrix.ySize()) aSum += aMatrix(x,2*aMatrix.ySize()-1-yp);
+ else aSum += aMatrix(x,yp);
+ }
+ }
+}
+
+template <class T>
+void recursiveSmoothX(CMatrix<T>& aMatrix, float aSigma) {
+ CVector<T> aVals1(aMatrix.xSize());
+ CVector<T> aVals2(aMatrix.xSize());
+ float aAlpha = 2.5/(sqrt(NMath::Pi)*aSigma);
+ float aExp = exp(-aAlpha);
+ float aExpSqr = aExp*aExp;
+ float a2Exp = 2.0*aExp;
+ float k = (1.0-aExp)*(1.0-aExp)/(1.0+2.0*aAlpha*aExp-aExpSqr);
+ float aPreMinus = aExp*(aAlpha-1.0);
+ float aPrePlus = aExp*(aAlpha+1.0);
+ for (int y = 0; y < aMatrix.ySize(); y++) {
+ aVals1(0) = (0.5f-k*aPreMinus)*aMatrix(0,y);
+ aVals1(1) = k*(aMatrix(1,y)+aPreMinus*aMatrix(0,y))+(a2Exp-aExpSqr)*aVals1(0);
+ for (int x = 2; x < aMatrix.xSize(); x++)
+ aVals1(x) = k*(aMatrix(x,y)+aPreMinus*aMatrix(x-1,y))+a2Exp*aVals1(x-1)-aExpSqr*aVals1(x-2);
+ aVals2(aMatrix.xSize()-1) = (0.5f+k*aPreMinus)*aMatrix(aMatrix.xSize()-1,y);
+ aVals2(aMatrix.xSize()-2) = k*((aPrePlus-aExpSqr)*aMatrix(aMatrix.xSize()-1,y))+(a2Exp-aExpSqr)*aVals2(aMatrix.xSize()-1);
+ for (int x = aMatrix.xSize()-3; x >= 0; x--)
+ aVals2(x) = k*(aPrePlus*aMatrix(x+1,y)-aExpSqr*aMatrix(x+2,y))+a2Exp*aVals2(x+1)-aExpSqr*aVals2(x+2);
+ for (int x = 0; x < aMatrix.xSize(); x++)
+ aMatrix(x,y) = aVals1(x)+aVals2(x);
+ }
+}
+
+template <class T>
+void recursiveSmoothY(CMatrix<T>& aMatrix, float aSigma) {
+ CVector<T> aVals1(aMatrix.ySize());
+ CVector<T> aVals2(aMatrix.ySize());
+ float aAlpha = 2.5/(sqrt(NMath::Pi)*aSigma);
+ float aExp = exp(-aAlpha);
+ float aExpSqr = aExp*aExp;
+ float a2Exp = 2.0*aExp;
+ float k = (1.0-aExp)*(1.0-aExp)/(1.0+2.0*aAlpha*aExp-aExpSqr);
+ float aPreMinus = aExp*(aAlpha-1.0);
+ float aPrePlus = aExp*(aAlpha+1.0);
+ for (int x = 0; x < aMatrix.xSize(); x++) {
+ aVals1(0) = (0.5f-k*aPreMinus)*aMatrix(x,0);
+ aVals1(1) = k*(aMatrix(x,1)+aPreMinus*aMatrix(x,0))+(a2Exp-aExpSqr)*aVals1(0);
+ for (int y = 2; y < aMatrix.ySize(); y++)
+ aVals1(y) = k*(aMatrix(x,y)+aPreMinus*aMatrix(x,y-1))+a2Exp*aVals1(y-1)-aExpSqr*aVals1(y-2);
+ aVals2(aMatrix.ySize()-1) = (0.5f+k*aPreMinus)*aMatrix(x,aMatrix.ySize()-1);
+ aVals2(aMatrix.ySize()-2) = k*((aPrePlus-aExpSqr)*aMatrix(x,aMatrix.ySize()-1))+(a2Exp-aExpSqr)*aVals2(aMatrix.ySize()-1);
+ for (int y = aMatrix.ySize()-3; y >= 0; y--)
+ aVals2(y) = k*(aPrePlus*aMatrix(x,y+1)-aExpSqr*aMatrix(x,y+2))+a2Exp*aVals2(y+1)-aExpSqr*aVals2(y+2);
+ for (int y = 0; y < aMatrix.ySize(); y++)
+ aMatrix(x,y) = aVals1(y)+aVals2(y);
+ }
+}
+
+template <class T>
+inline void recursiveSmooth(CMatrix<T>& aMatrix, float aSigma) {
+ recursiveSmoothX(aMatrix,aSigma);
+ recursiveSmoothY(aMatrix,aSigma);
+}
+
+// Linear 3D filtering ---------------------------------------------------------
+
+template <class T>
+inline void filter(CTensor<T>& aTensor, const CFilter<T>& aFilterX, const CFilter<T>& aFilterY, const CFilter<T>& aFilterZ) {
+ CTensor<T> tempTensor(aTensor.xSize(),aTensor.ySize(),aTensor.zSize());
+ filter(aTensor,tempTensor,aFilterX,1,1);
+ filter(tempTensor,aTensor,1,aFilterY,1);
+ filter(aTensor,tempTensor,1,1,aFilterZ);
+ aTensor = tempTensor;
+}
+
+template <class T>
+inline void filter(const CTensor<T>& aTensor, CTensor<T>& aResult, const CFilter<T>& aFilterX, const CFilter<T>& aFilterY, const CFilter<T>& aFilterZ) {
+ CTensor<T> tempTensor(aTensor.xSize(),aTensor.ySize(),aTensor.zSize());
+ filter(aTensor,aResult,aFilterX,1,1);
+ filter(aResult,tempTensor,1,aFilterY,1);
+ filter(tempTensor,aResult,1,1,aFilterZ);
+}
+
+template <class T>
+inline void filter(CTensor<T>& aTensor, const CFilter<T>& aFilter, const int aDummy1, const int aDummy2) {
+ CTensor<T> tempTensor(aTensor.xSize(),aTensor.ySize(),aTensor.zSize());
+ filter(aTensor,tempTensor,aFilter,1,1);
+ aTensor = tempTensor;
+}
+
+template <class T>
+void filter(const CTensor<T>& aTensor, CTensor<T>& aResult, const CFilter<T>& aFilter, const int aDummy1, const int aDummy2) {
+ if (aResult.xSize() != aTensor.xSize() || aResult.ySize() != aTensor.ySize() || aResult.zSize() != aTensor.zSize())
+ throw EFilterIncompatibleSize(aTensor.xSize()*aTensor.ySize()*aTensor.zSize(),aResult.xSize()*aResult.ySize()*aResult.zSize());
+ int x1 = -aFilter.A();
+ int x2 = aTensor.xSize()-aFilter.B();
+ int a2Size = 2*aTensor.xSize()-1;
+ for (int z = 0; z < aTensor.zSize(); z++)
+ for (int y = 0; y < aTensor.ySize(); y++) {
+ // Left rim
+ for (int x = 0; x < x1; x++) {
+ aResult(x,y,z) = 0;
+ for (int i = aFilter.A(); i < aFilter.B(); i++) {
+ if (x+i < 0) aResult(x,y,z) += aFilter[i]*aTensor(-1-x-i,y,z);
+ else if (x+i >= aTensor.xSize()) aResult(x,y,z) += aFilter[i]*aTensor(a2Size-x-i,y,z);
+ else aResult(x,y,z) += aFilter[i]*aTensor(x+i,y,z);
+ }
+ }
+ // Center
+ for (int x = x1; x < x2; x++) {
+ aResult(x,y,z) = 0;
+ for (int i = aFilter.A(); i < aFilter.B(); i++)
+ aResult(x,y,z) += aFilter[i]*aTensor(x+i,y,z);
+ }
+ // Right rim
+ for (int x = x2; x < aTensor.xSize(); x++) {
+ aResult(x,y,z) = 0;
+ for (int i = aFilter.A(); i < aFilter.B(); i++) {
+ if (x+i < 0) aResult(x,y,z) += aFilter[i]*aTensor(-1-x-i,y,z);
+ else if (x+i >= aTensor.xSize()) aResult(x,y,z) += aFilter[i]*aTensor(a2Size-x-i,y,z);
+ else aResult(x,y,z) += aFilter[i]*aTensor(x+i,y,z);
+ }
+ }
+ }
+}
+
+template <class T>
+inline void filter(CTensor<T>& aTensor, const int aDummy1, const CFilter<T>& aFilter, const int aDummy2) {
+ CTensor<T> tempTensor(aTensor.xSize(),aTensor.ySize(),aTensor.zSize());
+ filter(aTensor,tempTensor,1,aFilter,1);
+ aTensor = tempTensor;
+}
+
+template <class T>
+void filter(const CTensor<T>& aTensor, CTensor<T>& aResult, const int aDummy1, const CFilter<T>& aFilter, const int aDummy2) {
+ if (aResult.xSize() != aTensor.xSize() || aResult.ySize() != aTensor.ySize() || aResult.zSize() != aTensor.zSize())
+ throw EFilterIncompatibleSize(aTensor.xSize()*aTensor.ySize()*aTensor.zSize(),aResult.xSize()*aResult.ySize()*aResult.zSize());
+ int y1 = -aFilter.A();
+ int y2 = aTensor.ySize()-aFilter.B();
+ int a2Size = 2*aTensor.ySize()-1;
+ for (int z = 0; z < aTensor.zSize(); z++) {
+ // Upper rim
+ for (int y = 0; y < y1; y++)
+ for (int x = 0; x < aTensor.xSize(); x++) {
+ aResult(x,y,z) = 0;
+ for (int i = aFilter.A(); i < aFilter.B(); i++) {
+ if (y+i < 0) aResult(x,y,z) += aFilter[i]*aTensor(x,-1-y-i,z);
+ else if (y+i >= aTensor.ySize()) aResult(x,y,z) += aFilter[i]*aTensor(x,a2Size-y-i,z);
+ else aResult(x,y,z) += aFilter[i]*aTensor(x,y+i,z);
+ }
+ }
+ // Lower rim
+ for (int y = y2; y < aTensor.ySize(); y++)
+ for (int x = 0; x < aTensor.xSize(); x++) {
+ aResult(x,y,z) = 0;
+ for (int i = aFilter.A(); i < aFilter.B(); i++) {
+ if (y+i < 0) aResult(x,y,z) += aFilter[i]*aTensor(x,-1-y-i,z);
+ else if (y+i >= aTensor.ySize()) aResult(x,y,z) += aFilter[i]*aTensor(x,a2Size-y-i,z);
+ else aResult(x,y,z) += aFilter[i]*aTensor(x,y+i,z);
+ }
+ }
+ }
+ // Center
+ for (int z = 0; z < aTensor.zSize(); z++)
+ for (int y = y1; y < y2; y++)
+ for (int x = 0; x < aTensor.xSize(); x++) {
+ aResult(x,y,z) = 0;
+ for (int i = aFilter.A(); i < aFilter.B(); i++)
+ aResult(x,y,z) += aFilter[i]*aTensor(x,y+i,z);
+ }
+}
+
+template <class T>
+inline void filter(CTensor<T>& aTensor, const int aDummy1, const int aDummy2, const CFilter<T>& aFilter) {
+ CTensor<T> tempTensor(aTensor.xSize(),aTensor.ySize(),aTensor.zSize());
+ filter(aTensor,tempTensor,1,1,aFilter);
+ aTensor = tempTensor;
+}
+
+template <class T>
+void filter(const CTensor<T>& aTensor, CTensor<T>& aResult, const int aDummy1, const int aDummy2, const CFilter<T>& aFilter) {
+ if (aResult.xSize() != aTensor.xSize() || aResult.ySize() != aTensor.ySize() || aResult.zSize() != aTensor.zSize())
+ throw EFilterIncompatibleSize(aTensor.xSize()*aTensor.ySize()*aTensor.zSize(),aResult.xSize()*aResult.ySize()*aResult.zSize());
+ int z1 = -aFilter.A();
+ int z2 = aTensor.zSize()-aFilter.B();
+ if (z2 < 0) z2 = 0;
+ int a2Size = 2*aTensor.zSize()-1;
+ // Front rim
+ for (int z = 0; z < z1; z++)
+ for (int y = 0; y < aTensor.ySize(); y++)
+ for (int x = 0; x < aTensor.xSize(); x++) {
+ aResult(x,y,z) = 0;
+ for (int i = aFilter.A(); i < aFilter.B(); i++) {
+ if (z+i < 0) aResult(x,y,z) += aFilter[i]*aTensor(x,y,-1-z-i);
+ else if (z+i >= aTensor.zSize()) aResult(x,y,z) += aFilter[i]*aTensor(x,y,a2Size-z-i);
+ else aResult(x,y,z) += aFilter[i]*aTensor(x,y,z+i);
+ }
+ }
+ // Back rim
+ for (int z = z2; z < aTensor.zSize(); z++)
+ for (int y = 0; y < aTensor.ySize(); y++)
+ for (int x = 0; x < aTensor.xSize(); x++) {
+ aResult(x,y,z) = 0;
+ for (int i = aFilter.A(); i < aFilter.B(); i++) {
+ if (z+i < 0) aResult(x,y,z) += aFilter[i]*aTensor(x,y,-1-z-i);
+ else if (z+i >= aTensor.zSize()) aResult(x,y,z) += aFilter[i]*aTensor(x,y,a2Size-z-i);
+ else aResult(x,y,z) += aFilter[i]*aTensor(x,y,z+i);
+ }
+ }
+ // Center
+ for (int z = z1; z < z2; z++)
+ for (int y = 0; y < aTensor.ySize(); y++)
+ for (int x = 0; x < aTensor.xSize(); x++) {
+ aResult(x,y,z) = 0;
+ for (int i = aFilter.A(); i < aFilter.B(); i++)
+ aResult(x,y,z) += aFilter[i]*aTensor(x,y,z+i);
+ }
+}
+
+// boxfilterX
+template <class T>
+inline void boxFilterX(CTensor<T>& aTensor, int aWidth) {
+ CTensor<T> aTemp(aTensor);
+ boxFilterX(aTemp,aTensor,aWidth);
+}
+
+template <class T>
+void boxFilterX(const CTensor<T>& aTensor, CTensor<T>& aResult, int aWidth) {
+ if (aWidth % 2 == 0) aWidth += 1;
+ T* invWidth = new T[aWidth+1];
+ invWidth[0] = 1.0f;
+ for (int i = 1; i <= aWidth; i++)
+ invWidth[i] = 1.0/i;
+ int halfWidth = (aWidth >> 1);
+ int aRight = halfWidth;
+ if (aRight >= aTensor.xSize()) aRight = aTensor.xSize()-1;
+ for (int z = 0; z < aTensor.zSize(); z++)
+ for (int y = 0; y < aTensor.ySize(); y++) {
+ int aOffset = (z*aTensor.ySize()+y)*aTensor.xSize();
+ // Initialize
+ int aNum = 0;
+ T aSum = 0.0f;
+ for (int x = 0; x <= aRight; x++) {
+ aSum += aTensor.data()[aOffset+x]; aNum++;
+ }
+ // Shift
+ for (int x = 0; x < aTensor.xSize(); x++) {
+ aResult.data()[aOffset+x] = aSum*invWidth[aNum];
+ if (x-halfWidth >= 0) {
+ aSum -= aTensor.data()[aOffset+x-halfWidth]; aNum--;
+ }
+ if (x+halfWidth+1 < aTensor.xSize()) {
+ aSum += aTensor.data()[aOffset+x+halfWidth+1]; aNum++;
+ }
+ }
+ }
+ delete[] invWidth;
+}
+
+// boxfilterY
+template <class T>
+inline void boxFilterY(CTensor<T>& aTensor, int aWidth) {
+ CTensor<T> aTemp(aTensor);
+ boxFilterY(aTemp,aTensor,aWidth);
+}
+
+template <class T>
+void boxFilterY(const CTensor<T>& aTensor, CTensor<T>& aResult, int aWidth) {
+ if (aWidth % 2 == 0) aWidth += 1;
+ T* invWidth = new T[aWidth+1];
+ invWidth[0] = 1.0f;
+ for (int i = 1; i <= aWidth; i++)
+ invWidth[i] = 1.0/i;
+ int halfWidth = (aWidth >> 1);
+ int aBottom = halfWidth;
+ if (aBottom >= aTensor.ySize()) aBottom = aTensor.ySize()-1;
+ for (int z = 0; z < aTensor.zSize(); z++)
+ for (int x = 0; x < aTensor.xSize(); x++) {
+ // Initialize
+ int aNum = 0;
+ T aSum = 0.0f;
+ for (int y = 0; y <= aBottom; y++) {
+ aSum += aTensor(x,y,z); aNum++;
+ }
+ // Shift
+ for (int y = 0; y < aTensor.ySize(); y++) {
+ aResult(x,y,z) = aSum*invWidth[aNum];
+ if (y-halfWidth >= 0) {
+ aSum -= aTensor(x,y-halfWidth,z); aNum--;
+ }
+ if (y+halfWidth+1 < aTensor.ySize()) {
+ aSum += aTensor(x,y+halfWidth+1,z); aNum++;
+ }
+ }
+ }
+ delete[] invWidth;
+}
+
+// boxfilterZ
+template <class T>
+inline void boxFilterZ(CTensor<T>& aTensor, int aWidth) {
+ CTensor<T> aTemp(aTensor);
+ boxFilterZ(aTemp,aTensor,aWidth);
+}
+
+template <class T>
+void boxFilterZ(const CTensor<T>& aTensor, CTensor<T>& aResult, int aWidth) {
+ if (aWidth % 2 == 0) aWidth += 1;
+ T* invWidth = new T[aWidth+1];
+ invWidth[0] = 1.0f;
+ for (int i = 1; i <= aWidth; i++)
+ invWidth[i] = 1.0/i;
+ int halfWidth = (aWidth >> 1);
+ int aBottom = halfWidth;
+ if (aBottom >= aTensor.zSize()) aBottom = aTensor.zSize()-1;
+ for (int y = 0; y < aTensor.ySize(); y++)
+ for (int x = 0; x < aTensor.xSize(); x++) {
+ // Initialize
+ int aNum = 0;
+ T aSum = 0.0f;
+ for (int z = 0; z <= aBottom; z++) {
+ aSum += aTensor(x,y,z); aNum++;
+ }
+ // Shift
+ for (int z = 0; z < aTensor.zSize(); z++) {
+ aResult(x,y,z) = aSum*invWidth[aNum];
+ if (z-halfWidth >= 0) {
+ aSum -= aTensor(x,y,z-halfWidth); aNum--;
+ }
+ if (z+halfWidth+1 < aTensor.zSize()) {
+ aSum += aTensor(x,y,z+halfWidth+1); aNum++;
+ }
+ }
+ }
+ delete[] invWidth;
+}
+
+template <class T>
+void recursiveSmoothX(CTensor<T>& aTensor, float aSigma) {
+ CVector<T> aVals1(aTensor.xSize());
+ CVector<T> aVals2(aTensor.xSize());
+ float aAlpha = 2.5/(sqrt(NMath::Pi)*aSigma);
+ float aExp = exp(-aAlpha);
+ float aExpSqr = aExp*aExp;
+ float a2Exp = 2.0*aExp;
+ float k = (1.0-aExp)*(1.0-aExp)/(1.0+2.0*aAlpha*aExp-aExpSqr);
+ float aPreMinus = aExp*(aAlpha-1.0);
+ float aPrePlus = aExp*(aAlpha+1.0);
+ for (int z = 0; z < aTensor.zSize(); z++)
+ for (int y = 0; y < aTensor.ySize(); y++) {
+ int aOffset = (z*aTensor.ySize()+y)*aTensor.xSize();
+ aVals1(0) = (0.5-k*aPreMinus)*aTensor.data()[aOffset];
+ aVals1(1) = k*(aTensor.data()[aOffset+1]+aPreMinus*aTensor.data()[aOffset])+(2.0*aExp-aExpSqr)*aVals1(0);
+ for (int x = 2; x < aTensor.xSize(); x++)
+ aVals1(x) = k*(aTensor.data()[aOffset+x]+aPreMinus*aTensor.data()[aOffset+x-1])+a2Exp*aVals1(x-1)-aExpSqr*aVals1(x-2);
+ aVals2(aTensor.xSize()-1) = (0.5+k*aPreMinus)*aTensor.data()[aOffset+aTensor.xSize()-1];
+ aVals2(aTensor.xSize()-2) = k*((aPrePlus-aExpSqr)*aTensor.data()[aOffset+aTensor.xSize()-1])+(a2Exp-aExpSqr)*aVals2(aTensor.xSize()-1);
+ for (int x = aTensor.xSize()-3; x >= 0; x--)
+ aVals2(x) = k*(aPrePlus*aTensor.data()[aOffset+x+1]-aExpSqr*aTensor.data()[aOffset+x+2])+a2Exp*aVals2(x+1)-aExpSqr*aVals2(x+2);
+ for (int x = 0; x < aTensor.xSize(); x++)
+ aTensor.data()[aOffset+x] = aVals1(x)+aVals2(x);
+ }
+}
+
+template <class T>
+void recursiveSmoothY(CTensor<T>& aTensor, float aSigma) {
+ CVector<T> aVals1(aTensor.ySize());
+ CVector<T> aVals2(aTensor.ySize());
+ float aAlpha = 2.5/(sqrt(NMath::Pi)*aSigma);
+ float aExp = exp(-aAlpha);
+ float aExpSqr = aExp*aExp;
+ float a2Exp = 2.0*aExp;
+ float k = (1.0-aExp)*(1.0-aExp)/(1.0+2.0*aAlpha*aExp-aExpSqr);
+ float aPreMinus = aExp*(aAlpha-1.0);
+ float aPrePlus = aExp*(aAlpha+1.0);
+ for (int z = 0; z < aTensor.zSize(); z++)
+ for (int x = 0; x < aTensor.xSize(); x++) {
+ aVals1(0) = (0.5-k*aPreMinus)*aTensor(x,0,z);
+ aVals1(1) = k*(aTensor(x,1,z)+aPreMinus*aTensor(x,0,z))+(2.0*aExp-aExpSqr)*aVals1(0);
+ for (int y = 2; y < aTensor.ySize(); y++)
+ aVals1(y) = k*(aTensor(x,y,z)+aPreMinus*aTensor(x,y-1,z))+a2Exp*aVals1(y-1)-aExpSqr*aVals1(y-2);
+ aVals2(aTensor.ySize()-1) = (0.5+k*aPreMinus)*aTensor(x,aTensor.ySize()-1,z);
+ aVals2(aTensor.ySize()-2) = k*((aPrePlus-aExpSqr)*aTensor(x,aTensor.ySize()-1,z))+(a2Exp-aExpSqr)*aVals2(aTensor.ySize()-1);
+ for (int y = aTensor.ySize()-3; y >= 0; y--)
+ aVals2(y) = k*(aPrePlus*aTensor(x,y+1,z)-aExpSqr*aTensor(x,y+2,z))+a2Exp*aVals2(y+1)-aExpSqr*aVals2(y+2);
+ for (int y = 0; y < aTensor.ySize(); y++)
+ aTensor(x,y,z) = aVals1(y)+aVals2(y);
+ }
+}
+
+template <class T>
+void recursiveSmoothZ(CTensor<T>& aTensor, float aSigma) {
+ CVector<T> aVals1(aTensor.zSize());
+ CVector<T> aVals2(aTensor.zSize());
+ float aAlpha = 2.5/(sqrt(NMath::Pi)*aSigma);
+ float aExp = exp(-aAlpha);
+ float aExpSqr = aExp*aExp;
+ float a2Exp = 2.0*aExp;
+ float k = (1.0-aExp)*(1.0-aExp)/(1.0+2.0*aAlpha*aExp-aExpSqr);
+ float aPreMinus = aExp*(aAlpha-1.0);
+ float aPrePlus = aExp*(aAlpha+1.0);
+ for (int y = 0; y < aTensor.ySize(); y++)
+ for (int x = 0; x < aTensor.xSize(); x++) {
+ aVals1(0) = (0.5-k*aPreMinus)*aTensor(x,y,0);
+ aVals1(1) = k*(aTensor(x,y,1)+aPreMinus*aTensor(x,y,0))+(2.0*aExp-aExpSqr)*aVals1(0);
+ for (int z = 2; z < aTensor.zSize(); z++)
+ aVals1(z) = k*(aTensor(x,y,z)+aPreMinus*aTensor(x,y,z-1))+a2Exp*aVals1(z-1)-aExpSqr*aVals1(z-2);
+ aVals2(aTensor.zSize()-1) = (0.5+k*aPreMinus)*aTensor(x,y,aTensor.zSize()-1);
+ aVals2(aTensor.zSize()-2) = k*((aPrePlus-aExpSqr)*aTensor(x,y,aTensor.zSize()-1))+(a2Exp-aExpSqr)*aVals2(aTensor.zSize()-1);
+ for (int z = aTensor.zSize()-3; z >= 0; z--)
+ aVals2(z) = k*(aPrePlus*aTensor(x,y,z+1)-aExpSqr*aTensor(x,y,z+2))+a2Exp*aVals2(z+1)-aExpSqr*aVals2(z+2);
+ for (int z = 0; z < aTensor.zSize(); z++)
+ aTensor(x,y,z) = aVals1(z)+aVals2(z);
+ }
+}
+
+// Linear 4D filtering ---------------------------------------------------------
+
+template <class T>
+inline void filter(CTensor4D<T>& aTensor, const CFilter<T>& aFilterX, const CFilter<T>& aFilterY, const CFilter<T>& aFilterZ, const CFilter<T>& aFilterA) {
+ CTensor4D<T> tempTensor(aTensor.xSize(),aTensor.ySize(),aTensor.zSize());
+ filter(aTensor,tempTensor,aFilterX,1,1,1);
+ filter(tempTensor,aTensor,1,aFilterY,1,1);
+ filter(aTensor,tempTensor,1,1,aFilterZ,1);
+ filter(tempTensor,aTensor,1,1,1,aFilterA);
+}
+
+template <class T>
+inline void filter(CTensor4D<T>& aTensor, const CFilter<T>& aFilter, const int aDummy1, const int aDummy2, const int aDummy3) {
+ CTensor4D<T> tempTensor(aTensor.xSize(),aTensor.ySize(),aTensor.zSize(),aTensor.aSize());
+ filter(aTensor,tempTensor,aFilter,1,1,1);
+ aTensor = tempTensor;
+}
+
+template <class T>
+void filter(const CTensor4D<T>& aTensor, CTensor4D<T>& aResult, const CFilter<T>& aFilter, const int aDummy1, const int aDummy2, const int aDummy3) {
+ if (aResult.xSize() != aTensor.xSize() || aResult.ySize() != aTensor.ySize() || aResult.zSize() != aTensor.zSize() || aResult.aSize() != aTensor.aSize())
+ throw EFilterIncompatibleSize(aTensor.xSize()*aTensor.ySize()*aTensor.zSize()*aTensor.aSize(),aResult.xSize()*aResult.ySize()*aResult.zSize()*aResult.aSize());
+ int x1 = -aFilter.A();
+ int x2 = aTensor.xSize()-aFilter.B();
+ int a2Size = 2*aTensor.xSize()-1;
+ aResult = 0;
+ for (int a = 0; a < aTensor.aSize(); a++)
+ for (int z = 0; z < aTensor.zSize(); z++)
+ for (int y = 0; y < aTensor.ySize(); y++) {
+ int aOffset = aTensor.xSize()*(y+aTensor.ySize()*(z+aTensor.zSize()*a));
+ // Left rim
+ for (int x = 0; x < x1; x++)
+ for (int i = aFilter.A(); i < aFilter.B(); i++) {
+ if (x+i < 0) aResult.data()[aOffset+x] += aFilter[i]*aTensor.data()[aOffset-1-x-i];
+ else if (x+i >= aTensor.xSize()) aResult.data()[aOffset+x] += aFilter[i]*aTensor.data()[aOffset+a2Size-x-i];
+ else aResult.data()[aOffset+x] += aFilter[i]*aTensor.data()[x+i+aOffset];
+ }
+ // Center
+ for (int x = x1; x < x2; x++)
+ for (int i = aFilter.A(); i < aFilter.B(); i++)
+ aResult.data()[aOffset+x] += aFilter[i]*aTensor.data()[aOffset+x+i];
+ // Right rim
+ for (int x = x2; x < aTensor.xSize(); x++)
+ for (int i = aFilter.A(); i < aFilter.B(); i++) {
+ if (x+i < 0) aResult.data()[aOffset+x] += aFilter[i]*aTensor.data()[aOffset-1-x-i];
+ else if (x+i >= aTensor.xSize()) aResult.data()[aOffset+x] += aFilter[i]*aTensor.data()[aOffset+a2Size-x-i];
+ else aResult.data()[aOffset+x] += aFilter[i]*aTensor.data()[x+i+aOffset];
+ }
+ }
+}
+
+template <class T>
+inline void filter(CTensor4D<T>& aTensor, const int aDummy1, const CFilter<T>& aFilter, const int aDummy2, const int aDummy3) {
+ CTensor4D<T> tempTensor(aTensor.xSize(),aTensor.ySize(),aTensor.zSize(),aTensor.aSize());
+ filter(aTensor,tempTensor,1,aFilter,1,1);
+ aTensor = tempTensor;
+}
+
+template <class T>
+void filter(const CTensor4D<T>& aTensor, CTensor4D<T>& aResult, const int aDummy1, const CFilter<T>& aFilter, const int aDummy2, const int aDummy3) {
+ if (aResult.xSize() != aTensor.xSize() || aResult.ySize() != aTensor.ySize() || aResult.zSize() != aTensor.zSize() || aResult.aSize() != aTensor.aSize())
+ throw EFilterIncompatibleSize(aTensor.xSize()*aTensor.ySize()*aTensor.zSize()*aTensor.aSize(),aResult.xSize()*aResult.ySize()*aResult.zSize()*aResult.aSize());
+ int y1 = -aFilter.A();
+ int y2 = aTensor.ySize()-aFilter.B();
+ int a2Size = 2*aTensor.ySize()-1;
+ aResult = 0;
+ for (int a = 0; a < aTensor.aSize(); a++) {
+ for (int z = 0; z < aTensor.zSize(); z++) {
+ // Upper rim
+ for (int y = 0; y < y1; y++)
+ for (int x = 0; x < aTensor.xSize(); x++)
+ for (int i = aFilter.A(); i < aFilter.B(); i++) {
+ if (y+i < 0) aResult(x,y,z,a) += aFilter[i]*aTensor(x,-1-y-i,z,a);
+ else if (y+i >= aTensor.ySize()) aResult(x,y,z,a) += aFilter[i]*aTensor(x,a2Size-y-i,z,a);
+ else aResult(x,y,z,a) += aFilter[i]*aTensor(x,y+i,z,a);
+ }
+ // Lower rim
+ for (int y = y2; y < aTensor.ySize(); y++)
+ for (int x = 0; x < aTensor.xSize(); x++)
+ for (int i = aFilter.A(); i < aFilter.B(); i++) {
+ if (y+i < 0) aResult(x,y,z,a) += aFilter[i]*aTensor(x,-1-y-i,z,a);
+ else if (y+i >= aTensor.ySize()) aResult(x,y,z,a) += aFilter[i]*aTensor(x,a2Size-y-i,z,a);
+ else aResult(x,y,z,a) += aFilter[i]*aTensor(x,y+i,z,a);
+ }
+ }
+ // Center
+ for (int z = 0; z < aTensor.zSize(); z++)
+ for (int y = y1; y < y2; y++)
+ for (int x = 0; x < aTensor.xSize(); x++)
+ for (int i = aFilter.A(); i < aFilter.B(); i++)
+ aResult(x,y,z,a) += aFilter[i]*aTensor(x,y+i,z,a);
+ }
+}
+
+template <class T>
+inline void filter(CTensor4D<T>& aTensor, const int aDummy1, const int aDummy2, const CFilter<T>& aFilter, const int aDummy3) {
+ CTensor4D<T> tempTensor(aTensor.xSize(),aTensor.ySize(),aTensor.zSize(),aTensor.aSize());
+ filter(aTensor,tempTensor,1,1,aFilter,1);
+ aTensor = tempTensor;
+}
+
+template <class T>
+void filter(const CTensor4D<T>& aTensor, CTensor4D<T>& aResult, const int aDummy1, const int aDummy2, const CFilter<T>& aFilter, const int aDummy3) {
+ if (aResult.xSize() != aTensor.xSize() || aResult.ySize() != aTensor.ySize() || aResult.zSize() != aTensor.zSize() || aResult.aSize() != aTensor.aSize())
+ throw EFilterIncompatibleSize(aTensor.xSize()*aTensor.ySize()*aTensor.zSize()*aTensor.aSize(),aResult.xSize()*aResult.ySize()*aResult.zSize()*aResult.aSize());
+ int z1 = -aFilter.A();
+ int z2 = aTensor.zSize()-aFilter.B();
+ int a2Size = 2*aTensor.zSize()-1;
+ aResult = 0;
+ for (int a = 0; a < aTensor.aSize(); a++) {
+ // Front rim
+ for (int z = 0; z < z1; z++)
+ for (int y = 0; y < aTensor.ySize(); y++)
+ for (int x = 0; x < aTensor.xSize(); x++)
+ for (int i = aFilter.A(); i < aFilter.B(); i++) {
+ if (z+i < 0) aResult(x,y,z,a) += aFilter[i]*aTensor(x,y,-1-z-i,a);
+ else if (z+i >= aTensor.zSize()) aResult(x,y,z,a) += aFilter[i]*aTensor(x,y,a2Size-z-i,a);
+ else aResult(x,y,z,a) += aFilter[i]*aTensor(x,y,z+i,a);
+ }
+ // Back rim
+ for (int z = z2; z < aTensor.zSize(); z++)
+ for (int y = 0; y < aTensor.ySize(); y++)
+ for (int x = 0; x < aTensor.xSize(); x++)
+ for (int i = aFilter.A(); i < aFilter.B(); i++) {
+ if (z+i < 0) aResult(x,y,z,a) += aFilter[i]*aTensor(x,y,-1-z-i,a);
+ else if (z+i >= aTensor.zSize()) aResult(x,y,z,a) += aFilter[i]*aTensor(x,y,a2Size-z-i,a);
+ else aResult(x,y,z,a) += aFilter[i]*aTensor(x,y,z+i,a);
+ }
+ // Center
+ for (int z = z1; z < z2; z++)
+ for (int y = 0; y < aTensor.ySize(); y++)
+ for (int x = 0; x < aTensor.xSize(); x++)
+ for (int i = aFilter.A(); i < aFilter.B(); i++)
+ aResult(x,y,z,a) += aFilter[i]*aTensor(x,y,z+i,a);
+ }
+}
+
+template <class T>
+inline void filter(CTensor4D<T>& aTensor, const int aDummy1, const int aDummy2, const int aDummy3, const CFilter<T>& aFilter) {
+ CTensor4D<T> tempTensor(aTensor.xSize(),aTensor.ySize(),aTensor.zSize(),aTensor.aSize());
+ filter(aTensor,tempTensor,1,1,1,aFilter);
+ aTensor = tempTensor;
+}
+
+template <class T>
+void filter(const CTensor4D<T>& aTensor, CTensor4D<T>& aResult, const int aDummy1, const int aDummy2, const int aDummy3, const CFilter<T>& aFilter) {
+ if (aResult.xSize() != aTensor.xSize() || aResult.ySize() != aTensor.ySize() || aResult.zSize() != aTensor.zSize() || aResult.aSize() != aTensor.aSize())
+ throw EFilterIncompatibleSize(aTensor.xSize()*aTensor.ySize()*aTensor.zSize()*aTensor.aSize(),aResult.xSize()*aResult.ySize()*aResult.zSize()*aResult.aSize());
+ int a1 = -aFilter.A();
+ int a2 = aTensor.aSize()-aFilter.B();
+ int a2Size = 2*aTensor.aSize()-1;
+ aResult = 0;
+ // Front rim
+ for (int a = 0; a < a1; a++)
+ for (int z = 0; z < aTensor.zSize(); z++)
+ for (int y = 0; y < aTensor.ySize(); y++)
+ for (int x = 0; x < aTensor.xSize(); x++)
+ for (int i = aFilter.A(); i < aFilter.B(); i++) {
+ if (a+i < 0) aResult(x,y,z,a) += aFilter[i]*aTensor(x,y,z,-1-a-i);
+ else if (a+i >= aTensor.aSize()) aResult(x,y,z,a) += aFilter[i]*aTensor(x,y,z,a2Size-a-i);
+ else aResult(x,y,z,a) += aFilter[i]*aTensor(x,y,z,a+i);
+ }
+ // Back rim
+ for (int a = a2; a < aTensor.aSize(); a++)
+ for (int z = 0; z < aTensor.zSize(); z++)
+ for (int y = 0; y < aTensor.ySize(); y++)
+ for (int x = 0; x < aTensor.xSize(); x++)
+ for (int i = aFilter.A(); i < aFilter.B(); i++) {
+ if (a+i < 0) aResult(x,y,z,a) += aFilter[i]*aTensor(x,y,z,-1-a-i);
+ else if (a+i >= aTensor.aSize()) aResult(x,y,z,a) += aFilter[i]*aTensor(x,y,z,a2Size-a-i);
+ else aResult(x,y,z,a) += aFilter[i]*aTensor(x,y,z,a+i);
+ }
+ // Center
+ for (int a = a1; a < a2; a++)
+ for (int z = 0; z < aTensor.zSize(); z++)
+ for (int y = 0; y < aTensor.ySize(); y++)
+ for (int x = 0; x < aTensor.xSize(); x++)
+ for (int i = aFilter.A(); i < aFilter.B(); i++)
+ aResult(x,y,z,a) += aFilter[i]*aTensor(x,y,z,a+i);
+}
+
+template <class T>
+void recursiveSmoothX(CTensor4D<T>& aTensor, float aSigma) {
+ CVector<T> aVals1(aTensor.xSize());
+ CVector<T> aVals2(aTensor.xSize());
+ float aAlpha = 2.5/(sqrt(NMath::Pi)*aSigma);
+ float aExp = exp(-aAlpha);
+ float aExpSqr = aExp*aExp;
+ float a2Exp = 2.0*aExp;
+ float k = (1.0-aExp)*(1.0-aExp)/(1.0+2.0*aAlpha*aExp-aExpSqr);
+ float aPreMinus = aExp*(aAlpha-1.0);
+ float aPrePlus = aExp*(aAlpha+1.0);
+ for (int a = 0; a < aTensor.aSize(); a++)
+ for (int z = 0; z < aTensor.zSize(); z++)
+ for (int y = 0; y < aTensor.ySize(); y++) {
+ int aOffset = ((a*aTensor.zSize()+z)*aTensor.ySize()+y)*aTensor.xSize();
+ aVals1(0) = (0.5-k*aPreMinus)*aTensor.data()[aOffset];
+ aVals1(1) = k*(aTensor.data()[aOffset+1]+aPreMinus*aTensor.data()[aOffset])+(2.0*aExp-aExpSqr)*aVals1(0);
+ for (int x = 2; x < aTensor.xSize(); x++)
+ aVals1(x) = k*(aTensor.data()[aOffset+x]+aPreMinus*aTensor.data()[aOffset+x-1])+a2Exp*aVals1(x-1)-aExpSqr*aVals1(x-2);
+ aVals2(aTensor.xSize()-1) = (0.5+k*aPreMinus)*aTensor.data()[aOffset+aTensor.xSize()-1];
+ aVals2(aTensor.xSize()-2) = k*((aPrePlus-aExpSqr)*aTensor.data()[aOffset+aTensor.xSize()-1])+(a2Exp-aExpSqr)*aVals2(aTensor.xSize()-1);
+ for (int x = aTensor.xSize()-3; x >= 0; x--)
+ aVals2(x) = k*(aPrePlus*aTensor.data()[aOffset+x+1]-aExpSqr*aTensor.data()[aOffset+x+2])+a2Exp*aVals2(x+1)-aExpSqr*aVals2(x+2);
+ for (int x = 0; x < aTensor.xSize(); x++)
+ aTensor.data()[aOffset+x] = aVals1(x)+aVals2(x);
+ }
+}
+
+template <class T>
+void recursiveSmoothY(CTensor4D<T>& aTensor, float aSigma) {
+ CVector<T> aVals1(aTensor.ySize());
+ CVector<T> aVals2(aTensor.ySize());
+ CVector<T> aVals3(aTensor.ySize());
+ float aAlpha = 2.5/(sqrt(NMath::Pi)*aSigma);
+ float aExp = exp(-aAlpha);
+ float aExpSqr = aExp*aExp;
+ float a2Exp = 2.0*aExp;
+ float k = (1.0-aExp)*(1.0-aExp)/(1.0+2.0*aAlpha*aExp-aExpSqr);
+ float aPreMinus = aExp*(aAlpha-1.0);
+ float aPrePlus = aExp*(aAlpha+1.0);
+ for (int a = 0; a < aTensor.aSize(); a++)
+ for (int z = 0; z < aTensor.zSize(); z++)
+ for (int x = 0; x < aTensor.xSize(); x++) {
+ for (int y = 0; y < aTensor.ySize(); y++)
+ aVals3(y) = aTensor(x,y,z,a);
+ aVals1(0) = (0.5-k*aPreMinus)*aVals3(0);
+ aVals1(1) = k*(aVals3(1)+aPreMinus*aVals3(0))+(2.0*aExp-aExpSqr)*aVals1(0);
+ for (int y = 2; y < aTensor.ySize(); y++)
+ aVals1(y) = k*(aVals3(y)+aPreMinus*aVals3(y-1))+a2Exp*aVals1(y-1)-aExpSqr*aVals1(y-2);
+ aVals2(aTensor.ySize()-1) = (0.5+k*aPreMinus)*aVals3(aTensor.ySize()-1);
+ aVals2(aTensor.ySize()-2) = k*((aPrePlus-aExpSqr)*aVals3(aTensor.ySize()-1))+(a2Exp-aExpSqr)*aVals2(aTensor.ySize()-1);
+ for (int y = aTensor.ySize()-3; y >= 0; y--)
+ aVals2(y) = k*(aPrePlus*aVals3(y+1)-aExpSqr*aVals3(y+2))+a2Exp*aVals2(y+1)-aExpSqr*aVals2(y+2);
+ for (int y = 0; y < aTensor.ySize(); y++)
+ aTensor(x,y,z,a) = aVals1(y)+aVals2(y);
+ }
+}
+
+template <class T>
+void recursiveSmoothZ(CTensor4D<T>& aTensor, float aSigma) {
+ CVector<T> aVals1(aTensor.zSize());
+ CVector<T> aVals2(aTensor.zSize());
+ CVector<T> aVals3(aTensor.zSize());
+ float aAlpha = 2.5/(sqrt(NMath::Pi)*aSigma);
+ float aExp = exp(-aAlpha);
+ float aExpSqr = aExp*aExp;
+ float a2Exp = 2.0*aExp;
+ float k = (1.0-aExp)*(1.0-aExp)/(1.0+2.0*aAlpha*aExp-aExpSqr);
+ float aPreMinus = aExp*(aAlpha-1.0);
+ float aPrePlus = aExp*(aAlpha+1.0);
+ for (int a = 0; a < aTensor.aSize(); a++)
+ for (int y = 0; y < aTensor.ySize(); y++)
+ for (int x = 0; x < aTensor.xSize(); x++) {
+ for (int z = 0; z < aTensor.zSize(); z++)
+ aVals3(z) = aTensor(x,y,z,a);
+ aVals1(0) = (0.5-k*aPreMinus)*aVals3(0);
+ aVals1(1) = k*(aVals3(1)+aPreMinus*aVals3(0))+(2.0*aExp-aExpSqr)*aVals1(0);
+ for (int z = 2; z < aTensor.zSize(); z++)
+ aVals1(z) = k*(aVals3(z)+aPreMinus*aVals3(z-1))+a2Exp*aVals1(z-1)-aExpSqr*aVals1(z-2);
+ aVals2(aTensor.zSize()-1) = (0.5+k*aPreMinus)*aVals3(aTensor.zSize()-1);
+ aVals2(aTensor.zSize()-2) = k*((aPrePlus-aExpSqr)*aVals3(aTensor.zSize()-1))+(a2Exp-aExpSqr)*aVals2(aTensor.zSize()-1);
+ for (int z = aTensor.zSize()-3; z >= 0; z--)
+ aVals2(z) = k*(aPrePlus*aVals3(z+1)-aExpSqr*aVals3(z+2))+a2Exp*aVals2(z+1)-aExpSqr*aVals2(z+2);
+ for (int z = 0; z < aTensor.zSize(); z++)
+ aTensor(x,y,z,a) = aVals1(z)+aVals2(z);
+ }
+}
+
+template <class T>
+void recursiveSmoothA(CTensor4D<T>& aTensor, float aSigma) {
+ CVector<T> aVals1(aTensor.aSize());
+ CVector<T> aVals2(aTensor.aSize());
+ CVector<T> aVals3(aTensor.aSize());
+ float aAlpha = 2.5/(sqrt(NMath::Pi)*aSigma);
+ float aExp = exp(-aAlpha);
+ float aExpSqr = aExp*aExp;
+ float a2Exp = 2.0*aExp;
+ float k = (1.0-aExp)*(1.0-aExp)/(1.0+2.0*aAlpha*aExp-aExpSqr);
+ float aPreMinus = aExp*(aAlpha-1.0);
+ float aPrePlus = aExp*(aAlpha+1.0);
+ for (int z = 0; z < aTensor.zSize(); z++)
+ for (int y = 0; y < aTensor.ySize(); y++)
+ for (int x = 0; x < aTensor.xSize(); x++) {
+ for (int a = 0; a < aTensor.aSize(); a++)
+ aVals3(a) = aTensor(x,y,z,a);
+ aVals1(0) = (0.5-k*aPreMinus)*aVals3(0);
+ aVals1(1) = k*(aVals3(1)+aPreMinus*aVals3(0))+(2.0*aExp-aExpSqr)*aVals1(0);
+ for (int a = 2; a < aTensor.aSize(); a++)
+ aVals1(a) = k*(aVals3(a)+aPreMinus*aVals3(a-1))+a2Exp*aVals1(a-1)-aExpSqr*aVals1(a-2);
+ aVals2(aTensor.aSize()-1) = (0.5+k*aPreMinus)*aVals3(aTensor.aSize()-1);
+ aVals2(aTensor.aSize()-2) = k*((aPrePlus-aExpSqr)*aVals3(aTensor.aSize()-1))+(a2Exp-aExpSqr)*aVals2(aTensor.aSize()-1);
+ for (int a = aTensor.aSize()-3; a >= 0; a--)
+ aVals2(a) = k*(aPrePlus*aVals3(a+1)-aExpSqr*aVals3(a+2))+a2Exp*aVals2(a+1)-aExpSqr*aVals2(a+2);
+ for (int a = 0; a < aTensor.aSize(); a++)
+ aTensor(x,y,z,a) = aVals1(a)+aVals2(a);
+ }
+}
+
+// Nonlinear filters -----------------------------------------------------------
+
+// osher (2D)
+template <class T>
+void osher(CMatrix<T>& aData, int aIterations) {
+ CMatrix<T> aDiff(aData.xSize(),aData.ySize());
+ for (int t = 0; t < aIterations; t++) {
+ for (int y = 0; y < aData.ySize(); y++)
+ for (int x = 0; x < aData.xSize(); x++) {
+ T u00,u01,u02,u10,u11,u12,u20,u21,u22;
+ if (x > 0) {
+ if (y > 0) u00 = aData(x-1,y-1);
+ else u00 = aData(x-1,0);
+ u01 = aData(x-1,y);
+ if (y < aData.ySize()-1) u02 = aData(x-1,y+1);
+ else u02 = aData(x-1,y);
+ }
+ else {
+ if (y > 0) u00 = aData(0,y-1);
+ else u00 = aData(0,0);
+ u01 = aData(0,y);
+ if (y < aData.ySize()-1) u02 = aData(0,y+1);
+ else u02 = aData(0,y);
+ }
+ if (y > 0) u10 = aData(x,y-1);
+ else u10 = aData(x,y);
+ u11 = aData(x,y);
+ if (y < aData.ySize()-1) u12 = aData(x,y+1);
+ else u12 = aData(x,y);
+ if (x < aData.xSize()-1) {
+ if (y > 0) u20 = aData(x+1,y-1);
+ else u20 = aData(x+1,y);
+ u21 = aData(x+1,y);
+ if (y < aData.ySize()-1) u22 = aData(x+1,y+1);
+ else u22 = aData(x+1,y);
+ }
+ else {
+ if (y > 0) u20 = aData(x,y-1);
+ else u20 = aData(x,y);
+ u21 = aData(x,y);
+ if (y < aData.ySize()-1) u22 = aData(x,y+1);
+ else u22 = aData(x,y);
+ }
+ T ux = 0.5*(u21-u01);
+ T uy = 0.5*(u12-u10);
+ T uxuy = ux*uy;
+ T uxx = u01-2.0*u11+u21;
+ T uyy = u10-2.0*u11+u12;
+ T uxy;
+ if (uxuy < 0) uxy = 2.0*u11+u00+u22-u10-u12-u01-u21;
+ else uxy = u10+u12+u01+u21-2.0*u11-u02-u20;
+ T laPlace = uyy*uy*uy+uxy*uxuy+uxx*ux*ux;
+ T uxLeft = u11-u01;
+ T uxRight = u21-u11;
+ T uyUp = u11-u10;
+ T uyDown = u12-u11;
+ if (laPlace < 0) {
+ T aSum = 0;
+ if (uxRight > 0) aSum += uxRight*uxRight;
+ if (uxLeft < 0) aSum += uxLeft*uxLeft;
+ if (uyDown > 0) aSum += uyDown*uyDown;
+ if (uyUp < 0) aSum += uyUp*uyUp;
+ aDiff(x,y) = sqrt(aSum);
+ }
+ else if (laPlace > 0) {
+ T aSum = 0;
+ if (uxRight < 0) aSum += uxRight*uxRight;
+ if (uxLeft > 0) aSum += uxLeft*uxLeft;
+ if (uyDown < 0) aSum += uyDown*uyDown;
+ if (uyUp > 0) aSum += uyUp*uyUp;
+ aDiff(x,y) = -sqrt(aSum);
+ }
+ }
+ for (int i = 0; i < aData.size(); i++)
+ aData.data()[i] += 0.25*aDiff.data()[i];
+ }
+}
+
+template <class T>
+inline void osher(const CMatrix<T>& aData, CMatrix<T>& aResult, int aIterations) {
+ aResult = aData;
+ osher(aResult,aIterations);
+}
+
+}
+
+#endif
+
diff --git a/video_input/consistencyChecker/CMatrix.h b/video_input/consistencyChecker/CMatrix.h new file mode 100644 index 0000000..a49553e --- /dev/null +++ b/video_input/consistencyChecker/CMatrix.h @@ -0,0 +1,1396 @@ +// CMatrix
+// A two-dimensional array including basic matrix operations
+//
+// Author: Thomas Brox
+//-------------------------------------------------------------------------
+
+#ifndef CMATRIX_H
+#define CMATRIX_H
+
+#include <math.h>
+#include <stdio.h>
+#include <string.h>
+#include <stdlib.h>
+#include <iostream>
+#include <fstream>
+#include <string>
+#include <queue>
+#include <stack>
+#ifdef GNU_COMPILER
+ #include <strstream>
+#else
+ #include <sstream>
+#endif
+#include <CVector.h>
+
+template <class T>
+class CMatrix {
+public:
+ // standard constructor
+ inline CMatrix();
+ // constructor
+ inline CMatrix(const int aXSize, const int aYSize);
+ // copy constructor
+ CMatrix(const CMatrix<T>& aCopyFrom);
+ // constructor with implicit filling
+ CMatrix(const int aXSize, const int aYSize, const T aFillValue);
+ // destructor
+ virtual ~CMatrix();
+
+ // Changes the size of the matrix, data will be lost
+ void setSize(int aXSize, int aYSize);
+ // Downsamples the matrix
+ void downsampleBool(int aNewXSize, int aNewYSize, float aThreshold = 0.5);
+ void downsampleInt(int aNewXSize, int aNewYSize);
+ void downsample(int aNewXSize, int aNewYSize);
+ void downsample(int aNewXSize, int aNewYSize, CMatrix<float>& aConfidence);
+ void downsampleBilinear(int aNewXSize, int aNewYSize);
+ // Upsamples the matrix
+ void upsample(int aNewXSize, int aNewYSize);
+ void upsampleBilinear(int aNewXSize, int aNewYSize);
+// void upsampleBicubic(int aNewXSize, int aNewYSize);
+ // Scales the matrix (includes upsampling and downsampling)
+ void rescale(int aNewXSize, int aNewYSize);
+ // Creates an identity matrix
+ void identity(int aSize);
+ // Fills the matrix with the value aValue (see also operator =)
+ void fill(const T aValue);
+ // Fills a rectangular area with the value aValue
+ void fillRect(const T aValue, int ax1, int ay1, int ax2, int ay2);
+ // Copies a rectangular part from the matrix into aResult, the size of aResult will be adjusted
+ void cut(CMatrix<T>& aResult,const int x1, const int y1, const int x2, const int y2);
+ // Copies aCopyFrom at a certain position of the matrix
+ void paste(CMatrix<T>& aCopyFrom, int ax, int ay);
+ // Mirrors the boundaries, aFrom is the distance from the boundaries where the pixels are copied from,
+ // aTo is the distance from the boundaries they are copied to
+ void mirror(int aFrom, int aTo);
+ // Transforms the values so that they are all between aMin and aMax
+ // aInitialMin/Max are initializations for seeking the minimum and maximum, change if your
+ // data is not in this range or the data type T cannot hold these values
+ void normalize(T aMin, T aMax, T aInitialMin = -30000, T aInitialMax = 30000);
+ // Clips values that exceed the given range
+ void clip(T aMin, T aMax);
+
+ // Applies a similarity transform (translation, rotation, scaling) to the image
+ void applySimilarityTransform(CMatrix<T>& aWarped, CMatrix<bool>& aOutside, float tx, float ty, float cx, float cy, float phi, float scale);
+ // Applies a homography (linear projective transformation) to the image
+ void applyHomography(CMatrix<T>& aWarped, CMatrix<bool>& aOutside, const CMatrix<float>& H);
+
+ // Draws a line into the image
+ void drawLine(int dStartX, int dStartY, int dEndX, int dEndY, T aValue = 255);
+ // Inverts a gray value image
+ void invertImage();
+ // Extracts the connected component starting from (x,y)
+ // Component -> 255, Remaining area -> 0
+ void connectedComponent(int x, int y);
+
+ // Appends another matrix with the same column number
+ void append(CMatrix<T>& aMatrix);
+ // Inverts a square matrix with Gauss elimination
+ void inv();
+ // Transposes a square matrix
+ void trans();
+ // Multiplies with two vectors (from left and from right)
+ float scalar(CVector<T>& aLeft, CVector<T>& aRight);
+
+ // Reads a picture from a pgm-File
+ void readFromPGM(const char* aFilename);
+ // Saves the matrix as a picture in pgm-Format
+ void writeToPGM(const char *aFilename);
+ // Read matrix from text file
+ void readFromTXT(const char* aFilename, bool aHeader = true, int aXSize = 0, int aYSize = 0);
+ // Read matrix from Matlab ascii file
+ void readFromMatlabTXT(const char* aFilename, bool aHeader = true, int aXSize = 0, int aYSize = 0);
+ // Save matrix as text file
+ void writeToTXT(const char* aFilename, bool aHeader = true);
+ // Reads a projection matrix in a format used by Bodo Rosenhahn
+ void readBodoProjectionMatrix(const char* aFilename);
+
+ // Gives full access to matrix values
+ inline T& operator()(const int ax, const int ay) const;
+ // Fills the matrix with the value aValue (equivalent to fill())
+ inline CMatrix<T>& operator=(const T aValue);
+ // Copies the matrix aCopyFrom to this matrix (size of matrix might change)
+ CMatrix<T>& operator=(const CMatrix<T>& aCopyFrom);
+ // matrix sum
+ CMatrix<T>& operator+=(const CMatrix<T>& aMatrix);
+ // Adds a constant to the matrix
+ CMatrix<T>& operator+=(const T aValue);
+ // matrix difference
+ CMatrix<T>& operator-=(const CMatrix<T>& aMatrix);
+ // matrix product
+ CMatrix<T>& operator*=(const CMatrix<T>& aMatrix);
+ // Multiplication with a scalar
+ CMatrix<T>& operator*=(const T aValue);
+
+ // Comparison of two matrices
+ bool operator==(const CMatrix<T>& aMatrix);
+
+ // Returns the minimum value
+ T min() const;
+ // Returns the maximum value
+ T max() const;
+ // Returns the average value
+ T avg() const;
+ // Gives access to the matrix' size
+ inline int xSize() const;
+ inline int ySize() const;
+ inline int size() const;
+ // Returns one row from the matrix
+ void getVector(CVector<T>& aVector, int ay);
+ // Gives access to the internal data representation
+ inline T* data() const;
+protected:
+ int mXSize,mYSize;
+ T *mData;
+};
+
+// Returns a matrix where all negative elements are turned positive
+template <class T> CMatrix<T> abs(const CMatrix<T>& aMatrix);
+// Returns the tranposed matrix
+template <class T> CMatrix<T> trans(const CMatrix<T>& aMatrix);
+// matrix sum
+template <class T> CMatrix<T> operator+(const CMatrix<T>& aM1, const CMatrix<T>& aM2);
+// matrix difference
+template <class T> CMatrix<T> operator-(const CMatrix<T>& aM1, const CMatrix<T>& aM2);
+// matrix product
+template <class T> CMatrix<T> operator*(const CMatrix<T>& aM1, const CMatrix<T>& aM2);
+// Multiplication with a vector
+template <class T> CVector<T> operator*(const CMatrix<T>& aMatrix, const CVector<T>& aVector);
+// Multiplikation with a scalar
+template <class T> CMatrix<T> operator*(const CMatrix<T>& aMatrix, const T aValue);
+template <class T> inline CMatrix<T> operator*(const T aValue, const CMatrix<T>& aMatrix);
+// Provides basic output functionality (only appropriate for small matrices)
+template <class T> std::ostream& operator<<(std::ostream& aStream, const CMatrix<T>& aMatrix);
+
+// Exceptions thrown by CMatrix-------------------------------------------
+
+
+// Thrown when one tries to access an element of a matrix which is out of
+// the matrix' bounds
+struct EMatrixRangeOverflow {
+ EMatrixRangeOverflow(const int ax, const int ay) {
+ using namespace std;
+ cerr << "Exception EMatrixRangeOverflow: x = " << ax << ", y = " << ay << endl;
+ }
+};
+
+// Thrown when one tries to multiply two matrices where M1's column number
+// is not equal to M2's row number or when one tries to add two matrices
+// which have not the same size
+struct EIncompatibleMatrices {
+ EIncompatibleMatrices(const int x1, const int y1, const int x2, const int y2) {
+
+ using namespace std;
+ cerr << "Exception EIncompatibleMatrices: M1 = " << x1 << "x" << y1;
+ cerr << " M2 = " << x2 << "x" << y2 << endl;
+ }
+};
+
+// Thrown when a nonquadratic matrix is tried to be inversed
+struct ENonquadraticMatrix {
+ ENonquadraticMatrix(const int x, const int y) {
+ using namespace std;
+ cerr << "Exception ENonquadarticMatrix: M = " << x << "x" << y << endl;
+ }
+};
+
+// Thrown when a matrix is not positive definite
+struct ENonPositiveDefinite {
+ ENonPositiveDefinite() {
+ using namespace std;
+ cerr << "Exception ENonPositiveDefinite" << endl;
+ }
+};
+
+// Thrown when reading a file which does not keep to the PGM specification
+struct EInvalidFileFormat {
+ EInvalidFileFormat(const char* s) {
+ using namespace std;
+ cerr << "Exception EInvalidFileFormat: File is not in " << s << " format" << endl;
+ }
+};
+
+// I M P L E M E N T A T I O N --------------------------------------------
+//
+// You might wonder why there is implementation code in a header file.
+// The reason is that not all C++ compilers yet manage separate compilation
+// of templates. Inline functions cannot be compiled separately anyway.
+// So in this case the whole implementation code is added to the header
+// file.
+// Users of CMatrix should ignore everything that's beyond this line :)
+// ------------------------------------------------------------------------
+
+// P U B L I C ------------------------------------------------------------
+
+// standard constructor
+template <class T>
+inline CMatrix<T>::CMatrix() {
+ mData = 0; mXSize = mYSize = 0;
+}
+
+// constructor
+template <class T>
+inline CMatrix<T>::CMatrix(const int aXSize, const int aYSize)
+ : mXSize(aXSize), mYSize(aYSize) {
+ mData = new T[aXSize*aYSize];
+}
+
+// copy constructor
+template <class T>
+CMatrix<T>::CMatrix(const CMatrix<T>& aCopyFrom)
+ : mXSize(aCopyFrom.mXSize), mYSize(aCopyFrom.mYSize) {
+ if (aCopyFrom.mData == 0) mData = 0;
+ else {
+ int wholeSize = mXSize*mYSize;
+ mData = new T[wholeSize];
+ for (register int i = 0; i < wholeSize; i++)
+ mData[i] = aCopyFrom.mData[i];
+ }
+}
+
+// constructor with implicit filling
+template <class T>
+CMatrix<T>::CMatrix(const int aXSize, const int aYSize, const T aFillValue)
+ : mXSize(aXSize), mYSize(aYSize) {
+ mData = new T[aXSize*aYSize];
+ fill(aFillValue);
+}
+
+// destructor
+template <class T>
+CMatrix<T>::~CMatrix() {
+ delete [] mData;
+}
+
+// setSize
+template <class T>
+void CMatrix<T>::setSize(int aXSize, int aYSize) {
+ if (mData != 0) delete[] mData;
+ mData = new T[aXSize*aYSize];
+ mXSize = aXSize;
+ mYSize = aYSize;
+}
+
+// downsampleBool
+template <class T>
+void CMatrix<T>::downsampleBool(int aNewXSize, int aNewYSize, float aThreshold) {
+ CMatrix<float> aTemp(mXSize,mYSize);
+ int aSize = size();
+ for (int i = 0; i < aSize; i++)
+ aTemp.data()[i] = mData[i];
+ aTemp.downsample(aNewXSize,aNewYSize);
+ setSize(aNewXSize,aNewYSize);
+ aSize = size();
+ for (int i = 0; i < aSize; i++)
+ mData[i] = (aTemp.data()[i] >= aThreshold);
+}
+
+// downsampleInt
+template <class T>
+void CMatrix<T>::downsampleInt(int aNewXSize, int aNewYSize) {
+ T* newData = new int[aNewXSize*aNewYSize];
+ float factorX = ((float)mXSize)/aNewXSize;
+ float factorY = ((float)mYSize)/aNewYSize;
+ float ay = 0.0;
+ for (int y = 0; y < aNewYSize; y++) {
+ float ax = 0.0;
+ for (int x = 0; x < aNewXSize; x++) {
+ CVector<float> aHistogram(256,0.0);
+ for (float by = 0.0; by < factorY;) {
+ float restY = floor(by+1.0)-by;
+ if (restY+by >= factorY) restY = factorY-by;
+ for (float bx = 0.0; bx < factorX;) {
+ float restX = floor(bx+1.0)-bx;
+ if (restX+bx >= factorX) restX = factorX-bx;
+ aHistogram(operator()((int)(ax+bx),(int)(ay+by))) += restX*restY;
+ bx += restX;
+ }
+ by += restY;
+ }
+ float aMax = 0; int aMaxVal;
+ for (int i = 0; i < aHistogram.size(); i++)
+ if (aHistogram(i) > aMax) {
+ aMax = aHistogram(i);
+ aMaxVal = i;
+ }
+ newData[x+aNewXSize*y] = aMaxVal;
+ ax += factorX;
+ }
+ ay += factorY;
+ }
+ delete[] mData;
+ mData = newData;
+ mXSize = aNewXSize; mYSize = aNewYSize;
+}
+
+template <class T>
+void CMatrix<T>::downsample(int aNewXSize, int aNewYSize) {
+ // Downsample in x-direction
+ int aIntermedSize = aNewXSize*mYSize;
+ T* aIntermedData = new T[aIntermedSize];
+ if (aNewXSize < mXSize) {
+ for (int i = 0; i < aIntermedSize; i++)
+ aIntermedData[i] = 0.0;
+ T factor = ((float)mXSize)/aNewXSize;
+ for (int y = 0; y < mYSize; y++) {
+ int aFineOffset = y*mXSize;
+ int aCoarseOffset = y*aNewXSize;
+ int i = aFineOffset;
+ int j = aCoarseOffset;
+ int aLastI = aFineOffset+mXSize;
+ int aLastJ = aCoarseOffset+aNewXSize;
+ T rest = factor;
+ T part = 1.0;
+ do {
+ if (rest > 1.0) {
+ aIntermedData[j] += part*mData[i];
+ rest -= part;
+ part = 1.0;
+ i++;
+ if (rest <= 0.0) {
+ rest = factor;
+ j++;
+ }
+ }
+ else {
+ aIntermedData[j] += rest*mData[i];
+ part = 1.0-rest;
+ rest = factor;
+ j++;
+ }
+ }
+ while (i < aLastI && j < aLastJ);
+ }
+ }
+ else {
+ T* aTemp = aIntermedData;
+ aIntermedData = mData;
+ mData = aTemp;
+ }
+ // Downsample in y-direction
+ delete[] mData;
+ int aDataSize = aNewXSize*aNewYSize;
+ mData = new T[aDataSize];
+ if (aNewYSize < mYSize) {
+ for (int i = 0; i < aDataSize; i++)
+ mData[i] = 0.0;
+ float factor = ((float)mYSize)/aNewYSize;
+ for (int x = 0; x < aNewXSize; x++) {
+ int i = x;
+ int j = x;
+ int aLastI = mYSize*aNewXSize+x;
+ int aLastJ = aNewYSize*aNewXSize+x;
+ float rest = factor;
+ float part = 1.0;
+ do {
+ if (rest > 1.0) {
+ mData[j] += part*aIntermedData[i];
+ rest -= part;
+ part = 1.0;
+ i += aNewXSize;
+ if (rest <= 0.0) {
+ rest = factor;
+ j += aNewXSize;
+ }
+ }
+ else {
+ mData[j] += rest*aIntermedData[i];
+ part = 1.0-rest;
+ rest = factor;
+ j += aNewXSize;
+ }
+ }
+ while (i < aLastI && j < aLastJ);
+ }
+ }
+ else {
+ T* aTemp = mData;
+ mData = aIntermedData;
+ aIntermedData = aTemp;
+ }
+ // Normalize
+ float aNormalization = ((float)aDataSize)/size();
+ for (int i = 0; i < aDataSize; i++)
+ mData[i] *= aNormalization;
+ // Adapt size of matrix
+ mXSize = aNewXSize;
+ mYSize = aNewYSize;
+ delete[] aIntermedData;
+}
+
+template <class T>
+void CMatrix<T>::downsample(int aNewXSize, int aNewYSize, CMatrix<float>& aConfidence) {
+ int aNewSize = aNewXSize*aNewYSize;
+ T* newData = new T[aNewSize];
+ float* aCounter = new float[aNewSize];
+ for (int i = 0; i < aNewSize; i++) {
+ newData[i] = 0;
+ aCounter[i] = 0;
+ }
+ float factorX = ((float)aNewXSize)/mXSize;
+ float factorY = ((float)aNewYSize)/mYSize;
+ for (int y = 0; y < mYSize; y++)
+ for (int x = 0; x < mXSize; x++)
+ if (aConfidence(x,y) > 0) {
+ float ax = x*factorX;
+ float ay = y*factorY;
+ int x1 = (int)ax;
+ int y1 = (int)ay;
+ int x2 = x1+1;
+ int y2 = y1+1;
+ float alphax = ax-x1;
+ float betax = 1.0-alphax;
+ float alphay = ay-y1;
+ float betay = 1.0-alphay;
+ float conf = aConfidence(x,y);
+ T val = conf*operator()(x,y);
+ int i = x1+aNewXSize*y1;
+ newData[i] += betax*betay*val;
+ aCounter[i] += betax*betay*conf;
+ if (x2 < aNewXSize) {
+ i = x2+aNewXSize*y1;
+ newData[i] += alphax*betay*val;
+ aCounter[i] += alphax*betay*conf;
+ }
+ if (y2 < aNewYSize) {
+ i = x1+aNewXSize*y2;
+ newData[i] += betax*alphay*val;
+ aCounter[i] += betax*alphay*conf;
+ }
+ if (x2 < aNewXSize && y2 < aNewYSize) {
+ i = x2+aNewXSize*y2;
+ newData[i] += alphax*alphay*val;
+ aCounter[i] += alphax*alphay*conf;
+ }
+ }
+ for (int i = 0; i < aNewSize; i++)
+ if (aCounter[i] > 0) newData[i] /= aCounter[i];
+ // Adapt size of matrix
+ mXSize = aNewXSize;
+ mYSize = aNewYSize;
+ delete[] mData;
+ delete[] aCounter;
+ mData = newData;
+}
+
+// downsampleBilinear
+template <class T>
+void CMatrix<T>::downsampleBilinear(int aNewXSize, int aNewYSize) {
+ int aNewSize = aNewXSize*aNewYSize;
+ T* aNewData = new T[aNewSize];
+ float factorX = ((float)mXSize)/aNewXSize;
+ float factorY = ((float)mYSize)/aNewYSize;
+ for (int y = 0; y < aNewYSize; y++)
+ for (int x = 0; x < aNewXSize; x++) {
+ float ax = (x+0.5)*factorX-0.5;
+ float ay = (y+0.5)*factorY-0.5;
+ if (ax < 0) ax = 0.0;
+ if (ay < 0) ay = 0.0;
+ int x1 = (int)ax;
+ int y1 = (int)ay;
+ int x2 = x1+1;
+ int y2 = y1+1;
+ float alphaX = ax-x1;
+ float alphaY = ay-y1;
+ if (x1 < 0) x1 = 0;
+ if (y1 < 0) y1 = 0;
+ if (x2 >= mXSize) x2 = mXSize-1;
+ if (y2 >= mYSize) y2 = mYSize-1;
+ float a = (1.0-alphaX)*mData[x1+y1*mXSize]+alphaX*mData[x2+y1*mXSize];
+ float b = (1.0-alphaX)*mData[x1+y2*mXSize]+alphaX*mData[x2+y2*mXSize];
+ aNewData[x+y*aNewXSize] = (1.0-alphaY)*a+alphaY*b;
+ }
+ delete[] mData;
+ mData = aNewData;
+ mXSize = aNewXSize;
+ mYSize = aNewYSize;
+}
+
+template <class T>
+void CMatrix<T>::upsample(int aNewXSize, int aNewYSize) {
+ // Upsample in x-direction
+ int aIntermedSize = aNewXSize*mYSize;
+ T* aIntermedData = new T[aIntermedSize];
+ if (aNewXSize > mXSize) {
+ for (int i = 0; i < aIntermedSize; i++)
+ aIntermedData[i] = 0.0;
+ T factor = ((float)aNewXSize)/mXSize;
+ for (int y = 0; y < mYSize; y++) {
+ int aFineOffset = y*aNewXSize;
+ int aCoarseOffset = y*mXSize;
+ int i = aCoarseOffset;
+ int j = aFineOffset;
+ int aLastI = aCoarseOffset+mXSize;
+ int aLastJ = aFineOffset+aNewXSize;
+ T rest = factor;
+ T part = 1.0;
+ do {
+ if (rest > 1.0) {
+ aIntermedData[j] += part*mData[i];
+ rest -= part;
+ part = 1.0;
+ j++;
+ if (rest <= 0.0) {
+ rest = factor;
+ i++;
+ }
+ }
+ else {
+ aIntermedData[j] += rest*mData[i];
+ part = 1.0-rest;
+ rest = factor;
+ i++;
+ }
+ }
+ while (i < aLastI && j < aLastJ);
+ }
+ }
+ else {
+ T* aTemp = aIntermedData;
+ aIntermedData = mData;
+ mData = aTemp;
+ }
+ // Upsample in y-direction
+ delete[] mData;
+ int aDataSize = aNewXSize*aNewYSize;
+ mData = new T[aDataSize];
+ if (aNewYSize > mYSize) {
+ for (int i = 0; i < aDataSize; i++)
+ mData[i] = 0.0;
+ float factor = ((float)aNewYSize)/mYSize;
+ for (int x = 0; x < aNewXSize; x++) {
+ int i = x;
+ int j = x;
+ int aLastI = mYSize*aNewXSize;
+ int aLastJ = aNewYSize*aNewXSize;
+ float rest = factor;
+ float part = 1.0;
+ do {
+ if (rest > 1.0) {
+ mData[j] += part*aIntermedData[i];
+ rest -= part;
+ part = 1.0;
+ j += aNewXSize;
+ if (rest <= 0.0) {
+ rest = factor;
+ i += aNewXSize;
+ }
+ }
+ else {
+ mData[j] += rest*aIntermedData[i];
+ part = 1.0-rest;
+ rest = factor;
+ i += aNewXSize;
+ }
+ }
+ while (i < aLastI && j < aLastJ);
+ }
+ }
+ else {
+ T* aTemp = mData;
+ mData = aIntermedData;
+ aIntermedData = aTemp;
+ }
+ // Adapt size of matrix
+ mXSize = aNewXSize;
+ mYSize = aNewYSize;
+ delete[] aIntermedData;
+}
+
+// upsampleBilinear
+template <class T>
+void CMatrix<T>::upsampleBilinear(int aNewXSize, int aNewYSize) {
+ int aNewSize = aNewXSize*aNewYSize;
+ T* aNewData = new T[aNewSize];
+ float factorX = (float)(mXSize)/(aNewXSize);
+ float factorY = (float)(mYSize)/(aNewYSize);
+ for (int y = 0; y < aNewYSize; y++)
+ for (int x = 0; x < aNewXSize; x++) {
+ float ax = (x+0.5)*factorX-0.5;
+ float ay = (y+0.5)*factorY-0.5;
+ if (ax < 0) ax = 0.0;
+ if (ay < 0) ay = 0.0;
+ int x1 = (int)ax;
+ int y1 = (int)ay;
+ int x2 = x1+1;
+ int y2 = y1+1;
+ float alphaX = ax-x1;
+ float alphaY = ay-y1;
+ if (x1 < 0) x1 = 0;
+ if (y1 < 0) y1 = 0;
+ if (x2 >= mXSize) x2 = mXSize-1;
+ if (y2 >= mYSize) y2 = mYSize-1;
+ float a = (1.0-alphaX)*mData[x1+y1*mXSize]+alphaX*mData[x2+y1*mXSize];
+ float b = (1.0-alphaX)*mData[x1+y2*mXSize]+alphaX*mData[x2+y2*mXSize];
+ aNewData[x+y*aNewXSize] = (1.0-alphaY)*a+alphaY*b;
+ }
+ delete[] mData;
+ mData = aNewData;
+ mXSize = aNewXSize;
+ mYSize = aNewYSize;
+}
+
+template <class T>
+void CMatrix<T>::rescale(int aNewXSize, int aNewYSize) {
+ if (mXSize >= aNewXSize) {
+ if (mYSize >= aNewYSize) downsample(aNewXSize,aNewYSize);
+ else {
+ downsample(aNewXSize,mYSize);
+ upsample(aNewXSize,aNewYSize);
+ }
+ }
+ else {
+ if (mYSize >= aNewYSize) {
+ downsample(mXSize,aNewYSize);
+ upsample(aNewXSize,aNewYSize);
+ }
+ else upsample(aNewXSize,aNewYSize);
+ }
+}
+
+// identity
+template <class T>
+void CMatrix<T>::identity(int aSize) {
+ if (aSize != mXSize || aSize != mYSize) {
+ delete[] mData;
+ mData = new T[aSize*aSize];
+ mXSize = aSize;
+ mYSize = aSize;
+ }
+ fill(0);
+ for (int i = 0; i < aSize; i++)
+ operator()(i,i) = 1;
+}
+
+// fill
+template <class T>
+void CMatrix<T>::fill(const T aValue) {
+ int wholeSize = mXSize*mYSize;
+ for (register int i = 0; i < wholeSize; i++)
+ mData[i] = aValue;
+}
+
+// fillRect
+template <class T>
+void CMatrix<T>::fillRect(const T aValue, int ax1, int ay1, int ax2, int ay2) {
+ for (int y = ay1; y <= ay2; y++)
+ for (register int x = ax1; x <= ax2; x++)
+ operator()(x,y) = aValue;
+}
+
+// cut
+template <class T>
+void CMatrix<T>::cut(CMatrix<T>& aResult,const int x1, const int y1, const int x2, const int y2) {
+ aResult.mXSize = x2-x1+1;
+ aResult.mYSize = y2-y1+1;
+ delete[] aResult.mData;
+ aResult.mData = new T[aResult.mXSize*aResult.mYSize];
+ for (int y = y1; y <= y2; y++)
+ for (int x = x1; x <= x2; x++)
+ aResult(x-x1,y-y1) = operator()(x,y);
+}
+
+// paste
+template <class T>
+void CMatrix<T>::paste(CMatrix<T>& aCopyFrom, int ax, int ay) {
+ for (int y = 0; y < aCopyFrom.ySize(); y++)
+ for (int x = 0; x < aCopyFrom.xSize(); x++)
+ operator()(ax+x,ay+y) = aCopyFrom(x,y);
+}
+
+// mirror
+template <class T>
+void CMatrix<T>::mirror(int aFrom, int aTo) {
+ int aToXIndex = mXSize-aTo-1;
+ int aToYIndex = mYSize-aTo-1;
+ int aFromXIndex = mXSize-aFrom-1;
+ int aFromYIndex = mYSize-aFrom-1;
+ for (int y = aFrom; y <= aFromYIndex; y++) {
+ operator()(aTo,y) = operator()(aFrom,y);
+ operator()(aToXIndex,y) = operator()(aFromXIndex,y);
+ }
+ for (int x = aTo; x <= aToXIndex; x++) {
+ operator()(x,aTo) = operator()(x,aFrom);
+ operator()(x,aToYIndex) = operator()(x,aFromYIndex);
+ }
+}
+
+// normalize
+template <class T>
+void CMatrix<T>::normalize(T aMin, T aMax, T aInitialMin, T aInitialMax) {
+ int aSize = mXSize*mYSize;
+ T aCurrentMin = aInitialMax;
+ T aCurrentMax = aInitialMin;
+ for (int i = 0; i < aSize; i++)
+ if (mData[i] > aCurrentMax) aCurrentMax = mData[i];
+ else if (mData[i] < aCurrentMin) aCurrentMin = mData[i];
+ T aTemp = (aCurrentMax-aCurrentMin);
+ if (aTemp == 0) aTemp = 1;
+ else aTemp = (aMax-aMin)/aTemp;
+ for (int i = 0; i < aSize; i++) {
+ mData[i] -= aCurrentMin;
+ mData[i] *= aTemp;
+ mData[i] += aMin;
+ }
+}
+
+// clip
+template <class T>
+void CMatrix<T>::clip(T aMin, T aMax) {
+ int aSize = size();
+ for (int i = 0; i < aSize; i++)
+ if (mData[i] < aMin) mData[i] = aMin;
+ else if (mData[i] > aMax) mData[i] = aMax;
+}
+
+// applySimilarityTransform
+template <class T>
+void CMatrix<T>::applySimilarityTransform(CMatrix<T>& aWarped, CMatrix<bool>& aOutside, float tx, float ty, float cx, float cy, float phi, float scale) {
+ float cosphi = scale*cos(phi);
+ float sinphi = scale*sin(phi);
+ float ctx = cx+tx-cx*cosphi+cy*sinphi;
+ float cty = cy+ty-cy*cosphi-cx*sinphi;
+ aOutside = false;
+ int i = 0;
+ for (int y = 0; y < aWarped.ySize(); y++)
+ for (int x = 0; x < aWarped.xSize(); x++,i++) {
+ float xf = x; float yf = y;
+ float ax = xf*cosphi-yf*sinphi+ctx;
+ float ay = yf*cosphi+xf*sinphi+cty;
+ int x1 = (int)ax; int y1 = (int)ay;
+ float alphaX = ax-x1; float alphaY = ay-y1;
+ float betaX = 1.0-alphaX; float betaY = 1.0-alphaY;
+ if (x1 < 0 || y1 < 0 || x1+1 >= mXSize || y1+1 >= mYSize) aOutside.data()[i] = true;
+ else {
+ int j = y1*mXSize+x1;
+ float a = betaX*mData[j] +alphaX*mData[j+1];
+ float b = betaX*mData[j+mXSize]+alphaX*mData[j+1+mXSize];
+ aWarped.data()[i] = betaY*a+alphaY*b;
+ }
+ }
+}
+
+// applyHomography
+template <class T>
+void CMatrix<T>::applyHomography(CMatrix<T>& aWarped, CMatrix<bool>& aOutside, const CMatrix<float>& H) {
+ int aSize = size();
+ aOutside = false;
+ int i = 0;
+ for (int y = 0; y < aWarped.ySize(); y++)
+ for (int x = 0; x < aWarped.xSize(); x++,i++) {
+ float xf = x; float yf = y;
+ float ax = H.data()[0]*xf+H.data()[1]*yf+H.data()[2];
+ float ay = H.data()[3]*xf+H.data()[4]*yf+H.data()[5];
+ float az = H.data()[6]*xf+H.data()[7]*yf+H.data()[8];
+ float invaz = 1.0/az;
+ ax *= invaz; ay *= invaz;
+ int x1 = (int)ax; int y1 = (int)ay;
+ float alphaX = ax-x1; float alphaY = ay-y1;
+ float betaX = 1.0-alphaX; float betaY = 1.0-alphaY;
+ if (x1 < 0 || y1 < 0 || x1+1 >= mXSize || y1+1 >= mYSize) aOutside.data()[i] = true;
+ else {
+ int j = y1*mXSize+x1;
+ float a = betaX*mData[j] +alphaX*mData[j+1];
+ float b = betaX*mData[j+mXSize]+alphaX*mData[j+1+mXSize];
+ aWarped.data()[i] = betaY*a+alphaY*b;
+ }
+ }
+}
+
+// drawLine
+template <class T>
+void CMatrix<T>::drawLine(int dStartX, int dStartY, int dEndX, int dEndY, T aValue) {
+ // vertical line
+ if (dStartX == dEndX) {
+ if (dStartX < 0 || dStartX >= mXSize) return;
+ int x = dStartX;
+ if (dStartY < dEndY) {
+ for (int y = dStartY; y <= dEndY; y++)
+ if (y >= 0 && y < mYSize) mData[x+y*mXSize] = aValue;
+ }
+ else {
+ for (int y = dStartY; y >= dEndY; y--)
+ if (y >= 0 && y < mYSize) mData[x+y*mXSize] = aValue;
+ }
+ return;
+ }
+ // horizontal line
+ if (dStartY == dEndY) {
+ if (dStartY < 0 || dStartY >= mYSize) return;
+ int y = dStartY;
+ if (dStartX < dEndX) {
+ for (int x = dStartX; x <= dEndX; x++)
+ if (x >= 0 && x < mXSize) mData[x+y*mXSize] = aValue;
+ }
+ else {
+ for (int x = dStartX; x >= dEndX; x--)
+ if (x >= 0 && x < mXSize) mData[x+y*mXSize] = aValue;
+ }
+ return;
+ }
+ float m = float(dStartY - dEndY) / float(dStartX - dEndX);
+ float invm = 1.0/m;
+ if (fabs(m) > 1.0) {
+ if (dEndY > dStartY) {
+ for (int y = dStartY; y <= dEndY; y++) {
+ int x = (int)(0.5+dStartX+(y-dStartY)*invm);
+ if (x >= 0 && x < mXSize && y >= 0 && y < mYSize)
+ mData[x+y*mXSize] = aValue;
+ }
+ }
+ else {
+ for (int y = dStartY; y >= dEndY; y--) {
+ int x = (int)(0.5+dStartX+(y-dStartY)*invm);
+ if (x >= 0 && x < mXSize && y >= 0 && y < mYSize)
+ mData[x+y*mXSize] = aValue;
+ }
+ }
+ }
+ else {
+ if (dEndX > dStartX) {
+ for (int x = dStartX; x <= dEndX; x++) {
+ int y = (int)(0.5+dStartY+(x-dStartX)*m);
+ if (x >= 0 && x < mXSize && y >= 0 && y < mYSize)
+ mData[x+y*mXSize] = aValue;
+ }
+ }
+ else {
+ for (int x = dStartX; x >= dEndX; x--) {
+ int y = (int)(0.5+dStartY+(x-dStartX)*m);
+ if (x >= 0 && x < mXSize && y >= 0 && y < mYSize)
+ mData[x+y*mXSize] = aValue;
+ }
+ }
+ }
+}
+
+// invertImage
+template <class T>
+void CMatrix<T>::invertImage() {
+ int aSize = mXSize*mYSize;
+ for (int i = 0; i < aSize; i++)
+ mData[i] = 255-mData[i];
+}
+
+// connectedComponent
+typedef struct {short y, xl, xr, dy;} CSegment;
+
+template <class T>
+void CMatrix<T>::connectedComponent (int x, int y) {
+ std::stack<CSegment> aStack;
+ #define PUSH(Y,XL,XR,DY) if (Y+(DY)>=0 && Y+(DY)<mYSize)\
+ {CSegment S; S.y = Y; S.xl = XL; S.xr = XR;S.dy = DY;aStack.push(S);}
+ #define POP(Y,XL,XR,DY) {CSegment& S = aStack.top(); Y = S.y+(DY = S.dy);XL = S.xl; XR = S.xr; aStack.pop();}
+ T aCompValue = operator()(x,y);
+ CMatrix<bool> aConnected(mXSize,mYSize,false);
+ int l,x1,x2,dy;
+ PUSH(y,x,x,1);
+ PUSH(y+1,x,x,-1);
+ while (!aStack.empty()) {
+ POP(y,x1,x2,dy);
+ for (x=x1; x >= 0 && operator()(x,y) == aCompValue && !aConnected(x,y);x--)
+ aConnected(x,y) = true;
+ if (x >= x1) goto skip2;
+ l = x+1;
+ if (l < x1) PUSH(y,l,x1-1,-dy);
+ x = x1+1;
+ do {
+ for (; x < mXSize && operator()(x,y) == aCompValue && !aConnected(x,y); x++)
+ aConnected(x,y) = true;
+ PUSH(y,l,x-1,dy);
+ if (x>x2+1) PUSH(y,x2+1,x-1,-dy);
+ skip2: for (x++;x <= x2 && (operator()(x,y) != aCompValue || aConnected(x,y)); x++);
+ l = x;
+ }
+ while (x <= x2);
+ }
+ int aSize = size();
+ for (int i = 0; i < aSize; i++)
+ if (aConnected.data()[i]) mData[i] = 255;
+ else mData[i] = 0;
+ #undef PUSH
+ #undef POP
+}
+
+// append
+template <class T>
+void CMatrix<T>::append(CMatrix<T>& aMatrix) {
+ #ifdef _DEBUG
+ if (aMatrix.xSize() != mXSize) throw EIncompatibleMatrices(mXSize,mYSize,aMatrix.xSize(),aMatrix.ySize());
+ #endif
+ T* aNew = new T[mXSize*(mYSize+aMatrix.ySize())];
+ int aSize = mXSize*mYSize;
+ for (int i = 0; i < aSize; i++)
+ aNew[i] = mData[i];
+ int aSize2 = mXSize*aMatrix.ySize();
+ for (int i = 0; i < aSize2; i++)
+ aNew[i+aSize] = aMatrix.data()[i];
+ delete[] mData;
+ mData = aNew;
+ mYSize += aMatrix.ySize();
+}
+
+// inv
+template <class T>
+void CMatrix<T>::inv() {
+ if (mXSize != mYSize) throw ENonquadraticMatrix(mXSize,mYSize);
+ int* p = new int[mXSize];
+ T* hv = new T[mXSize];
+ CMatrix<T>& I(*this);
+ int n = mYSize;
+ for (int j = 0; j < n; j++)
+ p[j] = j;
+ for (int j = 0; j < n; j++) {
+ T max = fabs(I(j,j));
+ int r = j;
+ for (int i = j+1; i < n; i++)
+ if (fabs(I(j,i)) > max) {
+ max = fabs(I(j,i));
+ r = i;
+ }
+ // Matrix singular
+ if (max <= 0) return;
+ // Swap row j and r
+ if (r > j) {
+ for (int k = 0; k < n; k++) {
+ T hr = I(k,j);
+ I(k,j) = I(k,r);
+ I(k,r) = hr;
+ }
+ int hi = p[j];
+ p[j] = p[r];
+ p[r] = hi;
+ }
+ T hr = 1/I(j,j);
+ for (int i = 0; i < n; i++)
+ I(j,i) *= hr;
+ I(j,j) = hr;
+ hr *= -1;
+ for (int k = 0; k < n; k++)
+ if (k != j) {
+ for (int i = 0; i < n; i++)
+ if (i != j) I(k,i) -= I(j,i)*I(k,j);
+ I(k,j) *= hr;
+ }
+ }
+ for (int i = 0; i < n; i++) {
+ for (int k = 0; k < n; k++)
+ hv[p[k]] = I(k,i);
+ for (int k = 0; k < n; k++)
+ I(k,i) = hv[k];
+ }
+ delete[] p;
+ delete[] hv;
+}
+
+template <class T>
+void CMatrix<T>::trans() {
+ for (int y = 0; y < mYSize; y++)
+ for (int x = y; x < mXSize; x++) {
+ float temp = operator()(x,y);
+ operator()(x,y) = operator()(y,x);
+ operator()(y,x) = temp;
+ }
+}
+
+template <class T>
+float CMatrix<T>::scalar(CVector<T>& aLeft, CVector<T>& aRight) {
+ #ifdef _DEBUG
+ if ((aLeft.size() != mYSize) || (aRight.size() != mXSize))
+ throw EIncompatibleMatrices(mXSize,mYSize,aRight.size(),aLeft.size());
+ #endif
+ T* vec = new T[mYSize];
+ T* dat = mData;
+ for (int y = 0; y < mYSize; y++) {
+ vec[y] = 0;
+ for (int x = 0; x < mXSize; x++)
+ vec[y] += *(dat++)*aRight(x);
+ }
+ T aResult = 0.0;
+ for (int y = 0; y < mYSize; y++)
+ aResult += vec[y]*aLeft(y);
+ delete[] vec;
+ return aResult;
+}
+
+// readFromPGM
+template <class T>
+void CMatrix<T>::readFromPGM(const char* aFilename) {
+ FILE *aStream;
+ aStream = fopen(aFilename,"rb");
+ if (aStream == 0) std::cerr << "File not found: " << aFilename << std::endl;
+ int dummy;
+ // Find beginning of file (P5)
+ while (getc(aStream) != 'P');
+ if (getc(aStream) != '5') throw EInvalidFileFormat("PGM");
+ do dummy = getc(aStream); while (dummy != '\n' && dummy != ' ');
+ // Remove comments and empty lines
+ dummy = getc(aStream);
+ while (dummy == '#') {
+ while (getc(aStream) != '\n');
+ dummy = getc(aStream);
+ }
+ while (dummy == '\n')
+ dummy = getc(aStream);
+ // Read image size
+ mXSize = dummy-48;
+ while ((dummy = getc(aStream)) >= 48 && dummy < 58)
+ mXSize = 10*mXSize+dummy-48;
+ while ((dummy = getc(aStream)) < 48 || dummy >= 58);
+ mYSize = dummy-48;
+ while ((dummy = getc(aStream)) >= 48 && dummy < 58)
+ mYSize = 10*mYSize+dummy-48;
+ while (dummy != '\n' && dummy != ' ')
+ dummy = getc(aStream);
+ while ((dummy = getc(aStream)) >= 48 && dummy < 58);
+ if (dummy != '\n') while (getc(aStream) != '\n');
+ // Adjust size of data structure
+ delete[] mData;
+ mData = new T[mXSize*mYSize];
+ // Read image data
+ for (int i = 0; i < mXSize*mYSize; i++)
+ mData[i] = getc(aStream);
+ fclose(aStream);
+}
+
+// writeToPGM
+template <class T>
+void CMatrix<T>::writeToPGM(const char *aFilename) {
+ FILE *aStream;
+ aStream = fopen(aFilename,"wb");
+ // write header
+ char line[60];
+ sprintf(line,"P5\n%d %d\n255\n",mXSize,mYSize);
+ fwrite(line,strlen(line),1,aStream);
+ // write data
+ for (int i = 0; i < mXSize*mYSize; i++) {
+ char dummy = (char)mData[i];
+ fwrite(&dummy,1,1,aStream);
+ }
+ fclose(aStream);
+}
+
+// readFromTXT
+template <class T>
+void CMatrix<T>::readFromTXT(const char* aFilename, bool aHeader, int aXSize, int aYSize) {
+ std::ifstream aStream(aFilename);
+ // read header
+ if (aHeader) aStream >> mXSize >> mYSize;
+ else {
+ mXSize = aXSize;
+ mYSize = aYSize;
+ }
+ // Adjust size of data structure
+ delete[] mData;
+ mData = new T[mXSize*mYSize];
+ // read data
+ for (int i = 0; i < mXSize*mYSize; i++)
+ aStream >> mData[i];
+}
+
+// readFromMatlabTXT
+template <class T>
+void CMatrix<T>::readFromMatlabTXT(const char* aFilename, bool aHeader, int aXSize, int aYSize) {
+ std::ifstream aStream(aFilename);
+ // read header
+ float nx,ny;
+ if (aHeader) {
+ aStream >> nx >> ny;
+ mXSize = (int)nx; mYSize = (int)ny;
+ }
+ else {
+ mXSize = aXSize; mYSize = aYSize;
+ }
+ // Adjust size of data structure
+ delete[] mData;
+ mData = new T[mXSize*mYSize];
+ // read data
+ for (int i = 0; i < mXSize*mYSize; i++)
+ aStream >> mData[i];
+}
+
+//writeToTXT
+template <class T>
+void CMatrix<T>::writeToTXT(const char* aFilename, bool aHeader) {
+ std::ofstream aStream(aFilename);
+ // write header
+ if (aHeader) aStream << mXSize << " " << mYSize << std::endl;
+ // write data
+ int i = 0;
+ for (int y = 0; y < mYSize; y++) {
+ for (int x = 0; x < mXSize; x++, i++)
+ aStream << mData[i] << " ";
+ aStream << std::endl;
+ }
+}
+
+// readBodoProjectionMatrix
+template <class T>
+void CMatrix<T>::readBodoProjectionMatrix(const char* aFilename) {
+ readFromTXT(aFilename,false,4,3);
+}
+
+// operator ()
+template <class T>
+inline T& CMatrix<T>::operator()(const int ax, const int ay) const {
+ #ifdef _DEBUG
+ if (ax >= mXSize || ay >= mYSize || ax < 0 || ay < 0)
+ throw EMatrixRangeOverflow(ax,ay);
+ #endif
+ return mData[mXSize*ay+ax];
+}
+
+// operator =
+template <class T>
+inline CMatrix<T>& CMatrix<T>::operator=(const T aValue) {
+ fill(aValue);
+ return *this;
+}
+
+template <class T>
+CMatrix<T>& CMatrix<T>::operator=(const CMatrix<T>& aCopyFrom) {
+ if (this != &aCopyFrom) {
+ if (mData != 0) delete[] mData;
+ mXSize = aCopyFrom.mXSize;
+ mYSize = aCopyFrom.mYSize;
+ if (aCopyFrom.mData == 0) mData = 0;
+ else {
+ int wholeSize = mXSize*mYSize;
+ mData = new T[wholeSize];
+ for (register int i = 0; i < wholeSize; i++)
+ mData[i] = aCopyFrom.mData[i];
+ }
+ }
+ return *this;
+}
+
+// operator +=
+template <class T>
+CMatrix<T>& CMatrix<T>::operator+=(const CMatrix<T>& aMatrix) {
+ if ((mXSize != aMatrix.mXSize) || (mYSize != aMatrix.mYSize))
+ throw EIncompatibleMatrices(mXSize,mYSize,aMatrix.mXSize,aMatrix.mYSize);
+ int wholeSize = mXSize*mYSize;
+ for (int i = 0; i < wholeSize; i++)
+ mData[i] += aMatrix.mData[i];
+ return *this;
+}
+
+template <class T>
+CMatrix<T>& CMatrix<T>::operator+=(const T aValue) {
+ int wholeSize = mXSize*mYSize;
+ for (int i = 0; i < wholeSize; i++)
+ mData[i] += aValue;
+ return *this;
+}
+
+// operator -=
+template <class T>
+CMatrix<T>& CMatrix<T>::operator-=(const CMatrix<T>& aMatrix) {
+ if ((mXSize != aMatrix.mXSize) || (mYSize != aMatrix.mYSize))
+ throw EIncompatibleMatrices(mXSize,mYSize,aMatrix.mXSize,aMatrix.mYSize);
+ int wholeSize = mXSize*mYSize;
+ for (int i = 0; i < wholeSize; i++)
+ mData[i] -= aMatrix.mData[i];
+ return *this;
+}
+
+// operator *=
+template <class T>
+CMatrix<T>& CMatrix<T>::operator*=(const CMatrix<T>& aMatrix) {
+ if (mXSize != aMatrix.mYSize)
+ throw EIncompatibleMatrices(mXSize,mYSize,aMatrix.mXSize,aMatrix.mYSize);
+ T* oldData = mData;
+ mData = new T[mYSize*aMatrix.mXSize];
+ for (int y = 0; y < mYSize; y++)
+ for (int x = 0; x < aMatrix.mXSize; x++) {
+ mData[aMatrix.mXSize*y+x] = 0;
+ for (int i = 0; i < mXSize; i++)
+ mData[aMatrix.mXSize*y+x] += oldData[mXSize*y+i]*aMatrix(x,i);
+ }
+ delete[] oldData;
+ mXSize = aMatrix.mXSize;
+ return *this;
+}
+
+template <class T>
+CMatrix<T>& CMatrix<T>::operator*=(const T aValue) {
+ int wholeSize = mXSize*mYSize;
+ for (int i = 0; i < wholeSize; i++)
+ mData[i] *= aValue;
+ return *this;
+}
+
+// min
+template <class T>
+T CMatrix<T>::min() const {
+ T aMin = mData[0];
+ int aSize = mXSize*mYSize;
+ for (int i = 1; i < aSize; i++)
+ if (mData[i] < aMin) aMin = mData[i];
+ return aMin;
+}
+
+// max
+template <class T>
+T CMatrix<T>::max() const {
+ T aMax = mData[0];
+ int aSize = mXSize*mYSize;
+ for (int i = 1; i < aSize; i++)
+ if (mData[i] > aMax) aMax = mData[i];
+ return aMax;
+}
+
+// avg
+template <class T>
+T CMatrix<T>::avg() const {
+ T aAvg = 0;
+ int aSize = mXSize*mYSize;
+ for (int i = 0; i < aSize; i++)
+ aAvg += mData[i];
+ return aAvg/aSize;
+}
+
+// xSize
+template <class T>
+inline int CMatrix<T>::xSize() const {
+ return mXSize;
+}
+
+// ySize
+template <class T>
+inline int CMatrix<T>::ySize() const {
+ return mYSize;
+}
+
+// size
+template <class T>
+inline int CMatrix<T>::size() const {
+ return mXSize*mYSize;
+}
+
+// getVector
+template <class T>
+void CMatrix<T>::getVector(CVector<T>& aVector, int ay) {
+ int aOffset = mXSize*ay;
+ for (int x = 0; x < mXSize; x++)
+ aVector(x) = mData[x+aOffset];
+}
+
+// data()
+template <class T>
+inline T* CMatrix<T>::data() const {
+ return mData;
+}
+
+// N O N - M E M B E R F U N C T I O N S --------------------------------------
+
+// abs
+template <class T>
+CMatrix<T> abs(const CMatrix<T>& aMatrix) {
+ CMatrix<T> result(aMatrix.xSize(),aMatrix.ySize());
+ int wholeSize = aMatrix.size();
+ for (register int i = 0; i < wholeSize; i++) {
+ if (aMatrix.data()[i] < 0) result.data()[i] = -aMatrix.data()[i];
+ else result.data()[i] = aMatrix.data()[i];
+ }
+ return result;
+}
+
+// trans
+template <class T>
+CMatrix<T> trans(const CMatrix<T>& aMatrix) {
+ CMatrix<T> result(aMatrix.ySize(),aMatrix.xSize());
+ for (int y = 0; y < aMatrix.ySize(); y++)
+ for (int x = 0; x < aMatrix.xSize(); x++)
+ result(y,x) = aMatrix(x,y);
+ return result;
+}
+
+// operator +
+template <class T>
+CMatrix<T> operator+(const CMatrix<T>& aM1, const CMatrix<T>& aM2) {
+ if ((aM1.xSize() != aM2.xSize()) || (aM1.ySize() != aM2.ySize()))
+ throw EIncompatibleMatrices(aM1.xSize(),aM1.ySize(),aM2.xSize(),aM2.ySize());
+ CMatrix<T> result(aM1.xSize(),aM1.ySize());
+ int wholeSize = aM1.xSize()*aM1.ySize();
+ for (int i = 0; i < wholeSize; i++)
+ result.data()[i] = aM1.data()[i] + aM2.data()[i];
+ return result;
+}
+
+// operator -
+template <class T>
+CMatrix<T> operator-(const CMatrix<T>& aM1, const CMatrix<T>& aM2) {
+ if ((aM1.xSize() != aM2.xSize()) || (aM1.ySize() != aM2.ySize()))
+ throw EIncompatibleMatrices(aM1.xSize(),aM1.ySize(),aM2.xSize(),aM2.ySize());
+ CMatrix<T> result(aM1.xSize(),aM1.ySize());
+ int wholeSize = aM1.xSize()*aM1.ySize();
+ for (int i = 0; i < wholeSize; i++)
+ result.data()[i] = aM1.data()[i] - aM2.data()[i];
+ return result;
+}
+
+// operator *
+template <class T>
+CMatrix<T> operator*(const CMatrix<T>& aM1, const CMatrix<T>& aM2) {
+ if (aM1.xSize() != aM2.ySize())
+ throw EIncompatibleMatrices(aM1.xSize(),aM1.ySize(),aM2.xSize(),aM2.ySize());
+ CMatrix<T> result(aM2.xSize(),aM1.ySize(),0);
+ for (int y = 0; y < result.ySize(); y++)
+ for (int x = 0; x < result.xSize(); x++)
+ for (int i = 0; i < aM1.xSize(); i++)
+ result(x,y) += aM1(i,y)*aM2(x,i);
+ return result;
+}
+
+template <class T>
+CVector<T> operator*(const CMatrix<T>& aMatrix, const CVector<T>& aVector) {
+ if (aMatrix.xSize() != aVector.size())
+ throw EIncompatibleMatrices(aMatrix.xSize(),aMatrix.ySize(),1,aVector.size());
+ CVector<T> result(aMatrix.ySize(),0);
+ for (int y = 0; y < aMatrix.ySize(); y++)
+ for (int x = 0; x < aMatrix.xSize(); x++)
+ result(y) += aMatrix(x,y)*aVector(x);
+ return result;
+}
+
+template <class T>
+CMatrix<T> operator*(const CMatrix<T>& aMatrix, const T aValue) {
+ CMatrix<T> result(aMatrix.xSize(),aMatrix.ySize());
+ int wholeSize = aMatrix.xSize()*aMatrix.ySize();
+ for (int i = 0; i < wholeSize; i++)
+ result.data()[i] = aMatrix.data()[i]*aValue;
+ return result;
+}
+
+template <class T>
+inline CMatrix<T> operator*(const T aValue, const CMatrix<T>& aMatrix) {
+ return aMatrix*aValue;
+}
+
+// operator <<
+template <class T>
+std::ostream& operator<<(std::ostream& aStream, const CMatrix<T>& aMatrix) {
+ for (int y = 0; y < aMatrix.ySize(); y++) {
+ for (int x = 0; x < aMatrix.xSize(); x++)
+ aStream << aMatrix(x,y) << ' ';
+ aStream << std::endl;
+ }
+ return aStream;
+}
+
+
+// Comparison of two matrices
+template <class T> bool CMatrix<T>::operator==(const CMatrix<T>& aMatrix)
+{
+ if((*this).size()!=aMatrix.size())
+ return false;
+
+ for(int i=0; i<aMatrix.size();i++)
+ if(mData[i] != aMatrix.mData[i])
+ return false;
+ return true;
+}
+
+#endif
diff --git a/video_input/consistencyChecker/CTensor.h b/video_input/consistencyChecker/CTensor.h new file mode 100644 index 0000000..0f5af5c --- /dev/null +++ b/video_input/consistencyChecker/CTensor.h @@ -0,0 +1,1205 @@ +// CTensor
+// A three-dimensional array
+//
+// Author: Thomas Brox
+
+#ifndef CTENSOR_H
+#define CTENSOR_H
+
+#include <iostream>
+#include <fstream>
+#include <string>
+#include <sstream>
+#include <CMatrix.h>
+#include <NMath.h>
+
+inline int int_min(int x, int& y) { return (x<y)?x:y; }
+inline int int_max(int x, int& y) { return (x<y)?y:x; }
+
+template <class T>
+class CTensor {
+public:
+ // standard constructor
+ inline CTensor();
+ // constructor
+ inline CTensor(const int aXSize, const int aYSize, const int aZSize);
+ // copy constructor
+ CTensor(const CTensor<T>& aCopyFrom);
+ // constructor with implicit filling
+ CTensor(const int aXSize, const int aYSize, const int aZSize, const T aFillValue);
+ // destructor
+ virtual ~CTensor();
+
+ // Changes the size of the tensor, data will be lost
+ void setSize(int aXSize, int aYSize, int aZSize);
+ // Downsamples the tensor
+ void downsample(int aNewXSize, int aNewYSize);
+ void downsample(int aNewXSize, int aNewYSize, CMatrix<float>& aConfidence);
+ void downsample(int aNewXSize, int aNewYSize, CTensor<float>& aConfidence);
+ // Upsamples the tensor
+ void upsample(int aNewXSize, int aNewYSize);
+ void upsampleBilinear(int aNewXSize, int aNewYSize);
+ // Fills the tensor with the value aValue (see also operator =)
+ void fill(const T aValue);
+ // Fills a rectangular area with the value aValue
+ void fillRect(const CVector<T>& aValue, int ax1, int ay1, int ax2, int ay2);
+ // Copies a box from the tensor into aResult, the size of aResult will be adjusted
+ void cut(CTensor<T>& aResult, int x1, int y1, int z1, int x2, int y2, int z2);
+ // Copies aCopyFrom at a certain position of the tensor
+ void paste(CTensor<T>& aCopyFrom, int ax, int ay, int az);
+ // Mirrors the boundaries, aFrom is the distance from the boundaries where the pixels are copied from,
+ // aTo is the distance from the boundaries they are copied to
+ void mirrorLayers(int aFrom, int aTo);
+ // Transforms the values so that they are all between aMin and aMax
+ // aInitialMin/Max are initializations for seeking the minimum and maximum, change if your
+ // data is not in this range or the data type T cannot hold these values
+ void normalizeEach(T aMin, T aMax, T aInitialMin = -30000, T aInitialMax = 30000);
+ void normalize(T aMin, T aMax, int aChannel, T aInitialMin = -30000, T aInitialMax = 30000);
+ void normalize(T aMin, T aMax, T aInitialMin = -30000, T aInitialMax = 30000);
+ // Converts from RGB to CIELab color space and vice-versa
+ void rgbToCielab();
+ void cielabToRGB();
+ // Draws a line into the image (only for mZSize = 3)
+ void drawLine(int dStartX, int dStartY, int dEndX, int dEndY, T aValue1 = 255, T aValue2 = 255, T aValue3 = 255);
+ void drawRect(int dStartX, int dStartY, int dEndX, int dEndY, T aValue1 = 255, T aValue2 = 255, T aValue3 = 255);
+
+ // Applies a similarity transform (translation, rotation, scaling) to the image
+ void applySimilarityTransform(CTensor<T>& aWarped, CMatrix<bool>& aOutside, float tx, float ty, float cx, float cy, float phi, float scale);
+ // Applies a homography (linear projective transformation) to the image
+ void applyHomography(CTensor<T>& aWarped, CMatrix<bool>& aOutside, const CMatrix<float>& H);
+
+ // Reads the tensor from a file in Mathematica format
+ void readFromMathematicaFile(const char* aFilename);
+ // Writes the tensor to a file in Mathematica format
+ void writeToMathematicaFile(const char* aFilename);
+ // Reads the tensor from a movie file in IM format
+ void readFromIMFile(const char* aFilename);
+ // Writes the tensor to a movie file in IM format
+ void writeToIMFile(const char* aFilename);
+ // Reads an image from a PGM file
+ void readFromPGM(const char* aFilename);
+ // Writes the tensor in PGM-Format
+ void writeToPGM(const char* aFilename);
+ // Extends a XxYx1 tensor to a XxYx3 tensor with three identical layers
+ void makeColorTensor();
+ // Reads a color image from a PPM file
+ void readFromPPM(const char* aFilename);
+ // Writes the tensor in PPM-Format
+ void writeToPPM(const char* aFilename);
+ // Reads the tensor from a PDM file
+ void readFromPDM(const char* aFilename);
+ // Writes the tensor in PDM-Format
+ void writeToPDM(const char* aFilename, char aFeatureType);
+
+ // Gives full access to tensor's values
+ inline T& operator()(const int ax, const int ay, const int az) const;
+ // Read access with bilinear interpolation
+ CVector<T> operator()(const float ax, const float ay) const;
+ // Fills the tensor with the value aValue (equivalent to fill())
+ inline CTensor<T>& operator=(const T aValue);
+ // Copies the tensor aCopyFrom to this tensor (size of tensor might change)
+ CTensor<T>& operator=(const CTensor<T>& aCopyFrom);
+ // Adds a tensor of same size
+ CTensor<T>& operator+=(const CTensor<T>& aMatrix);
+ // Adds a constant to the tensor
+ CTensor<T>& operator+=(const T aValue);
+ // Multiplication with a scalar
+ CTensor<T>& operator*=(const T aValue);
+
+ // Returns the minimum value
+ T min() const;
+ // Returns the maximum value
+ T max() const;
+ // Returns the average value
+ T avg() const;
+ // Returns the average value of a specific layer
+ T avg(int az) const;
+ // Gives access to the tensor's size
+ inline int xSize() const;
+ inline int ySize() const;
+ inline int zSize() const;
+ inline int size() const;
+ // Returns the az layer of the tensor as matrix (slow and fast version)
+ CMatrix<T> getMatrix(const int az) const;
+ void getMatrix(CMatrix<T>& aMatrix, const int az) const;
+ // Copies the matrix components of aMatrix into the az layer of the tensor
+ void putMatrix(CMatrix<T>& aMatrix, const int az);
+ // Gives access to the internal data representation (use sparingly)
+ inline T* data() const;
+
+ // Possible interpretations of the third tensor dimension for PDM format
+ static const char cSpacial = 'S';
+ static const char cVector = 'V';
+ static const char cColor = 'C';
+ static const char cSymmetricMatrix = 'Y';
+protected:
+ int mXSize,mYSize,mZSize;
+ T *mData;
+};
+
+// Provides basic output functionality (only appropriate for very small tensors)
+template <class T> std::ostream& operator<<(std::ostream& aStream, const CTensor<T>& aTensor);
+
+// Exceptions thrown by CTensor-------------------------------------------------
+
+// Thrown when one tries to access an element of a tensor which is out of
+// the tensor's bounds
+struct ETensorRangeOverflow {
+ ETensorRangeOverflow(const int ax, const int ay, const int az) {
+ using namespace std;
+ cerr << "Exception ETensorRangeOverflow: x = " << ax << ", y = " << ay << ", z = " << az << endl;
+ }
+};
+
+// Thrown when the size of a tensor does not match the needed size for a certain operation
+struct ETensorIncompatibleSize {
+ ETensorIncompatibleSize(int ax, int ay, int ax2, int ay2) {
+ using namespace std;
+ cerr << "Exception ETensorIncompatibleSize: x = " << ax << ":" << ax2;
+ cerr << ", y = " << ay << ":" << ay2 << endl;
+ }
+ ETensorIncompatibleSize(int ax, int ay, int az) {
+ std::cerr << "Exception ETensorIncompatibleTensorSize: x = " << ax << ", y = " << ay << ", z= " << az << std::endl;
+ }
+};
+
+// I M P L E M E N T A T I O N --------------------------------------------
+//
+// You might wonder why there is implementation code in a header file.
+// The reason is that not all C++ compilers yet manage separate compilation
+// of templates. Inline functions cannot be compiled separately anyway.
+// So in this case the whole implementation code is added to the header
+// file.
+// Users of CTensor should ignore everything that's beyond this line :)
+// ------------------------------------------------------------------------
+
+// P U B L I C ------------------------------------------------------------
+
+// standard constructor
+template <class T>
+inline CTensor<T>::CTensor() {
+ mData = 0;
+ mXSize = mYSize = mZSize = 0;
+}
+
+// constructor
+template <class T>
+inline CTensor<T>::CTensor(const int aXSize, const int aYSize, const int aZSize)
+ : mXSize(aXSize), mYSize(aYSize), mZSize(aZSize) {
+ mData = new T[aXSize*aYSize*aZSize];
+}
+
+// copy constructor
+template <class T>
+CTensor<T>::CTensor(const CTensor<T>& aCopyFrom)
+ : mXSize(aCopyFrom.mXSize), mYSize(aCopyFrom.mYSize), mZSize(aCopyFrom.mZSize) {
+ int wholeSize = mXSize*mYSize*mZSize;
+ mData = new T[wholeSize];
+ for (register int i = 0; i < wholeSize; i++)
+ mData[i] = aCopyFrom.mData[i];
+}
+
+// constructor with implicit filling
+template <class T>
+CTensor<T>::CTensor(const int aXSize, const int aYSize, const int aZSize, const T aFillValue)
+ : mXSize(aXSize), mYSize(aYSize), mZSize(aZSize) {
+ mData = new T[aXSize*aYSize*aZSize];
+ fill(aFillValue);
+}
+
+// destructor
+template <class T>
+CTensor<T>::~CTensor() {
+ delete[] mData;
+}
+
+// setSize
+template <class T>
+void CTensor<T>::setSize(int aXSize, int aYSize, int aZSize) {
+ if (mData != 0) delete[] mData;
+ mData = new T[aXSize*aYSize*aZSize];
+ mXSize = aXSize;
+ mYSize = aYSize;
+ mZSize = aZSize;
+}
+
+//downsample
+template <class T>
+void CTensor<T>::downsample(int aNewXSize, int aNewYSize) {
+ T* mData2 = new T[aNewXSize*aNewYSize*mZSize];
+ int aSize = aNewXSize*aNewYSize;
+ for (int z = 0; z < mZSize; z++) {
+ CMatrix<T> aTemp(mXSize,mYSize);
+ getMatrix(aTemp,z);
+ aTemp.downsample(aNewXSize,aNewYSize);
+ for (int i = 0; i < aSize; i++)
+ mData2[i+z*aSize] = aTemp.data()[i];
+ }
+ delete[] mData;
+ mData = mData2;
+ mXSize = aNewXSize;
+ mYSize = aNewYSize;
+}
+
+template <class T>
+void CTensor<T>::downsample(int aNewXSize, int aNewYSize, CMatrix<float>& aConfidence) {
+ T* mData2 = new T[aNewXSize*aNewYSize*mZSize];
+ int aSize = aNewXSize*aNewYSize;
+ for (int z = 0; z < mZSize; z++) {
+ CMatrix<T> aTemp(mXSize,mYSize);
+ getMatrix(aTemp,z);
+ aTemp.downsample(aNewXSize,aNewYSize,aConfidence);
+ for (int i = 0; i < aSize; i++)
+ mData2[i+z*aSize] = aTemp.data()[i];
+ }
+ delete[] mData;
+ mData = mData2;
+ mXSize = aNewXSize;
+ mYSize = aNewYSize;
+}
+
+template <class T>
+void CTensor<T>::downsample(int aNewXSize, int aNewYSize, CTensor<float>& aConfidence) {
+ T* mData2 = new T[aNewXSize*aNewYSize*mZSize];
+ int aSize = aNewXSize*aNewYSize;
+ CMatrix<float> aConf(mXSize,mYSize);
+ for (int z = 0; z < mZSize; z++) {
+ CMatrix<T> aTemp(mXSize,mYSize);
+ getMatrix(aTemp,z);
+ aConfidence.getMatrix(aConf,z);
+ aTemp.downsample(aNewXSize,aNewYSize,aConf);
+ for (int i = 0; i < aSize; i++)
+ mData2[i+z*aSize] = aTemp.data()[i];
+ }
+ delete[] mData;
+ mData = mData2;
+ mXSize = aNewXSize;
+ mYSize = aNewYSize;
+}
+
+// upsample
+template <class T>
+void CTensor<T>::upsample(int aNewXSize, int aNewYSize) {
+ T* mData2 = new T[aNewXSize*aNewYSize*mZSize];
+ int aSize = aNewXSize*aNewYSize;
+ for (int z = 0; z < mZSize; z++) {
+ CMatrix<T> aTemp(mXSize,mYSize);
+ getMatrix(aTemp,z);
+ aTemp.upsample(aNewXSize,aNewYSize);
+ for (int i = 0; i < aSize; i++)
+ mData2[i+z*aSize] = aTemp.data()[i];
+ }
+ delete[] mData;
+ mData = mData2;
+ mXSize = aNewXSize;
+ mYSize = aNewYSize;
+}
+
+// upsampleBilinear
+template <class T>
+void CTensor<T>::upsampleBilinear(int aNewXSize, int aNewYSize) {
+ T* mData2 = new T[aNewXSize*aNewYSize*mZSize];
+ int aSize = aNewXSize*aNewYSize;
+ for (int z = 0; z < mZSize; z++) {
+ CMatrix<T> aTemp(mXSize,mYSize);
+ getMatrix(aTemp,z);
+ aTemp.upsampleBilinear(aNewXSize,aNewYSize);
+ for (int i = 0; i < aSize; i++)
+ mData2[i+z*aSize] = aTemp.data()[i];
+ }
+ delete[] mData;
+ mData = mData2;
+ mXSize = aNewXSize;
+ mYSize = aNewYSize;
+}
+
+// fill
+template <class T>
+void CTensor<T>::fill(const T aValue) {
+ int wholeSize = mXSize*mYSize*mZSize;
+ for (register int i = 0; i < wholeSize; i++)
+ mData[i] = aValue;
+}
+
+// fillRect
+template <class T>
+void CTensor<T>::fillRect(const CVector<T>& aValue, int ax1, int ay1, int ax2, int ay2) {
+ for (int z = 0; z < mZSize; z++) {
+ T val = aValue(z);
+ for (int y = int_max(0,ay1); y <= int_min(ySize()-1,ay2); y++)
+ for (register int x = int_max(0,ax1); x <= int_min(xSize()-1,ax2); x++)
+ operator()(x,y,z) = val;
+ }
+}
+
+// cut
+template <class T>
+void CTensor<T>::cut(CTensor<T>& aResult, int x1, int y1, int z1, int x2, int y2, int z2) {
+ aResult.mXSize = x2-x1+1;
+ aResult.mYSize = y2-y1+1;
+ aResult.mZSize = z2-z1+1;
+ delete[] aResult.mData;
+ aResult.mData = new T[aResult.mXSize*aResult.mYSize*aResult.mZSize];
+ for (int z = z1; z <= z2; z++)
+ for (int y = y1; y <= y2; y++)
+ for (int x = x1; x <= x2; x++)
+ aResult(x-x1,y-y1,z-z1) = operator()(x,y,z);
+}
+
+// paste
+template <class T>
+void CTensor<T>::paste(CTensor<T>& aCopyFrom, int ax, int ay, int az) {
+ for (int z = 0; z < aCopyFrom.zSize(); z++)
+ for (int y = 0; y < aCopyFrom.ySize(); y++)
+ for (int x = 0; x < aCopyFrom.xSize(); x++)
+ operator()(ax+x,ay+y,az+z) = aCopyFrom(x,y,z);
+}
+
+// mirrorLayers
+template <class T>
+void CTensor<T>::mirrorLayers(int aFrom, int aTo) {
+ for (int z = 0; z < mZSize; z++) {
+ int aToXIndex = mXSize-aTo-1;
+ int aToYIndex = mYSize-aTo-1;
+ int aFromXIndex = mXSize-aFrom-1;
+ int aFromYIndex = mYSize-aFrom-1;
+ for (int y = aFrom; y <= aFromYIndex; y++) {
+ operator()(aTo,y,z) = operator()(aFrom,y,z);
+ operator()(aToXIndex,y,z) = operator()(aFromXIndex,y,z);
+ }
+ for (int x = aTo; x <= aToXIndex; x++) {
+ operator()(x,aTo,z) = operator()(x,aFrom,z);
+ operator()(x,aToYIndex,z) = operator()(x,aFromYIndex,z);
+ }
+ }
+}
+
+// normalize
+template <class T>
+void CTensor<T>::normalizeEach(T aMin, T aMax, T aInitialMin, T aInitialMax) {
+ for (int k = 0; k < mZSize; k++)
+ normalize(aMin,aMax,k,aInitialMin,aInitialMax);
+}
+
+template <class T>
+void CTensor<T>::normalize(T aMin, T aMax, int aChannel, T aInitialMin, T aInitialMax) {
+ int aChannelSize = mXSize*mYSize;
+ T aCurrentMin = aInitialMax;
+ T aCurrentMax = aInitialMin;
+ int aIndex = aChannelSize*aChannel;
+ for (int i = 0; i < aChannelSize; i++) {
+ if (mData[aIndex] > aCurrentMax) aCurrentMax = mData[aIndex];
+ else if (mData[aIndex] < aCurrentMin) aCurrentMin = mData[aIndex];
+ aIndex++;
+ }
+ T aTemp1 = aCurrentMin - aMin;
+ T aTemp2 = (aCurrentMax-aCurrentMin);
+ if (aTemp2 == 0) aTemp2 = 1;
+ else aTemp2 = (aMax-aMin)/aTemp2;
+ aIndex = aChannelSize*aChannel;
+ for (int i = 0; i < aChannelSize; i++) {
+ mData[aIndex] -= aTemp1;
+ mData[aIndex] *= aTemp2;
+ aIndex++;
+ }
+}
+
+// drawLine
+template <class T>
+void CTensor<T>::drawLine(int dStartX, int dStartY, int dEndX, int dEndY, T aValue1, T aValue2, T aValue3) {
+ int aOffset1 = mXSize*mYSize;
+ int aOffset2 = 2*aOffset1;
+ // vertical line
+ if (dStartX == dEndX) {
+ if (dStartX < 0 || dStartX >= mXSize) return;
+ int x = dStartX;
+ if (dStartY < dEndY) {
+ for (int y = dStartY; y <= dEndY; y++)
+ if (y >= 0 && y < mYSize) {
+ mData[x+y*mXSize] = aValue1;
+ mData[x+y*mXSize+aOffset1] = aValue2;
+ mData[x+y*mXSize+aOffset2] = aValue3;
+ }
+ }
+ else {
+ for (int y = dStartY; y >= dEndY; y--)
+ if (y >= 0 && y < mYSize) {
+ mData[x+y*mXSize] = aValue1;
+ mData[x+y*mXSize+aOffset1] = aValue2;
+ mData[x+y*mXSize+aOffset2] = aValue3;
+ }
+ }
+ return;
+ }
+ // horizontal line
+ if (dStartY == dEndY) {
+ if (dStartY < 0 || dStartY >= mYSize) return;
+ int y = dStartY;
+ if (dStartX < dEndX) {
+ for (int x = dStartX; x <= dEndX; x++)
+ if (x >= 0 && x < mXSize) {
+ mData[x+y*mXSize] = aValue1;
+ mData[x+y*mXSize+aOffset1] = aValue2;
+ mData[x+y*mXSize+aOffset2] = aValue3;
+ }
+ }
+ else {
+ for (int x = dStartX; x >= dEndX; x--)
+ if (x >= 0 && x < mXSize) {
+ mData[x+y*mXSize] = aValue1;
+ mData[x+y*mXSize+aOffset1] = aValue2;
+ mData[x+y*mXSize+aOffset2] = aValue3;
+ }
+ }
+ return;
+ }
+ float m = float(dStartY - dEndY) / float(dStartX - dEndX);
+ float invm = 1.0/m;
+ if (fabs(m) > 1.0) {
+ if (dEndY > dStartY) {
+ for (int y = dStartY; y <= dEndY; y++) {
+ int x = (int)(0.5+dStartX+(y-dStartY)*invm);
+ if (x >= 0 && x < mXSize && y >= 0 && y < mYSize) {
+ mData[x+y*mXSize] = aValue1;
+ mData[x+y*mXSize+aOffset1] = aValue2;
+ mData[x+y*mXSize+aOffset2] = aValue3;
+ }
+ }
+ }
+ else {
+ for (int y = dStartY; y >= dEndY; y--) {
+ int x = (int)(0.5+dStartX+(y-dStartY)*invm);
+ if (x >= 0 && x < mXSize && y >= 0 && y < mYSize) {
+ mData[x+y*mXSize] = aValue1;
+ mData[x+y*mXSize+aOffset1] = aValue2;
+ mData[x+y*mXSize+aOffset2] = aValue3;
+ }
+ }
+ }
+ }
+ else {
+ if (dEndX > dStartX) {
+ for (int x = dStartX; x <= dEndX; x++) {
+ int y = (int)(0.5+dStartY+(x-dStartX)*m);
+ if (x >= 0 && x < mXSize && y >= 0 && y < mYSize) {
+ mData[x+y*mXSize] = aValue1;
+ mData[x+y*mXSize+aOffset1] = aValue2;
+ mData[x+y*mXSize+aOffset2] = aValue3;
+ }
+ }
+ }
+ else {
+ for (int x = dStartX; x >= dEndX; x--) {
+ int y = (int)(0.5+dStartY+(x-dStartX)*m);
+ if (x >= 0 && x < mXSize && y >= 0 && y < mYSize) {
+ mData[x+y*mXSize] = aValue1;
+ mData[x+y*mXSize+aOffset1] = aValue2;
+ mData[x+y*mXSize+aOffset2] = aValue3;
+ }
+ }
+ }
+ }
+}
+
+// drawRect
+template <class T>
+void CTensor<T>::drawRect(int dStartX, int dStartY, int dEndX, int dEndY, T aValue1, T aValue2, T aValue3) {
+ drawLine(dStartX,dStartY,dEndX,dStartY,aValue1,aValue2,aValue3);
+ drawLine(dStartX,dEndY,dEndX,dEndY,aValue1,aValue2,aValue3);
+ drawLine(dStartX,dStartY,dStartX,dEndY,aValue1,aValue2,aValue3);
+ drawLine(dEndX,dStartY,dEndX,dEndY,aValue1,aValue2,aValue3);
+}
+
+template <class T>
+void CTensor<T>::normalize(T aMin, T aMax, T aInitialMin, T aInitialMax) {
+ int aSize = mXSize*mYSize*mZSize;
+ T aCurrentMin = aInitialMax;
+ T aCurrentMax = aInitialMin;
+ for (int i = 0; i < aSize; i++) {
+ if (mData[i] > aCurrentMax) aCurrentMax = mData[i];
+ else if (mData[i] < aCurrentMin) aCurrentMin = mData[i];
+ }
+ T aTemp1 = aCurrentMin - aMin;
+ T aTemp2 = (aCurrentMax-aCurrentMin);
+ if (aTemp2 == 0) aTemp2 = 1;
+ else aTemp2 = (aMax-aMin)/aTemp2;
+ for (int i = 0; i < aSize; i++) {
+ mData[i] -= aTemp1;
+ mData[i] *= aTemp2;
+ }
+}
+
+template <class T>
+void CTensor<T>::rgbToCielab() {
+ for (int y = 0; y < mYSize; y++)
+ for (int x = 0; x < mXSize; x++) {
+ float R = operator()(x,y,0)*0.003921569;
+ float G = operator()(x,y,1)*0.003921569;
+ float B = operator()(x,y,2)*0.003921569;
+ if (R>0.0031308) R = pow((R + 0.055)*0.9478673, 2.4); else R *= 0.077399381;
+ if (G>0.0031308) G = pow((G + 0.055)*0.9478673, 2.4); else G *= 0.077399381;
+ if (B>0.0031308) B = pow((B + 0.055)*0.9478673, 2.4); else B *= 0.077399381;
+ //Observer. = 2?, Illuminant = D65
+ float X = R * 0.4124 + G * 0.3576 + B * 0.1805;
+ float Y = R * 0.2126 + G * 0.7152 + B * 0.0722;
+ float Z = R * 0.0193 + G * 0.1192 + B * 0.9505;
+ X *= 1.052111;
+ Z *= 0.918417;
+ if (X > 0.008856) X = pow(X,0.33333333333); else X = 7.787*X + 0.137931034;
+ if (Y > 0.008856) Y = pow(Y,0.33333333333); else Y = 7.787*Y + 0.137931034;
+ if (Z > 0.008856) Z = pow(Z,0.33333333333); else Z = 7.787*Z + 0.137931034;
+ operator()(x,y,0) = 1000.0*((295.8*Y) - 40.8)/255.0;
+ operator()(x,y,1) = 128.0+637.5*(X-Y);
+ operator()(x,y,2) = 128.0+255.0*(Y-Z);
+ }
+}
+
+template <class T>
+void CTensor<T>::cielabToRGB() {
+ for (int y = 0; y < mYSize; y++)
+ for (int x = 0; x < mXSize; x++) {
+ float L = operator()(x,y,0)*0.255;
+ float A = operator()(x,y,1);
+ float B = operator()(x,y,2);
+ float Y = (L+40.8)*0.00338066;
+ float X = (A-128.0+637.5*Y)*0.0015686;
+ float Z = (128.0+255.0*Y-B)*0.00392157;
+ float temp = Y*Y*Y;
+ if (temp > 0.008856) Y = temp;
+ else Y = (Y-0.137931034)*0.12842;
+ temp = X*X*X;
+ if (temp > 0.008856) X = temp;
+ else X = (X-0.137931034)*0.12842;
+ temp = Z*Z*Z;
+ if (temp > 0.008856) Z = temp;
+ else Z = (Z-0.137931034)*0.12842;
+ X *= 0.95047;
+ Y *= 1.0;
+ Z *= 1.08883;
+ float r = 3.2406*X-1.5372*Y-0.4986*Z;
+ float g = -0.9689*X+1.8758*Y+0.0415*Z;
+ float b = 0.0557*X-0.204*Y+1.057*Z;
+ if (r < 0) r = 0;
+ temp = 1.055*pow(r,0.41667)-0.055;
+ if (temp > 0.0031308) r = temp;
+ else r *= 12.92;
+ if (g < 0) g = 0;
+ temp = 1.055*pow(g,0.41667)-0.055;
+ if (temp > 0.0031308) g = temp;
+ else g *= 12.92;
+ if (b < 0) b = 0;
+ temp = 1.055*pow(b,0.41667)-0.055;
+ if (temp > 0.0031308) b = temp;
+ else b *= 12.92;
+ operator()(x,y,0) = 255.0*r;
+ operator()(x,y,1) = 255.0*g;
+ operator()(x,y,2) = 255.0*b;
+ }
+}
+
+// applySimilarityTransform
+template <class T>
+void CTensor<T>::applySimilarityTransform(CTensor<T>& aWarped, CMatrix<bool>& aOutside, float tx, float ty, float cx, float cy, float phi, float scale) {
+ float cosphi = scale*cos(phi);
+ float sinphi = scale*sin(phi);
+ int aSize = mXSize*mYSize;
+ int aWarpedSize = aWarped.xSize()*aWarped.ySize();
+ float ctx = cx+tx-cx*cosphi+cy*sinphi;
+ float cty = cy+ty-cy*cosphi-cx*sinphi;
+ aOutside = false;
+ int i = 0;
+ for (int y = 0; y < aWarped.ySize(); y++)
+ for (int x = 0; x < aWarped.xSize(); x++,i++) {
+ float xf = x; float yf = y;
+ float ax = xf*cosphi-yf*sinphi+ctx;
+ float ay = yf*cosphi+xf*sinphi+cty;
+ int x1 = (int)ax; int y1 = (int)ay;
+ float alphaX = ax-x1; float alphaY = ay-y1;
+ float betaX = 1.0-alphaX; float betaY = 1.0-alphaY;
+ if (x1 < 0 || y1 < 0 || x1+1 >= mXSize || y1+1 >= mYSize) aOutside.data()[i] = true;
+ else {
+ int j = y1*mXSize+x1;
+ for (int k = 0; k < mZSize; k++) {
+ float a = betaX*mData[j] +alphaX*mData[j+1];
+ float b = betaX*mData[j+mXSize]+alphaX*mData[j+1+mXSize];
+ aWarped.data()[i+k*aWarpedSize] = betaY*a+alphaY*b;
+ j += aSize;
+ }
+ }
+ }
+}
+
+// applyHomography
+template <class T>
+void CTensor<T>::applyHomography(CTensor<T>& aWarped, CMatrix<bool>& aOutside, const CMatrix<float>& H) {
+ int aSize = mXSize*mYSize;
+ int aWarpedSize = aWarped.xSize()*aWarped.ySize();
+ aOutside = false;
+ int i = 0;
+ for (int y = 0; y < aWarped.ySize(); y++)
+ for (int x = 0; x < aWarped.xSize(); x++,i++) {
+ float xf = x; float yf = y;
+ float ax = H.data()[0]*xf+H.data()[1]*yf+H.data()[2];
+ float ay = H.data()[3]*xf+H.data()[4]*yf+H.data()[5];
+ float az = H.data()[6]*xf+H.data()[7]*yf+H.data()[8];
+ float invaz = 1.0/az;
+ ax *= invaz; ay *= invaz;
+ int x1 = (int)ax; int y1 = (int)ay;
+ float alphaX = ax-x1; float alphaY = ay-y1;
+ float betaX = 1.0-alphaX; float betaY = 1.0-alphaY;
+ if (x1 < 0 || y1 < 0 || x1+1 >= mXSize || y1+1 >= mYSize) aOutside.data()[i] = true;
+ else {
+ int j = y1*mXSize+x1;
+ for (int k = 0; k < mZSize; k++) {
+ float a = betaX*mData[j] +alphaX*mData[j+1];
+ float b = betaX*mData[j+mXSize]+alphaX*mData[j+1+mXSize];
+ aWarped.data()[i+k*aWarpedSize] = betaY*a+alphaY*b;
+ j += aSize;
+ }
+ }
+ }
+}
+
+// -----------------------------------------------------------------------------
+// File I/O
+// -----------------------------------------------------------------------------
+
+// readFromMathematicaFile
+template <class T>
+void CTensor<T>::readFromMathematicaFile(const char* aFilename) {
+ using namespace std;
+ // Read the whole file and store data in aData
+ // Ignore blanks, tabs and lines
+ // Also ignore Mathematica comments (* ... *)
+ ifstream aStream(aFilename);
+ string aData;
+ char aChar;
+ bool aBracketFound = false;
+ bool aStarFound = false;
+ bool aCommentFound = false;
+ while (aStream.get(aChar))
+ if (aChar != ' ' && aChar != '\t' && aChar != '\n') {
+ if (aCommentFound) {
+ if (!aStarFound && aChar == '*') aStarFound = true;
+ else {
+ if (aStarFound && aChar == ')') aCommentFound = false;
+ aStarFound = false;
+ }
+ }
+ else {
+ if (!aBracketFound && aChar == '(') aBracketFound = true;
+ else {
+ if (aBracketFound && aChar == '*') aCommentFound = true;
+ else aData += aChar;
+ aBracketFound = false;
+ }
+ }
+ }
+ // Count the number of braces and double braces to figure out z- and y-Size of tensor
+ int aDoubleBraceCount = 0;
+ int aBraceCount = 0;
+ int aPos = 0;
+ while ((aPos = aData.find_first_of('{',aPos)+1) > 0) {
+ aBraceCount++;
+ if (aData[aPos] == '{' && aData[aPos+1] != '{') aDoubleBraceCount++;
+ }
+ // Count the number of commas in the first section to figure out xSize of tensor
+ int aCommaCount = 0;
+ aPos = 0;
+ while (aData[aPos] != '}') {
+ if (aData[aPos] == ',') aCommaCount++;
+ aPos++;
+ }
+ // Adapt size of tensor
+ if (mData != 0) delete[] mData;
+ mXSize = aCommaCount+1;
+ mYSize = (aBraceCount-1-aDoubleBraceCount) / aDoubleBraceCount;
+ mZSize = aDoubleBraceCount;
+ mData = new T[mXSize*mYSize*mZSize];
+ // Analyse file ---------------
+ aPos = 0;
+ if (aData[aPos] != '{') throw EInvalidFileFormat("Mathematica");
+ aPos++;
+ for (int z = 0; z < mZSize; z++) {
+ if (aData[aPos] != '{') throw EInvalidFileFormat("Mathematica");
+ aPos++;
+ for (int y = 0; y < mYSize; y++) {
+ if (aData[aPos] != '{') throw EInvalidFileFormat("Mathematica");
+ aPos++;
+ for (int x = 0; x < mXSize; x++) {
+ int oldPos = aPos;
+ if (x+1 < mXSize) aPos = aData.find_first_of(',',aPos);
+ else aPos = aData.find_first_of('}',aPos);
+ #ifdef GNU_COMPILER
+ string s = aData.substr(oldPos,aPos-oldPos);
+ istrstream is(s.c_str());
+ #else
+ string s = aData.substr(oldPos,aPos-oldPos);
+ istringstream is(s);
+ #endif
+ T aItem;
+ is >> aItem;
+ operator()(x,y,z) = aItem;
+ aPos++;
+ }
+ if (y+1 < mYSize) {
+ if (aData[aPos] != ',') throw EInvalidFileFormat("Mathematica");
+ aPos++;
+ while (aData[aPos] != '{')
+ aPos++;
+ }
+ }
+ aPos++;
+ if (z+1 < mZSize) {
+ if (aData[aPos] != ',') throw EInvalidFileFormat("Mathematica");
+ aPos++;
+ while (aData[aPos] != '{')
+ aPos++;
+ }
+ }
+}
+
+// writeToMathematicaFile
+template <class T>
+void CTensor<T>::writeToMathematicaFile(const char* aFilename) {
+ using namespace std;
+ ofstream aStream(aFilename);
+ aStream << '{';
+ for (int z = 0; z < mZSize; z++) {
+ aStream << '{';
+ for (int y = 0; y < mYSize; y++) {
+ aStream << '{';
+ for (int x = 0; x < mXSize; x++) {
+ aStream << operator()(x,y,z);
+ if (x+1 < mXSize) aStream << ',';
+ }
+ aStream << '}';
+ if (y+1 < mYSize) aStream << ",\n";
+ }
+ aStream << '}';
+ if (z+1 < mZSize) aStream << ",\n";
+ }
+ aStream << '}';
+}
+
+// readFromIMFile
+template <class T>
+void CTensor<T>::readFromIMFile(const char* aFilename) {
+ FILE *aStream;
+ aStream = fopen(aFilename,"rb");
+ // Read image data
+ for (int i = 0; i < mXSize*mYSize*mZSize; i++)
+ mData[i] = getc(aStream);
+ fclose(aStream);
+}
+
+// writeToIMFile
+template <class T>
+void CTensor<T>::writeToIMFile(const char *aFilename) {
+ FILE *aStream;
+ aStream = fopen(aFilename,"wb");
+ // write data
+ for (int i = 0; i < mXSize*mYSize*mZSize; i++) {
+ char dummy = (char)mData[i];
+ fwrite(&dummy,1,1,aStream);
+ }
+ fclose(aStream);
+}
+
+// readFromPGM
+template <class T>
+void CTensor<T>::readFromPGM(const char* aFilename) {
+ FILE *aStream;
+ aStream = fopen(aFilename,"rb");
+ if (aStream == 0) std::cerr << "File not found: " << aFilename << std::endl;
+ int dummy;
+ // Find beginning of file (P5)
+ while (getc(aStream) != 'P');
+ if (getc(aStream) != '5') throw EInvalidFileFormat("PGM");
+ do
+ dummy = getc(aStream);
+ while (dummy != '\n' && dummy != ' ');
+ // Remove comments and empty lines
+ dummy = getc(aStream);
+ while (dummy == '#') {
+ while (getc(aStream) != '\n');
+ dummy = getc(aStream);
+ }
+ while (dummy == '\n')
+ dummy = getc(aStream);
+ // Read image size
+ mXSize = dummy-48;
+ while ((dummy = getc(aStream)) >= 48 && dummy < 58)
+ mXSize = 10*mXSize+dummy-48;
+ while ((dummy = getc(aStream)) < 48 || dummy >= 58);
+ mYSize = dummy-48;
+ while ((dummy = getc(aStream)) >= 48 && dummy < 58)
+ mYSize = 10*mYSize+dummy-48;
+ mZSize = 1;
+ while (dummy != '\n' && dummy != ' ')
+ dummy = getc(aStream);
+ while (dummy != '\n' && dummy != ' ')
+ dummy = getc(aStream);
+ // Adjust size of data structure
+ delete[] mData;
+ mData = new T[mXSize*mYSize];
+ // Read image data
+ for (int i = 0; i < mXSize*mYSize; i++)
+ mData[i] = getc(aStream);
+ fclose(aStream);
+}
+
+// writeToPGM
+template <class T>
+void CTensor<T>::writeToPGM(const char* aFilename) {
+ int rows = (int)floor(sqrt(mZSize));
+ int cols = (int)ceil(mZSize*1.0/rows);
+ FILE* outimage = fopen(aFilename, "wb");
+ fprintf(outimage, "P5 \n");
+ fprintf(outimage, "%ld %ld \n255\n", cols*mXSize,rows*mYSize);
+ for (int r = 0; r < rows; r++)
+ for (int y = 0; y < mYSize; y++)
+ for (int c = 0; c < cols; c++)
+ for (int x = 0; x < mXSize; x++) {
+ unsigned char aHelp;
+ if (r*cols+c >= mZSize) aHelp = 0;
+ else aHelp = (unsigned char)operator()(x,y,r*cols+c);
+ fwrite (&aHelp, sizeof(unsigned char), 1, outimage);
+ }
+ fclose(outimage);
+}
+
+// makeColorTensor
+template <class T>
+void CTensor<T>::makeColorTensor() {
+ if (mZSize != 1) return;
+ int aSize = mXSize*mYSize;
+ int a2Size = 2*aSize;
+ T* aNewData = new T[aSize*3];
+ for (int i = 0; i < aSize; i++)
+ aNewData[i] = aNewData[i+aSize] = aNewData[i+a2Size] = mData[i];
+ mZSize = 3;
+ delete[] mData;
+ mData = aNewData;
+}
+
+// readFromPPM
+template <class T>
+void CTensor<T>::readFromPPM(const char* aFilename) {
+ FILE *aStream;
+ aStream = fopen(aFilename,"rb");
+ if (aStream == 0)
+ std::cerr << "File not found: " << aFilename << std::endl;
+ int dummy;
+ // Find beginning of file (P6)
+ while (getc(aStream) != 'P');
+ dummy = getc(aStream);
+ if (dummy == '5') mZSize = 1;
+ else if (dummy == '6') mZSize = 3;
+ else throw EInvalidFileFormat("PPM");
+ do dummy = getc(aStream); while (dummy != '\n' && dummy != ' ');
+ // Remove comments and empty lines
+ dummy = getc(aStream);
+ while (dummy == '#') {
+ while (getc(aStream) != '\n');
+ dummy = getc(aStream);
+ }
+ while (dummy == '\n')
+ dummy = getc(aStream);
+ // Read image size
+ mXSize = dummy-48;
+ while ((dummy = getc(aStream)) >= 48 && dummy < 58)
+ mXSize = 10*mXSize+dummy-48;
+ while ((dummy = getc(aStream)) < 48 || dummy >= 58);
+ mYSize = dummy-48;
+ while ((dummy = getc(aStream)) >= 48 && dummy < 58)
+ mYSize = 10*mYSize+dummy-48;
+ while (dummy != '\n' && dummy != ' ')
+ dummy = getc(aStream);
+ while (dummy < 48 || dummy >= 58) dummy = getc(aStream);
+ while ((dummy = getc(aStream)) >= 48 && dummy < 58);
+ if (dummy != '\n') while (getc(aStream) != '\n');
+ // Adjust size of data structure
+ delete[] mData;
+ mData = new T[mXSize*mYSize*mZSize];
+ // Read image data
+ int aSize = mXSize*mYSize;
+ if (mZSize == 1)
+ for (int i = 0; i < aSize; i++)
+ mData[i] = getc(aStream);
+ else {
+ int aSizeTwice = aSize+aSize;
+ for (int i = 0; i < aSize; i++) {
+ mData[i] = getc(aStream);
+ mData[i+aSize] = getc(aStream);
+ mData[i+aSizeTwice] = getc(aStream);
+ }
+ }
+ fclose(aStream);
+}
+
+// writeToPPM
+template <class T>
+void CTensor<T>::writeToPPM(const char* aFilename) {
+ FILE* outimage = fopen(aFilename, "wb");
+ fprintf(outimage, "P6 \n");
+ fprintf(outimage, "%d %d \n255\n", mXSize,mYSize);
+ for (int y = 0; y < mYSize; y++)
+ for (int x = 0; x < mXSize; x++) {
+ unsigned char aHelp = (unsigned char)operator()(x,y,0);
+ fwrite (&aHelp, sizeof(unsigned char), 1, outimage);
+ aHelp = (unsigned char)operator()(x,y,1);
+ fwrite (&aHelp, sizeof(unsigned char), 1, outimage);
+ aHelp = (unsigned char)operator()(x,y,2);
+ fwrite (&aHelp, sizeof(unsigned char), 1, outimage);
+ }
+ fclose(outimage);
+}
+
+// readFromPDM
+template <class T>
+void CTensor<T>::readFromPDM(const char* aFilename) {
+ std::ifstream aStream(aFilename);
+ std::string s;
+ // Read header
+ aStream >> s;
+ if (s != "P9") throw EInvalidFileFormat("PDM");
+ char aFeatureType;
+ aStream >> aFeatureType;
+ aStream >> s;
+ aStream >> mXSize;
+ aStream >> mYSize;
+ aStream >> mZSize;
+ aStream >> s;
+ // Adjust size of data structure
+ delete[] mData;
+ mData = new T[mXSize*mYSize*mZSize];
+ // Read data
+ for (int i = 0; i < mXSize*mYSize*mZSize; i++)
+ aStream >> mData[i];
+}
+
+// writeToPDM
+template <class T>
+void CTensor<T>::writeToPDM(const char* aFilename, char aFeatureType) {
+ std::ofstream aStream(aFilename);
+ // write header
+ aStream << "P9" << std::endl;
+ aStream << aFeatureType << "SS" << std::endl;
+ aStream << mZSize << ' ' << mYSize << ' ' << mXSize << std::endl;
+ aStream << "F" << std::endl;
+ // write data
+ for (int i = 0; i < mXSize*mYSize*mZSize; i++) {
+ aStream << mData[i];
+ if (i % 8 == 0) aStream << std::endl;
+ else aStream << ' ';
+ }
+}
+
+// operator ()
+template <class T>
+inline T& CTensor<T>::operator()(const int ax, const int ay, const int az) const {
+ #ifdef _DEBUG
+ if (ax >= mXSize || ay >= mYSize || az >= mZSize || ax < 0 || ay < 0 || az < 0)
+ throw ETensorRangeOverflow(ax,ay,az);
+ #endif
+ return mData[mXSize*(mYSize*az+ay)+ax];
+}
+
+template <class T>
+CVector<T> CTensor<T>::operator()(const float ax, const float ay) const {
+ CVector<T> aResult(mZSize);
+ int x1 = (int)ax;
+ int y1 = (int)ay;
+ int x2 = x1+1;
+ int y2 = y1+1;
+ #ifdef _DEBUG
+ if (x2 >= mXSize || y2 >= mYSize || x1 < 0 || y1 < 0) throw ETensorRangeOverflow(ax,ay,0);
+ #endif
+ float alphaX = ax-x1; float alphaXTrans = 1.0-alphaX;
+ float alphaY = ay-y1; float alphaYTrans = 1.0-alphaY;
+ for (int k = 0; k < mZSize; k++) {
+ float a = alphaXTrans*operator()(x1,y1,k)+alphaX*operator()(x2,y1,k);
+ float b = alphaXTrans*operator()(x1,y2,k)+alphaX*operator()(x2,y2,k);
+ aResult(k) = alphaYTrans*a+alphaY*b;
+ }
+ return aResult;
+}
+
+// operator =
+template <class T>
+inline CTensor<T>& CTensor<T>::operator=(const T aValue) {
+ fill(aValue);
+ return *this;
+}
+
+template <class T>
+CTensor<T>& CTensor<T>::operator=(const CTensor<T>& aCopyFrom) {
+ if (this != &aCopyFrom) {
+ delete[] mData;
+ if (aCopyFrom.mData == 0) {
+ mData = 0; mXSize = 0; mYSize = 0; mZSize = 0;
+ }
+ else {
+ mXSize = aCopyFrom.mXSize;
+ mYSize = aCopyFrom.mYSize;
+ mZSize = aCopyFrom.mZSize;
+ int wholeSize = mXSize*mYSize*mZSize;
+ mData = new T[wholeSize];
+ for (register int i = 0; i < wholeSize; i++)
+ mData[i] = aCopyFrom.mData[i];
+ }
+ }
+ return *this;
+}
+
+// operator +=
+template <class T>
+CTensor<T>& CTensor<T>::operator+=(const CTensor<T>& aTensor) {
+ #ifdef _DEBUG
+ if (mXSize != aTensor.mXSize || mYSize != aTensor.mYSize || mZSize != aTensor.mZSize)
+ throw ETensorIncompatibleSize(mXSize,mYSize,mZSize);
+ #endif
+ int wholeSize = size();
+ for (int i = 0; i < wholeSize; i++)
+ mData[i] += aTensor.mData[i];
+ return *this;
+}
+
+// operator +=
+template <class T>
+CTensor<T>& CTensor<T>::operator+=(const T aValue) {
+ int wholeSize = mXSize*mYSize*mZSize;
+ for (int i = 0; i < wholeSize; i++)
+ mData[i] += aValue;
+ return *this;
+}
+
+// operator *=
+template <class T>
+CTensor<T>& CTensor<T>::operator*=(const T aValue) {
+ int wholeSize = mXSize*mYSize*mZSize;
+ for (int i = 0; i < wholeSize; i++)
+ mData[i] *= aValue;
+ return *this;
+}
+
+// min
+template <class T>
+T CTensor<T>::min() const {
+ T aMin = mData[0];
+ int aSize = mXSize*mYSize*mZSize;
+ for (int i = 1; i < aSize; i++)
+ if (mData[i] < aMin) aMin = mData[i];
+ return aMin;
+}
+
+// max
+template <class T>
+T CTensor<T>::max() const {
+ T aMax = mData[0];
+ int aSize = mXSize*mYSize*mZSize;
+ for (int i = 1; i < aSize; i++)
+ if (mData[i] > aMax) aMax = mData[i];
+ return aMax;
+}
+
+// avg
+template <class T>
+T CTensor<T>::avg() const {
+ T aAvg = 0;
+ for (int z = 0; z < mZSize; z++)
+ aAvg += avg(z);
+ return aAvg/mZSize;
+}
+
+template <class T>
+T CTensor<T>::avg(int az) const {
+ T aAvg = 0;
+ int aSize = mXSize*mYSize;
+ int aTemp = (az+1)*aSize;
+ for (int i = az*aSize; i < aTemp; i++)
+ aAvg += mData[i];
+ return aAvg/aSize;
+}
+
+// xSize
+template <class T>
+inline int CTensor<T>::xSize() const {
+ return mXSize;
+}
+
+// ySize
+template <class T>
+inline int CTensor<T>::ySize() const {
+ return mYSize;
+}
+
+// zSize
+template <class T>
+inline int CTensor<T>::zSize() const {
+ return mZSize;
+}
+
+// size
+template <class T>
+inline int CTensor<T>::size() const {
+ return mXSize*mYSize*mZSize;
+}
+
+// getMatrix
+template <class T>
+CMatrix<T> CTensor<T>::getMatrix(const int az) const {
+ CMatrix<T> aTemp(mXSize,mYSize);
+ int aMatrixSize = mXSize*mYSize;
+ int aOffset = az*aMatrixSize;
+ for (int i = 0; i < aMatrixSize; i++)
+ aTemp.data()[i] = mData[i+aOffset];
+ return aTemp;
+}
+
+// getMatrix
+template <class T>
+void CTensor<T>::getMatrix(CMatrix<T>& aMatrix, const int az) const {
+ if (aMatrix.xSize() != mXSize || aMatrix.ySize() != mYSize)
+ throw ETensorIncompatibleSize(aMatrix.xSize(),aMatrix.ySize(),mXSize,mYSize);
+ int aMatrixSize = mXSize*mYSize;
+ int aOffset = az*aMatrixSize;
+ for (int i = 0; i < aMatrixSize; i++)
+ aMatrix.data()[i] = mData[i+aOffset];
+}
+
+// putMatrix
+template <class T>
+void CTensor<T>::putMatrix(CMatrix<T>& aMatrix, const int az) {
+ if (aMatrix.xSize() != mXSize || aMatrix.ySize() != mYSize)
+ throw ETensorIncompatibleSize(aMatrix.xSize(),aMatrix.ySize(),mXSize,mYSize);
+ int aMatrixSize = mXSize*mYSize;
+ int aOffset = az*aMatrixSize;
+ for (int i = 0; i < aMatrixSize; i++)
+ mData[i+aOffset] = aMatrix.data()[i];
+}
+
+// data()
+template <class T>
+inline T* CTensor<T>::data() const {
+ return mData;
+}
+
+// N O N - M E M B E R F U N C T I O N S --------------------------------------
+
+// operator <<
+template <class T>
+std::ostream& operator<<(std::ostream& aStream, const CTensor<T>& aTensor) {
+ for (int z = 0; z < aTensor.zSize(); z++) {
+ for (int y = 0; y < aTensor.ySize(); y++) {
+ for (int x = 0; x < aTensor.xSize(); x++)
+ aStream << aTensor(x,y,z) << ' ';
+ aStream << std::endl;
+ }
+ aStream << std::endl;
+ }
+ return aStream;
+}
+
+#endif
diff --git a/video_input/consistencyChecker/CTensor4D.h b/video_input/consistencyChecker/CTensor4D.h new file mode 100644 index 0000000..6feeb5d --- /dev/null +++ b/video_input/consistencyChecker/CTensor4D.h @@ -0,0 +1,656 @@ +// CTensor4D
+// A four-dimensional array
+//
+// Author: Thomas Brox
+// Last change: 05.11.2001
+//-------------------------------------------------------------------------
+// Note:
+// There is a difference between the GNU Compiler's STL and the standard
+// concerning the definition and usage of string streams as well as substrings.
+// Thus if using a GNU Compiler you should write #define GNU_COMPILER at the
+// beginning of your program.
+//
+// Another Note:
+// Linker problems occured in connection with <vector> from the STL.
+// In this case you should include this file in a namespace.
+// Example:
+// namespace NTensor4D {
+// #include <CTensor4D.h>
+// }
+// After including other packages you can then write:
+// using namespace NTensor4D;
+
+#ifndef CTENSOR4D_H
+#define CTENSOR4D_H
+
+#include <iostream>
+#include <fstream>
+#include <string>
+#ifdef GNU_COMPILER
+ #include <strstream>
+#else
+ #include <sstream>
+#endif
+#include "CTensor.h"
+
+template <class T>
+class CTensor4D {
+public:
+ // constructor
+ inline CTensor4D();
+ inline CTensor4D(const int aXSize, const int aYSize, const int aZSize, const int aASize);
+ // copy constructor
+ CTensor4D(const CTensor4D<T>& aCopyFrom);
+ // constructor with implicit filling
+ CTensor4D(const int aXSize, const int aYSize, const int aZSize, const int aASize, const T aFillValue);
+ // destructor
+ virtual ~CTensor4D();
+
+ // Changes the size of the tensor, data will be lost
+ void setSize(int aXSize, int aYSize, int aZSize, int aASize);
+ // Downsamples the tensor
+ void downsample(int aNewXSize, int aNewYSize);
+ void downsample(int aNewXSize, int aNewYSize, int aNewZSize);
+ // Upsamples the tensor
+ void upsample(int aNewXSize, int aNewYSize);
+ void upsampleBilinear(int aNewXSize, int aNewYSize);
+ void upsampleTrilinear(int aNewXSize, int aNewYSize, int aNewZSize);
+ // Fills the tensor with the value aValue (see also operator =)
+ void fill(const T aValue);
+ // Copies a box from the tensor into aResult, the size of aResult will be adjusted
+ void cut(CTensor4D<T>& aResult, int x1, int y1, int z1, int a1, int x2, int y2, int z2, int a2);
+ // Reads data from a list of PPM or PGM files given in a text file
+ void readFromFile(char* aFilename);
+ // Writes a set of colour images to a large PPM image
+ void writeToPPM(const char* aFilename, int aCols = 0, int aRows = 0);
+
+ // Gives full access to tensor's values
+ inline T& operator()(const int ax, const int ay, const int az, const int aa) const;
+ // Read access with bilinear interpolation
+ CVector<T> operator()(const float ax, const float ay, const int aa) const;
+ // Fills the tensor with the value aValue (equivalent to fill())
+ inline CTensor4D<T>& operator=(const T aValue);
+ // Copies the tensor aCopyFrom to this tensor (size of tensor might change)
+ CTensor4D<T>& operator=(const CTensor4D<T>& aCopyFrom);
+ // Multiplication with a scalar
+ CTensor4D<T>& operator*=(const T aValue);
+ // Component-wise addition
+ CTensor4D<T>& operator+=(const CTensor4D<T>& aTensor);
+
+ // Gives access to the tensor's size
+ inline int xSize() const;
+ inline int ySize() const;
+ inline int zSize() const;
+ inline int aSize() const;
+ inline int size() const;
+ // Returns the aath layer of the 4D-tensor as 3D-tensor
+ CTensor<T> getTensor3D(const int aa) const;
+ // Removes one dimension and returns the resulting 3D-tensor
+ void getTensor3D(CTensor<T>& aTensor, int aIndex, int aDim = 3) const;
+ // Copies the components of a 3D-tensor in the aDimth layer of the 4D-tensor
+ void putTensor3D(CTensor<T>& aTensor, int aIndex, int aDim = 3);
+ // Removes two dimensions and returns the resulting matrix
+ void getMatrix(CMatrix<T>& aMatrix, int aZIndex, int aAIndex) const;
+ // Copies the components of a 3D-tensor in the aDimth layer of the 4D-tensor
+ void putMatrix(CMatrix<T>& aMatrix, int aZIndex, int aAIndex);
+ // Gives access to the internal data representation (use sparingly)
+ inline T* data() const;
+protected:
+ int mXSize,mYSize,mZSize,mASize;
+ T *mData;
+};
+
+// Provides basic output functionality (only appropriate for very small tensors)
+template <class T> std::ostream& operator<<(std::ostream& aStream, const CTensor4D<T>& aTensor);
+
+// Exceptions thrown by CTensor-------------------------------------------------
+
+// Thrown when one tries to access an element of a tensor which is out of
+// the tensor's bounds
+struct ETensor4DRangeOverflow {
+ ETensor4DRangeOverflow(const int ax, const int ay, const int az, const int aa) {
+ using namespace std;
+ cerr << "Exception ETensor4DRangeOverflow: x = " << ax << ", y = " << ay << ", z = " << az << ", a = " << aa << endl;
+ }
+};
+
+// Thrown from getTensor3D if the parameter's size does not match with the size
+// of this tensor
+struct ETensor4DIncompatibleSize {
+ ETensor4DIncompatibleSize(int ax, int ay, int az, int ax2, int ay2, int az2) {
+ using namespace std;
+ cerr << "Exception ETensor4DIncompatibleSize: x = " << ax << ":" << ax2;
+ cerr << ", y = " << ay << ":" << ay2;
+ cerr << ", z = " << az << ":" << az2 << endl;
+ }
+};
+
+// Thrown from readFromFile if the file format is unknown
+struct ETensor4DInvalidFileFormat {
+ ETensor4DInvalidFileFormat() {
+ using namespace std;
+ cerr << "Exception ETensor4DInvalidFileFormat" << endl;
+ }
+};
+
+// I M P L E M E N T A T I O N --------------------------------------------
+//
+// You might wonder why there is implementation code in a header file.
+// The reason is that not all C++ compilers yet manage separate compilation
+// of templates. Inline functions cannot be compiled separately anyway.
+// So in this case the whole implementation code is added to the header
+// file.
+// Users of CTensor4D should ignore everything that's beyond this line :)
+// ------------------------------------------------------------------------
+
+// P U B L I C ------------------------------------------------------------
+
+// constructor
+template <class T>
+inline CTensor4D<T>::CTensor4D() {
+ mData = 0; mXSize = 0; mYSize = 0; mZSize = 0; mASize = 0;
+}
+
+// constructor
+template <class T>
+inline CTensor4D<T>::CTensor4D(const int aXSize, const int aYSize, const int aZSize, const int aASize)
+ : mXSize(aXSize), mYSize(aYSize), mZSize(aZSize), mASize(aASize) {
+ mData = new T[aXSize*aYSize*aZSize*aASize];
+}
+
+// copy constructor
+template <class T>
+CTensor4D<T>::CTensor4D(const CTensor4D<T>& aCopyFrom)
+ : mXSize(aCopyFrom.mXSize), mYSize(aCopyFrom.mYSize), mZSize(aCopyFrom.mZSize), mASize(aCopyFrom.mASize) {
+ int wholeSize = mXSize*mYSize*mZSize*mASize;
+ mData = new T[wholeSize];
+ for (register int i = 0; i < wholeSize; i++)
+ mData[i] = aCopyFrom.mData[i];
+}
+
+// constructor with implicit filling
+template <class T>
+CTensor4D<T>::CTensor4D(const int aXSize, const int aYSize, const int aZSize, const int aASize, const T aFillValue)
+ : mXSize(aXSize), mYSize(aYSize), mZSize(aZSize), mASize(aASize) {
+ mData = new T[aXSize*aYSize*aZSize*aASize];
+ fill(aFillValue);
+}
+
+// destructor
+template <class T>
+CTensor4D<T>::~CTensor4D() {
+ delete[] mData;
+}
+
+// setSize
+template <class T>
+void CTensor4D<T>::setSize(int aXSize, int aYSize, int aZSize, int aASize) {
+ if (mData != 0) delete[] mData;
+ mData = new T[aXSize*aYSize*aZSize*aASize];
+ mXSize = aXSize;
+ mYSize = aYSize;
+ mZSize = aZSize;
+ mASize = aASize;
+}
+
+//downsample
+template <class T>
+void CTensor4D<T>::downsample(int aNewXSize, int aNewYSize) {
+ T* mData2 = new T[aNewXSize*aNewYSize*mZSize*mASize];
+ int aSize = aNewXSize*aNewYSize;
+ for (int a = 0; a < mASize; a++)
+ for (int z = 0; z < mZSize; z++) {
+ CMatrix<T> aTemp(mXSize,mYSize);
+ getMatrix(aTemp,z,a);
+ aTemp.downsample(aNewXSize,aNewYSize);
+ for (int i = 0; i < aSize; i++)
+ mData2[i+(a*mZSize+z)*aSize] = aTemp.data()[i];
+ }
+ delete[] mData;
+ mData = mData2;
+ mXSize = aNewXSize;
+ mYSize = aNewYSize;
+}
+
+template <class T>
+void CTensor4D<T>::downsample(int aNewXSize, int aNewYSize, int aNewZSize) {
+ T* mData2 = new T[aNewXSize*aNewYSize*aNewZSize*mASize];
+ int aSize = aNewXSize*aNewYSize*aNewZSize;
+ for (int a = 0; a < mASize; a++) {
+ CTensor<T> aTemp(mXSize,mYSize,mZSize);
+ getTensor3D(aTemp,a);
+ aTemp.downsample(aNewXSize,aNewYSize,aNewZSize);
+ for (int i = 0; i < aSize; i++)
+ mData2[i+a*aSize] = aTemp.data()[i];
+ }
+ delete[] mData;
+ mData = mData2;
+ mXSize = aNewXSize;
+ mYSize = aNewYSize;
+ mZSize = aNewZSize;
+}
+
+// upsample
+template <class T>
+void CTensor4D<T>::upsample(int aNewXSize, int aNewYSize) {
+ T* mData2 = new T[aNewXSize*aNewYSize*mZSize*mASize];
+ int aSize = aNewXSize*aNewYSize;
+ for (int a = 0; a < mASize; a++)
+ for (int z = 0; z < mZSize; z++) {
+ CMatrix<T> aTemp(mXSize,mYSize);
+ getMatrix(aTemp,z,a);
+ aTemp.upsample(aNewXSize,aNewYSize);
+ for (int i = 0; i < aSize; i++)
+ mData2[i+(a*mZSize+z)*aSize] = aTemp.data()[i];
+ }
+ delete[] mData;
+ mData = mData2;
+ mXSize = aNewXSize;
+ mYSize = aNewYSize;
+}
+
+// upsampleBilinear
+template <class T>
+void CTensor4D<T>::upsampleBilinear(int aNewXSize, int aNewYSize) {
+ T* mData2 = new T[aNewXSize*aNewYSize*mZSize*mASize];
+ int aSize = aNewXSize*aNewYSize;
+ for (int a = 0; a < mASize; a++)
+ for (int z = 0; z < mZSize; z++) {
+ CMatrix<T> aTemp(mXSize,mYSize);
+ getMatrix(aTemp,z,a);
+ aTemp.upsampleBilinear(aNewXSize,aNewYSize);
+ for (int i = 0; i < aSize; i++)
+ mData2[i+(a*mZSize+z)*aSize] = aTemp.data()[i];
+ }
+ delete[] mData;
+ mData = mData2;
+ mXSize = aNewXSize;
+ mYSize = aNewYSize;
+}
+
+// upsampleTrilinear
+template <class T>
+void CTensor4D<T>::upsampleTrilinear(int aNewXSize, int aNewYSize, int aNewZSize) {
+ T* mData2 = new T[aNewXSize*aNewYSize*aNewZSize*mASize];
+ int aSize = aNewXSize*aNewYSize*aNewZSize;
+ for (int a = 0; a < mASize; a++) {
+ CTensor<T> aTemp(mXSize,mYSize,mZSize);
+ getTensor3D(aTemp,a);
+ aTemp.upsampleTrilinear(aNewXSize,aNewYSize,aNewZSize);
+ for (int i = 0; i < aSize; i++)
+ mData2[i+a*aSize] = aTemp.data()[i];
+ }
+ delete[] mData;
+ mData = mData2;
+ mXSize = aNewXSize;
+ mYSize = aNewYSize;
+ mZSize = aNewZSize;
+}
+
+// fill
+template <class T>
+void CTensor4D<T>::fill(const T aValue) {
+ int wholeSize = mXSize*mYSize*mZSize*mASize;
+ for (register int i = 0; i < wholeSize; i++)
+ mData[i] = aValue;
+}
+
+// cut
+template <class T>
+void CTensor4D<T>::cut(CTensor4D<T>& aResult, int x1, int y1, int z1, int a1, int x2, int y2, int z2, int a2) {
+ aResult.mXSize = x2-x1+1;
+ aResult.mYSize = y2-y1+1;
+ aResult.mZSize = z2-z1+1;
+ aResult.mASize = a2-a1+1;
+ delete[] aResult.mData;
+ aResult.mData = new T[aResult.mXSize*aResult.mYSize*aResult.mZSize*aResult.mASize];
+ for (int a = a1; a <= a2; a++)
+ for (int z = z1; z <= z2; z++)
+ for (int y = y1; y <= y2; y++)
+ for (int x = x1; x <= x2; x++)
+ aResult(x-x1,y-y1,z-z1,a-a1) = operator()(x,y,z,a);
+}
+
+// readFromFile
+template <class T>
+void CTensor4D<T>::readFromFile(char* aFilename) {
+ if (mData != 0) delete[] mData;
+ std::string s;
+ std::string aPath = aFilename;
+ aPath.erase(aPath.find_last_of('\\')+1,100);
+ mASize = 0;
+ {
+ std::ifstream aStream(aFilename);
+ while (!aStream.eof()) {
+ aStream >> s;
+ if (s != "") {
+ mASize++;
+ if (mASize == 1) {
+ s.erase(0,s.find_last_of('.'));
+ if (s == ".ppm" || s == ".PPM") mZSize = 3;
+ else if (s == ".pgm" || s == ".PGM") mZSize = 1;
+ else throw ETensor4DInvalidFileFormat();
+ }
+ }
+ }
+ }
+ std::ifstream aStream(aFilename);
+ aStream >> s;
+ s = aPath+s;
+ // PGM
+ if (mZSize == 1) {
+ CMatrix<float> aTemp;
+ aTemp.readFromPGM(s.c_str());
+ mXSize = aTemp.xSize();
+ mYSize = aTemp.ySize();
+ int aSize = mXSize*mYSize;
+ mData = new T[aSize*mASize];
+ for (int i = 0; i < aSize; i++)
+ mData[i] = aTemp.data()[i];
+ for (int a = 1; a < mASize; a++) {
+ aStream >> s;
+ s = aPath+s;
+ aTemp.readFromPGM(s.c_str());
+ for (int i = 0; i < aSize; i++)
+ mData[i+a*aSize] = aTemp.data()[i];
+ }
+ }
+ // PPM
+ else {
+ CTensor<float> aTemp;
+ aTemp.readFromPPM(s.c_str());
+ mXSize = aTemp.xSize();
+ mYSize = aTemp.ySize();
+ int aSize = 3*mXSize*mYSize;
+ mData = new T[aSize*mASize];
+ for (int i = 0; i < aSize; i++)
+ mData[i] = aTemp.data()[i];
+ for (int a = 1; a < mASize; a++) {
+ aStream >> s;
+ s = aPath+s;
+ aTemp.readFromPPM(s.c_str());
+ for (int i = 0; i < aSize; i++)
+ mData[i+a*aSize] = aTemp.data()[i];
+ }
+ }
+}
+
+// writeToPPM
+template <class T>
+void CTensor4D<T>::writeToPPM(const char* aFilename, int aCols, int aRows) {
+ int rows = (int)floor(sqrt(mASize));
+ if (aRows != 0) rows = aRows;
+ int cols = (int)ceil(mASize*1.0/rows);
+ if (aCols != 0) cols = aCols;
+ FILE* outimage = fopen(aFilename, "wb");
+ fprintf(outimage, "P6 \n");
+ fprintf(outimage, "%ld %ld \n255\n", cols*mXSize,rows*mYSize);
+ for (int r = 0; r < rows; r++)
+ for (int y = 0; y < mYSize; y++)
+ for (int c = 0; c < cols; c++)
+ for (int x = 0; x < mXSize; x++) {
+ unsigned char aHelp;
+ if (r*cols+c >= mASize) aHelp = 0;
+ else aHelp = (unsigned char)operator()(x,y,0,r*cols+c);
+ fwrite (&aHelp, sizeof(unsigned char), 1, outimage);
+ if (r*cols+c >= mASize) aHelp = 0;
+ else aHelp = (unsigned char)operator()(x,y,1,r*cols+c);
+ fwrite (&aHelp, sizeof(unsigned char), 1, outimage);
+ if (r*cols+c >= mASize) aHelp = 0;
+ else aHelp = (unsigned char)operator()(x,y,2,r*cols+c);
+ fwrite (&aHelp, sizeof(unsigned char), 1, outimage);
+ }
+ fclose(outimage);
+}
+
+// operator ()
+template <class T>
+inline T& CTensor4D<T>::operator()(const int ax, const int ay, const int az, const int aa) const {
+ #ifdef DEBUG
+ if (ax >= mXSize || ay >= mYSize || az >= mZSize || aa >= mASize || ax < 0 || ay < 0 || az < 0 || aa < 0)
+ throw ETensorRangeOverflow(ax,ay,az,aa);
+ #endif
+ return mData[mXSize*(mYSize*(mZSize*aa+az)+ay)+ax];
+}
+
+template <class T>
+CVector<T> CTensor4D<T>::operator()(const float ax, const float ay, const int aa) const {
+ CVector<T> aResult(mZSize);
+ int x1 = (int)ax;
+ int y1 = (int)ay;
+ int x2 = x1+1;
+ int y2 = y1+1;
+ #ifdef _DEBUG
+ if (x2 >= mXSize || y2 >= mYSize || x1 < 0 || y1 < 0) throw ETensorRangeOverflow(ax,ay,0);
+ #endif
+ float alphaX = ax-x1; float alphaXTrans = 1.0-alphaX;
+ float alphaY = ay-y1; float alphaYTrans = 1.0-alphaY;
+ for (int k = 0; k < mZSize; k++) {
+ float a = alphaXTrans*operator()(x1,y1,k,aa)+alphaX*operator()(x2,y1,k,aa);
+ float b = alphaXTrans*operator()(x1,y2,k,aa)+alphaX*operator()(x2,y2,k,aa);
+ aResult(k) = alphaYTrans*a+alphaY*b;
+ }
+ return aResult;
+}
+
+// operator =
+template <class T>
+inline CTensor4D<T>& CTensor4D<T>::operator=(const T aValue) {
+ fill(aValue);
+ return *this;
+}
+
+template <class T>
+CTensor4D<T>& CTensor4D<T>::operator=(const CTensor4D<T>& aCopyFrom) {
+ if (this != &aCopyFrom) {
+ if (mData != 0) delete[] mData;
+ mXSize = aCopyFrom.mXSize;
+ mYSize = aCopyFrom.mYSize;
+ mZSize = aCopyFrom.mZSize;
+ mASize = aCopyFrom.mASize;
+ int wholeSize = mXSize*mYSize*mZSize*mASize;
+ mData = new T[wholeSize];
+ for (register int i = 0; i < wholeSize; i++)
+ mData[i] = aCopyFrom.mData[i];
+ }
+ return *this;
+}
+
+// operator *=
+template <class T>
+CTensor4D<T>& CTensor4D<T>::operator*=(const T aValue) {
+ int wholeSize = mXSize*mYSize*mZSize*mASize;
+ for (int i = 0; i < wholeSize; i++)
+ mData[i] *= aValue;
+ return *this;
+}
+
+// operator +=
+template <class T>
+CTensor4D<T>& CTensor4D<T>::operator+=(const CTensor4D<T>& aTensor) {
+ #ifdef _DEBUG
+ if (mXSize != aTensor.mXSize || mYSize != aTensor.mYSize || mZSize != aTensor.mZSize || mASize != aTensor.mASize)
+ throw ETensorIncompatibleSize(mXSize,mYSize,mZSize);
+ #endif
+ int wholeSize = size();
+ for (int i = 0; i < wholeSize; i++)
+ mData[i] += aTensor.mData[i];
+ return *this;
+}
+
+// xSize
+template <class T>
+inline int CTensor4D<T>::xSize() const {
+
+ return mXSize;
+}
+
+// ySize
+template <class T>
+inline int CTensor4D<T>::ySize() const {
+ return mYSize;
+}
+
+// zSize
+template <class T>
+inline int CTensor4D<T>::zSize() const {
+ return mZSize;
+}
+
+// aSize
+template <class T>
+inline int CTensor4D<T>::aSize() const {
+ return mASize;
+}
+
+// size
+template <class T>
+inline int CTensor4D<T>::size() const {
+ return mXSize*mYSize*mZSize*mASize;
+}
+
+// getTensor3D
+template <class T>
+CTensor<T> CTensor4D<T>::getTensor3D(const int aa) const {
+ CTensor<T> aTemp(mXSize,mYSize,mZSize);
+ int aTensorSize = mXSize*mYSize*mZSize;
+ int aOffset = aa*aTensorSize;
+ for (int i = 0; i < aTensorSize; i++)
+ aTemp.data()[i] = mData[i+aOffset];
+ return aTemp;
+}
+
+// getTensor3D
+template <class T>
+void CTensor4D<T>::getTensor3D(CTensor<T>& aTensor, int aIndex, int aDim) const {
+ int aSize;
+ int aOffset;
+ switch (aDim) {
+ case 3:
+ if (aTensor.xSize() != mXSize || aTensor.ySize() != mYSize || aTensor.zSize() != mZSize)
+ throw ETensor4DIncompatibleSize(aTensor.xSize(),aTensor.ySize(),aTensor.zSize(),mXSize,mYSize,mZSize);
+ aSize = mXSize*mYSize*mZSize;
+ aOffset = aIndex*aSize;
+ for (int i = 0; i < aSize; i++)
+ aTensor.data()[i] = mData[i+aOffset];
+ break;
+ case 2:
+ if (aTensor.xSize() != mXSize || aTensor.ySize() != mYSize || aTensor.zSize() != mASize)
+ throw ETensor4DIncompatibleSize(aTensor.xSize(),aTensor.ySize(),aTensor.zSize(),mXSize,mYSize,mASize);
+ aSize = mXSize*mYSize;
+ aOffset = aIndex*aSize;
+ for (int a = 0; a < mASize; a++)
+ for (int i = 0; i < aSize; i++)
+ aTensor.data()[i+a*aSize] = mData[i+aOffset+a*aSize*mZSize];
+ break;
+ case 1:
+ if (aTensor.xSize() != mXSize || aTensor.ySize() != mZSize || aTensor.zSize() != mASize)
+ throw ETensor4DIncompatibleSize(aTensor.xSize(),aTensor.ySize(),aTensor.zSize(),mXSize,mZSize,mASize);
+ for (int a = 0; a < mASize; a++)
+ for (int z = 0; z < mZSize; z++)
+ for (int x = 0; x < mXSize; x++)
+ aTensor(x,z,a) = operator()(x,aIndex,z,a);
+ break;
+ case 0:
+ if (aTensor.xSize() != mYSize || aTensor.ySize() != mZSize || aTensor.zSize() != mASize)
+ throw ETensor4DIncompatibleSize(aTensor.xSize(),aTensor.ySize(),aTensor.zSize(),mYSize,mZSize,mASize);
+ for (int a = 0; a < mASize; a++)
+ for (int z = 0; z < mZSize; z++)
+ for (int y = 0; y < mYSize; y++)
+ aTensor(y,z,a) = operator()(aIndex,y,z,a);
+ break;
+ default: getTensor3D(aTensor,aIndex);
+ }
+}
+
+// putTensor3D
+template <class T>
+void CTensor4D<T>::putTensor3D(CTensor<T>& aTensor, int aIndex, int aDim) {
+ int aSize;
+ int aOffset;
+ switch (aDim) {
+ case 3:
+ if (aTensor.xSize() != mXSize || aTensor.ySize() != mYSize || aTensor.zSize() != mZSize)
+ throw ETensor4DIncompatibleSize(aTensor.xSize(),aTensor.ySize(),aTensor.zSize(),mXSize,mYSize,mZSize);
+ aSize = mXSize*mYSize*mZSize;
+ aOffset = aIndex*aSize;
+ for (int i = 0; i < aSize; i++)
+ mData[i+aOffset] = aTensor.data()[i];
+ break;
+ case 2:
+ if (aTensor.xSize() != mXSize || aTensor.ySize() != mYSize || aTensor.zSize() != mASize)
+ throw ETensor4DIncompatibleSize(aTensor.xSize(),aTensor.ySize(),aTensor.zSize(),mXSize,mYSize,mASize);
+ aSize = mXSize*mYSize;
+ aOffset = aIndex*aSize;
+ for (int a = 0; a < mASize; a++)
+ for (int i = 0; i < aSize; i++)
+ mData[i+aOffset+a*aSize*mZSize] = aTensor.data()[i+a*aSize];
+ break;
+ case 1:
+ if (aTensor.xSize() != mXSize || aTensor.ySize() != mZSize || aTensor.zSize() != mASize)
+ throw ETensor4DIncompatibleSize(aTensor.xSize(),aTensor.ySize(),aTensor.zSize(),mXSize,mZSize,mASize);
+ for (int a = 0; a < mASize; a++)
+ for (int z = 0; z < mZSize; z++)
+ for (int x = 0; x < mXSize; x++)
+ operator()(x,aIndex,z,a) = aTensor(x,z,a);
+ break;
+ case 0:
+ if (aTensor.xSize() != mYSize || aTensor.ySize() != mZSize || aTensor.zSize() != mASize)
+ throw ETensor4DIncompatibleSize(aTensor.xSize(),aTensor.ySize(),aTensor.zSize(),mYSize,mZSize,mASize);
+ for (int a = 0; a < mASize; a++)
+ for (int z = 0; z < mZSize; z++)
+ for (int y = 0; y < mYSize; y++)
+ operator()(aIndex,y,z,a) = aTensor(y,z,a);
+ break;
+ default: putTensor3D(aTensor,aIndex);
+ }
+}
+
+// getMatrix
+template <class T>
+void CTensor4D<T>::getMatrix(CMatrix<T>& aMatrix, int aZIndex, int aAIndex) const {
+ if (aMatrix.xSize() != mXSize || aMatrix.ySize() != mYSize)
+ throw ETensor4DIncompatibleSize(aMatrix.xSize(),aMatrix.ySize(),1,mXSize,mYSize,1);
+ int aSize = mXSize*mYSize;
+ int aOffset = aSize*(aAIndex*mZSize+aZIndex);
+ for (int i = 0; i < aSize; i++)
+ aMatrix.data()[i] = mData[i+aOffset];
+}
+
+// putMatrix
+template <class T>
+void CTensor4D<T>::putMatrix(CMatrix<T>& aMatrix, int aZIndex, int aAIndex) {
+ if (aMatrix.xSize() != mXSize || aMatrix.ySize() != mYSize)
+ throw ETensor4DIncompatibleSize(aMatrix.xSize(),aMatrix.ySize(),1,mXSize,mYSize,1);
+ int aSize = mXSize*mYSize;
+ int aOffset = aSize*(aAIndex*mZSize+aZIndex);
+ for (int i = 0; i < aSize; i++)
+ mData[i+aOffset] = aMatrix.data()[i];
+}
+
+// data()
+template <class T>
+inline T* CTensor4D<T>::data() const {
+ return mData;
+}
+
+// N O N - M E M B E R F U N C T I O N S --------------------------------------
+
+// operator <<
+template <class T>
+std::ostream& operator<<(std::ostream& aStream, const CTensor4D<T>& aTensor) {
+ for (int a = 0; a < aTensor.aSize(); a++) {
+ for (int z = 0; z < aTensor.zSize(); z++) {
+ for (int y = 0; y < aTensor.ySize(); y++) {
+ for (int x = 0; x < aTensor.xSize(); x++)
+ aStream << aTensor(x,y,z) << ' ';
+ aStream << std::endl;
+ }
+ aStream << std::endl;
+ }
+ aStream << std::endl;
+ }
+ return aStream;
+}
+
+#endif
diff --git a/video_input/consistencyChecker/CVector.h b/video_input/consistencyChecker/CVector.h new file mode 100644 index 0000000..aacb0fa --- /dev/null +++ b/video_input/consistencyChecker/CVector.h @@ -0,0 +1,519 @@ +// CVector
+// A one-dimensional array including basic vector operations
+//
+// Author: Thomas Brox
+// Last change: 23.05.2005
+//-------------------------------------------------------------------------
+#ifndef CVECTOR_H
+#define CVECTOR_H
+
+#include <iostream>
+#include <fstream>
+
+template <class T> class CMatrix;
+template <class T> class CTensor;
+
+template <class T>
+class CVector {
+public:
+ // constructor
+ inline CVector();
+ // constructor
+ inline CVector(const int aSize);
+ // copy constructor
+ CVector(const CVector<T>& aCopyFrom);
+ // constructor (from array)
+ CVector(const T* aPointer, const int aSize);
+ // constructor with implicit filling
+ CVector(const int aSize, const T aFillValue);
+ // destructor
+ virtual ~CVector();
+
+ // Changes the size of the vector (data is lost)
+ void setSize(int aSize);
+ // Fills the vector with the specified value (see also operator=)
+ void fill(const T aValue);
+ // Appends the values of another vector
+ void append(CVector<T>& aVector);
+ // Normalizes the length of the vector to 1
+ void normalize();
+ // Normalizes the component sum to 1
+ void normalizeSum();
+ // Reads values from a text file
+ void readFromTXT(const char* aFilename);
+ // Writes values to a text file
+ void writeToTXT(char* aFilename);
+ // Returns the sum of all values
+ T sum();
+ // Returns the minimum value
+ T min();
+ // Returns the maximum value
+ T max();
+ // Returns the Euclidean norm
+ T norm();
+
+ // Converts vector to homogeneous coordinates, i.e., all components are divided by last component
+ CVector<T>& homogen();
+ // Remove the last component
+ inline void homogen_nD();
+ // Computes the cross product between this vector and aVector
+ void cross(CVector<T>& aVector);
+
+ // Gives full access to the vector's values
+ inline T& operator()(const int aIndex) const;
+ inline T& operator[](const int aIndex) const;
+ // Fills the vector with the specified value (equivalent to fill)
+ inline CVector<T>& operator=(const T aValue);
+ // Copies a vector into this vector (size might change)
+ CVector<T>& operator=(const CVector<T>& aCopyFrom);
+ // Copies values from a matrix to the vector (size might change)
+ CVector<T>& operator=(const CMatrix<T>& aCopyFrom);
+ // Copies values from a tensor to the vector (size might change)
+ CVector<T>& operator=(const CTensor<T>& aCopyFrom);
+ // Adds another vector
+ CVector<T>& operator+=(const CVector<T>& aVector);
+ // Substracts another vector
+ CVector<T>& operator-=(const CVector<T>& aVector);
+ // Multiplies the vector with a scalar
+ CVector<T>& operator*=(const T aValue);
+ // Scalar product
+ T operator*=(const CVector<T>& aVector);
+ // Checks (non-)equivalence to another vector
+ bool operator==(const CVector<T>& aVector);
+ inline bool operator!=(const CVector<T>& aVector);
+
+ // Gives access to the vector's size
+ inline int size() const;
+ // Gives access to the internal data representation
+ inline T* data() const {return mData;}
+protected:
+ int mSize;
+ T* mData;
+};
+
+// Adds two vectors
+template <class T> CVector<T> operator+(const CVector<T>& vec1, const CVector<T>& vec2);
+// Substracts two vectors
+template <class T> CVector<T> operator-(const CVector<T>& vec1, const CVector<T>& vec2);
+// Multiplies vector with a scalar
+template <class T> CVector<T> operator*(const CVector<T>& aVector, const T aValue);
+template <class T> CVector<T> operator*(const T aValue, const CVector<T>& aVector);
+// Computes the scalar product of two vectors
+template <class T> T operator*(const CVector<T>& vec1, const CVector<T>& vec2);
+// Computes cross product of two vectors
+template <class T> CVector<T> operator/(const CVector<T>& vec1, const CVector<T>& vec2);
+// Sends the vector to an output stream
+template <class T> std::ostream& operator<<(std::ostream& aStream, const CVector<T>& aVector);
+
+// Exceptions thrown by CVector--------------------------------------------
+
+// Thrown if one tries to access an element of a vector which is out of
+// the vector's bounds
+struct EVectorRangeOverflow {
+ EVectorRangeOverflow(const int aIndex) {
+ using namespace std;
+ cerr << "Exception EVectorRangeOverflow: Index = " << aIndex << endl;
+ }
+};
+
+struct EVectorIncompatibleSize {
+ EVectorIncompatibleSize(int aSize1, int aSize2) {
+ using namespace std;
+ cerr << "Exception EVectorIncompatibleSize: " << aSize1 << " <> " << aSize2 << endl;
+ }
+};
+
+
+// I M P L E M E N T A T I O N --------------------------------------------
+//
+// You might wonder why there is implementation code in a header file.
+// The reason is that not all C++ compilers yet manage separate compilation
+// of templates. Inline functions cannot be compiled separately anyway.
+// So in this case the whole implementation code is added to the header
+// file.
+// Users of CVector should ignore everything that's beyond this line.
+// ------------------------------------------------------------------------
+
+// P U B L I C ------------------------------------------------------------
+// constructor
+template <class T>
+inline CVector<T>::CVector() : mSize(0) {
+ mData = new T[0];
+}
+
+// constructor
+template <class T>
+inline CVector<T>::CVector(const int aSize)
+ : mSize(aSize) {
+ mData = new T[aSize];
+}
+
+// copy constructor
+template <class T>
+CVector<T>::CVector(const CVector<T>& aCopyFrom)
+ : mSize(aCopyFrom.mSize) {
+ mData = new T[mSize];
+ for (int i = 0; i < mSize; i++)
+ mData[i] = aCopyFrom.mData[i];
+}
+
+// constructor (from array)
+template <class T>
+CVector<T>::CVector(const T* aPointer, const int aSize)
+ : mSize(aSize) {
+ mData = new T[mSize];
+ for (int i = 0; i < mSize; i++)
+ mData[i] = aPointer[i];
+}
+
+// constructor with implicit filling
+template <class T>
+CVector<T>::CVector(const int aSize, const T aFillValue)
+ : mSize(aSize) {
+ mData = new T[aSize];
+ fill(aFillValue);
+}
+
+// destructor
+template <class T>
+CVector<T>::~CVector() {
+ delete[] mData;
+}
+
+// setSize
+template <class T>
+void CVector<T>::setSize(int aSize) {
+ if (mData != 0) delete[] mData;
+ mData = new T[aSize];
+ mSize = aSize;
+}
+
+// fill
+template <class T>
+void CVector<T>::fill(const T aValue) {
+ for (register int i = 0; i < mSize; i++)
+ mData[i] = aValue;
+}
+
+// append
+template <class T>
+void CVector<T>::append(CVector<T>& aVector) {
+ T* aNewData = new T[mSize+aVector.size()];
+ for (int i = 0; i < mSize; i++)
+ aNewData[i] = mData[i];
+ for (int i = 0; i < aVector.size(); i++)
+ aNewData[i+mSize] = aVector(i);
+ mSize += aVector.size();
+ delete[] mData;
+ mData = aNewData;
+}
+
+// normalize
+template <class T>
+void CVector<T>::normalize() {
+ T aSum = 0;
+ for (register int i = 0; i < mSize; i++)
+ aSum += mData[i]*mData[i];
+ if (aSum == 0) return;
+ aSum = 1.0/sqrt(aSum);
+ for (register int i = 0; i < mSize; i++)
+ mData[i] *= aSum;
+}
+
+// normalizeSum
+template <class T>
+void CVector<T>::normalizeSum() {
+ T aSum = 0;
+ for (register int i = 0; i < mSize; i++)
+ aSum += mData[i];
+ if (aSum == 0) return;
+ aSum = 1.0/aSum;
+ for (register int i = 0; i < mSize; i++)
+ mData[i] *= aSum;
+}
+
+// readFromTXT
+template<class T>
+void CVector<T>::readFromTXT(const char* aFilename) {
+ std::ifstream aStream(aFilename);
+ mSize = 0;
+ float aDummy;
+ while (!aStream.eof()) {
+ aStream >> aDummy;
+ mSize++;
+ }
+ aStream.close();
+ std::ifstream aStream2(aFilename);
+ delete mData;
+ mData = new T[mSize];
+ for (int i = 0; i < mSize; i++)
+ aStream2 >> mData[i];
+}
+
+// writeToTXT
+template<class T>
+void CVector<T>::writeToTXT(char* aFilename) {
+ std::ofstream aStream(aFilename);
+ for (int i = 0; i < mSize; i++)
+ aStream << mData[i] << std::endl;
+}
+
+// sum
+template <class T>
+T CVector<T>::sum() {
+ T val = mData[0];
+ for (int i = 1; i < mSize; i++)
+ val += mData[i];
+ return val;
+}
+
+// min
+template <class T>
+T CVector<T>::min() {
+ T bestValue = mData[0];
+ for (int i = 1; i < mSize; i++)
+ if (mData[i] < bestValue) bestValue = mData[i];
+ return bestValue;
+}
+
+// max
+template <class T>
+T CVector<T>::max() {
+ T bestValue = mData[0];
+ for (int i = 1; i < mSize; i++)
+ if (mData[i] > bestValue) bestValue = mData[i];
+ return bestValue;
+}
+
+// norm
+template <class T>
+T CVector<T>::norm() {
+ T aSum = 0.0;
+ for (int i = 0; i < mSize; i++)
+ aSum += mData[i]*mData[i];
+ return sqrt(aSum);
+}
+
+// homogen
+template <class T>
+CVector<T>& CVector<T>::homogen() {
+ if (mSize > 1 && mData[mSize-1] != 0) {
+ T invVal = 1.0/mData[mSize-1];
+ for (int i = 0; i < mSize; i++)
+ mData[i] *= invVal;
+ }
+ return (*this);
+}
+
+// homogen_nD
+template <class T>
+inline void CVector<T>::homogen_nD() {
+ mSize--;
+}
+
+// cross
+template <class T>
+void CVector<T>::cross(CVector<T>& aVector) {
+ T aHelp0 = aVector(2)*mData[1] - aVector(1)*mData[2];
+ T aHelp1 = aVector(0)*mData[2] - aVector(2)*mData[0];
+ T aHelp2 = aVector(1)*mData[0] - aVector(0)*mData[1];
+ mData[0] = aHelp0;
+ mData[1] = aHelp1;
+ mData[2] = aHelp2;
+}
+
+// operator()
+template <class T>
+inline T& CVector<T>::operator()(const int aIndex) const {
+ #ifdef _DEBUG
+ if (aIndex >= mSize || aIndex < 0)
+ throw EVectorRangeOverflow(aIndex);
+ #endif
+ return mData[aIndex];
+}
+
+// operator[]
+template <class T>
+inline T& CVector<T>::operator[](const int aIndex) const {
+ return operator()(aIndex);
+}
+
+// operator=
+template <class T>
+inline CVector<T>& CVector<T>::operator=(const T aValue) {
+ fill(aValue);
+ return *this;
+}
+
+template <class T>
+CVector<T>& CVector<T>::operator=(const CVector<T>& aCopyFrom) {
+ if (this != &aCopyFrom) {
+ if (mSize != aCopyFrom.size()) {
+ delete[] mData;
+ mSize = aCopyFrom.size();
+ mData = new T[mSize];
+ }
+ for (register int i = 0; i < mSize; i++)
+ mData[i] = aCopyFrom.mData[i];
+ }
+ return *this;
+}
+
+template <class T>
+CVector<T>& CVector<T>::operator=(const CMatrix<T>& aCopyFrom) {
+ if (mSize != aCopyFrom.size()) {
+ delete[] mData;
+ mSize = aCopyFrom.size();
+ mData = new T[mSize];
+ }
+ for (register int i = 0; i < mSize; i++)
+ mData[i] = aCopyFrom.data()[i];
+ return *this;
+}
+
+template <class T>
+CVector<T>& CVector<T>::operator=(const CTensor<T>& aCopyFrom) {
+ if (mSize != aCopyFrom.size()) {
+ delete[] mData;
+ mSize = aCopyFrom.size();
+ mData = new T[mSize];
+ }
+ for (register int i = 0; i < mSize; i++)
+ mData[i] = aCopyFrom.data()[i];
+ return *this;
+}
+
+// operator +=
+template <class T>
+CVector<T>& CVector<T>::operator+=(const CVector<T>& aVector) {
+ #ifdef _DEBUG
+ if (mSize != aVector.size()) throw EVectorIncompatibleSize(mSize,aVector.size());
+ #endif
+ for (int i = 0; i < mSize; i++)
+ mData[i] += aVector(i);
+ return *this;
+}
+
+// operator -=
+template <class T>
+CVector<T>& CVector<T>::operator-=(const CVector<T>& aVector) {
+ #ifdef _DEBUG
+ if (mSize != aVector.size()) throw EVectorIncompatibleSize(mSize,aVector.size());
+ #endif
+ for (int i = 0; i < mSize; i++)
+ mData[i] -= aVector(i);
+ return *this;
+}
+
+// operator *=
+template <class T>
+CVector<T>& CVector<T>::operator*=(const T aValue) {
+ for (int i = 0; i < mSize; i++)
+ mData[i] *= aValue;
+ return *this;
+}
+
+template <class T>
+T CVector<T>::operator*=(const CVector<T>& aVector) {
+ #ifdef _DEBUG
+ if (mSize != aVector.size()) throw EVectorIncompatibleSize(mSize,aVector.size());
+ #endif
+ T aSum = 0.0;
+ for (int i = 0; i < mSize; i++)
+ aSum += mData[i]*aVector(i);
+ return aSum;
+}
+
+// operator ==
+template <class T>
+bool CVector<T>::operator==(const CVector<T>& aVector) {
+ if (mSize != aVector.size()) return false;
+ int i = 0;
+ while (i < mSize && aVector(i) == mData[i])
+ i++;
+ return (i == mSize);
+}
+
+// operator !=
+template <class T>
+inline bool CVector<T>::operator!=(const CVector<T>& aVector) {
+ return !((*this)==aVector);
+}
+
+// size
+template <class T>
+inline int CVector<T>::size() const {
+ return mSize;
+}
+
+// N O N - M E M B E R F U N C T I O N S -------------------------------------
+
+// operator +
+template <class T>
+CVector<T> operator+(const CVector<T>& vec1, const CVector<T>& vec2) {
+ #ifdef _DEBUG
+ if (vec1.size() != vec2.size()) throw EVectorIncompatibleSize(vec1.size(),vec2.size());
+ #endif
+ CVector<T> result(vec1.size());
+ for (int i = 0; i < vec1.size(); i++)
+ result(i) = vec1[i]+vec2[i];
+ return result;
+}
+
+// operator -
+template <class T>
+CVector<T> operator-(const CVector<T>& vec1, const CVector<T>& vec2) {
+ #ifdef _DEBUG
+ if (vec1.size() != vec2.size()) throw EVectorIncompatibleSize(vec1.size(),vec2.size());
+ #endif
+ CVector<T> result(vec1.size());
+ for (int i = 0; i < vec1.size(); i++)
+ result(i) = vec1(i)-vec2(i);
+ return result;
+}
+
+// operator *
+template <class T>
+CVector<T> operator*(const T aValue, const CVector<T>& aVector) {
+ CVector<T> result(aVector.size());
+ for (int i = 0; i < aVector.size(); i++)
+ result(i) = aValue*aVector(i);
+ return result;
+}
+
+template <class T>
+CVector<T> operator*(const CVector<T>& aVector, const T aValue) {
+ return operator*(aValue,aVector);
+}
+
+template <class T>
+T operator*(const CVector<T>& vec1, const CVector<T>& vec2) {
+ #ifdef _DEBUG
+ if (vec1.size() != vec2.size()) throw EVectorIncompatibleSize(vec1.size(),vec2.size());
+ #endif
+ T aSum = 0.0;
+ for (int i = 0; i < vec1.size(); i++)
+ aSum += vec1(i)*vec2(i);
+ return aSum;
+}
+
+// operator /
+template <class T>
+CVector<T> operator/(const CVector<T>& vec1, const CVector<T>& vec2) {
+ CVector<T> result(3);
+ result[0]=vec1[1]*vec2[2] - vec1[2]*vec2[1];
+ result[1]=vec1[2]*vec2[0] - vec1[0]*vec2[2];
+ result[2]=vec1[0]*vec2[1] - vec1[1]*vec2[0];
+ return result;
+}
+
+// operator <<
+template <class T>
+std::ostream& operator<<(std::ostream& aStream, const CVector<T>& aVector) {
+ for (int i = 0; i < aVector.size(); i++)
+ aStream << aVector(i) << '|';
+ aStream << std::endl;
+ return aStream;
+}
+
+#endif
diff --git a/video_input/consistencyChecker/Makefile b/video_input/consistencyChecker/Makefile new file mode 100644 index 0000000..94725e2 --- /dev/null +++ b/video_input/consistencyChecker/Makefile @@ -0,0 +1,3 @@ +default: + g++ -O3 -fPIC consistencyChecker.cpp NMath.cpp -I. -o consistencyChecker -L. + diff --git a/video_input/consistencyChecker/NMath.cpp b/video_input/consistencyChecker/NMath.cpp new file mode 100644 index 0000000..3a58b16 --- /dev/null +++ b/video_input/consistencyChecker/NMath.cpp @@ -0,0 +1,664 @@ +// Copyright: Thomas Brox
+
+#include <math.h>
+#include <stdlib.h>
+#include <NMath.h>
+
+namespace NMath {
+
+ const float Pi = 3.1415926536;
+
+ // faculty
+ int faculty(int n) {
+ int aResult = 1;
+ for (int i = 2; i <= n; i++)
+ aResult *= i;
+ return aResult;
+ }
+
+ // binCoeff
+ int binCoeff(const int n, const int k) {
+ if (k > (n >> 1)) return binCoeff(n,n-k);
+ int aResult = 1;
+ for (int i = n; i > (n-k); i--)
+ aResult *= i;
+ for (int j = 2; j <= k; j++)
+ aResult /= j;
+ return aResult;
+ }
+
+ // tangent
+ float tangent(const float x1, const float y1, const float x2, const float y2) {
+ float alpha;
+ float xDiff = x2-x1;
+ float yDiff = y2-y1;
+ if (yDiff > 0) {
+ if (xDiff == 0) alpha = 0.5*Pi;
+ else if (xDiff > 0) alpha = atan(yDiff/xDiff);
+ else alpha = Pi+atan(yDiff/xDiff);
+ }
+ else {
+ if (xDiff == 0) alpha = -0.5*Pi;
+ else if (xDiff > 0) alpha = atan(yDiff/xDiff);
+ else alpha = -Pi+atan(yDiff/xDiff);
+ }
+ return alpha;
+ }
+
+ // absAngleDifference
+ float absAngleDifference(const float aFirstAngle, const float aSecondAngle) {
+ float aAlphaDiff = abs(aFirstAngle - aSecondAngle);
+ if (aAlphaDiff > Pi) aAlphaDiff = 2*Pi-aAlphaDiff;
+ return aAlphaDiff;
+ }
+
+ // angleDifference
+ float angleDifference(const float aFirstAngle, const float aSecondAngle) {
+ float aAlphaDiff = aFirstAngle - aSecondAngle;
+ if (aAlphaDiff > Pi) aAlphaDiff = -2*Pi+aAlphaDiff;
+ else if (aAlphaDiff < -Pi) aAlphaDiff = 2*Pi+aAlphaDiff;
+ return aAlphaDiff;
+ }
+
+ // angleSum
+ float angleSum(const float aFirstAngle, const float aSecondAngle) {
+ float aSum = aFirstAngle + aSecondAngle;
+ if (aSum > Pi) aSum = -2*Pi+aSum;
+ else if (aSum < -Pi) aSum = 2*Pi+aSum;
+ return aSum;
+ }
+
+ // round
+ int round(const float aValue) {
+ float temp1 = floor(aValue);
+ float temp2 = ceil(aValue);
+ if (aValue-temp1 < 0.5) return (int)temp1;
+ else return (int)temp2;
+ }
+
+ // PATransformation
+ // Cyclic Jacobi method for determining the eigenvalues and eigenvectors
+ // of a symmetric matrix.
+ // Ref.: H.R. Schwarz: Numerische Mathematik. Teubner, Stuttgart, 1988.
+ // pp. 243-246.
+ void PATransformation(const CMatrix<float>& aMatrix, CVector<float>& aEigenvalues, CMatrix<float>& aEigenvectors, bool aOrdering) {
+ static const float eps = 0.0001;
+ static const float delta = 0.000001;
+ static const float eps2 = eps*eps;
+ float sum,theta,t,c,r,s,g,h;
+ // Initialization
+ CMatrix<float> aCopy(aMatrix);
+ int n = aEigenvalues.size();
+ aEigenvectors = 0;
+ for (int i = 0; i < n; i++)
+ aEigenvectors(i,i) = 1;
+ // Loop
+ do {
+ // check whether accuracy is reached
+ sum = 0.0;
+ for (int i = 1; i < n; i++)
+ for (int j = 0; j <= i-1; j++)
+ sum += aCopy(i,j)*aCopy(i,j);
+ if (sum+sum > eps2) {
+ for (int p = 0; p < n-1; p++)
+ for (int q = p+1; q < n; q++)
+ if (fabs(aCopy(q,p)) >= eps2) {
+ theta = (aCopy(q,q) - aCopy(p,p)) / (2.0 * aCopy(q,p));
+ t = 1.0;
+ if (fabs(theta) > delta) t = 1.0 / (theta + theta/fabs(theta) * sqrt (theta*theta + 1.0));
+ c = 1.0 / sqrt (1.0 + t*t);
+ s = c*t;
+ r = s / (1.0 + c);
+ aCopy(p,p) -= t * aCopy(q,p);
+ aCopy(q,q) += t * aCopy(q,p);
+ aCopy(q,p) = 0;
+ for (int j = 0; j <= p-1; j++) {
+ g = aCopy(q,j) + r * aCopy(p,j);
+ h = aCopy(p,j) - r * aCopy(q,j);
+ aCopy(p,j) -= s*g;
+ aCopy(q,j) += s*h;
+ }
+ for (int i = p+1; i <= q-1; i++) {
+ g = aCopy(q,i) + r * aCopy(i,p);
+ h = aCopy(i,p) - r * aCopy(q,i);
+ aCopy(i,p) -= s * g;
+ aCopy(q,i) += s * h;
+ }
+ for (int i = q+1; i < n; i++) {
+ g = aCopy(i,q) + r * aCopy(i,p);
+ h = aCopy(i,p) - r * aCopy(i,q);
+ aCopy(i,p) -= s * g;
+ aCopy(i,q) += s * h;
+ }
+ for (int i = 0; i < n; i++) {
+ g = aEigenvectors(i,q) + r * aEigenvectors(i,p);
+ h = aEigenvectors(i,p) - r * aEigenvectors(i,q);
+ aEigenvectors(i,p) -= s * g;
+ aEigenvectors(i,q) += s * h;
+ }
+ }
+ }
+ }
+ // Return eigenvalues
+ while (sum+sum > eps2);
+ for (int i = 0; i < n; i++)
+ aEigenvalues(i) = aCopy(i,i);
+ if (aOrdering) {
+ // Order eigenvalues and eigenvectors
+ for (int i = 0; i < n-1; i++) {
+ int k = i;
+ for (int j = i+1; j < n; j++)
+ if (fabs(aEigenvalues(j)) > fabs(aEigenvalues(k))) k = j;
+ if (k != i) {
+ // Switch eigenvalue i and k
+ float help = aEigenvalues(k);
+ aEigenvalues(k) = aEigenvalues(i);
+ aEigenvalues(i) = help;
+ // Switch eigenvector i and k
+ for (int j = 0; j < n; j++) {
+ help = aEigenvectors(j,k);
+ aEigenvectors(j,k) = aEigenvectors(j,i);
+ aEigenvectors(j,i) = help;
+ }
+ }
+ }
+ }
+ }
+
+ // PABackTransformation
+ void PABacktransformation(const CMatrix<float>& aEigenvectors, const CVector<float>& aEigenvalues, CMatrix<float>& aMatrix) {
+ for (int i = 0; i < aEigenvalues.size(); i++)
+ for (int j = 0; j <= i; j++) {
+ float sum = aEigenvalues(0) * aEigenvectors(i,0) * aEigenvectors(j,0);
+ for (int k = 1; k < aEigenvalues.size(); k++)
+ sum += aEigenvalues(k) * aEigenvectors(i,k) * aEigenvectors(j,k);
+ aMatrix(i,j) = sum;
+ }
+ for (int i = 0; i < aEigenvalues.size(); i++)
+ for (int j = i+1; j < aEigenvalues.size(); j++)
+ aMatrix(i,j) = aMatrix(j,i);
+ }
+
+ // svd (nach Numerical Recipes in C basierend auf Forsythe et al.: Computer Methods for
+ // Mathematical Computations (Englewood Cliffs, NJ: Prentice-Hall), Chapter 9, 1977,
+ // Code übernommen von Bodo Rosenhahn)
+ void svd(CMatrix<float>& U, CMatrix<float>& S, CMatrix<float>& V, bool aOrdering, int aIterations) {
+ static float at,bt,ct;
+ static float maxarg1,maxarg2;
+ #define PYTHAG(a,b) ((at=fabs(a)) > (bt=fabs(b)) ? (ct=bt/at,at*sqrt(1.0+ct*ct)) : (bt ? (ct=at/bt,bt*sqrt(1.0+ct*ct)): 0.0))
+ #define MAX(a,b) (maxarg1=(a),maxarg2=(b),(maxarg1) > (maxarg2) ? (maxarg1) : (maxarg2))
+ #define MIN(a,b) ((a) >(b) ? (b) : (a))
+ #define SIGN(a,b) ((b) >= 0.0 ? fabs(a) : -fabs(a))
+ int flag,i,its,j,jj,k,l,nm;
+ float c,f,h,s,x,y,z;
+ float anorm=0.0,g=0.0,scale=0.0;
+ int aXSize = U.xSize();
+ int aYSize = U.ySize();
+ CVector<float> aBuffer(aXSize);
+ for (i = 0; i < aXSize; i++) {
+ l=i+1;
+ aBuffer(i)=scale*g;
+ g=s=scale=0.0;
+ if (i < aYSize) {
+ for (k = i; k < aYSize; k++)
+ scale += fabs(U(i,k));
+ if (scale) {
+ for (k = i; k < aYSize; k++) {
+ U(i,k) /= scale;
+ s += U(i,k)*U(i,k);
+ }
+ f = U(i,i);
+ g = -SIGN(sqrt(s),f);
+ h = f*g-s;
+ U(i,i) = f-g;
+ for (j = l; j < aXSize; j++) {
+ for (s = 0.0, k = i; k < aYSize; k++)
+ s += U(i,k)*U(j,k);
+ f = s/h;
+ for (k = i; k < aYSize; k++)
+ U(j,k) += f*U(i,k);
+ }
+ for ( k = i; k < aYSize; k++)
+ U(i,k) *= scale;
+ }
+ }
+ S(i,i) = scale*g;
+ g=s=scale=0.0;
+ if (i < aYSize && i != aXSize-1) {
+ for (k = l; k < aXSize; k++)
+ scale += fabs(U(k,i));
+ if (scale != 0) {
+ for (k = l; k < aXSize; k++) {
+ U(k,i) /= scale;
+ s += U(k,i)*U(k,i);
+ }
+ f = U(l,i);
+ g = -SIGN(sqrt(s),f);
+ h = f*g-s;
+ U(l,i) = f-g;
+ for (k = l; k < aXSize; k++)
+ aBuffer(k) = U(k,i)/h;
+ for (j = l; j < aYSize; j++) {
+ for (s = 0.0, k = l; k < aXSize; k++)
+ s += U(k,j)*U(k,i);
+ for (k = l; k < aXSize; k++)
+ U(k,j) += s*aBuffer(k);
+ }
+ for (k = l; k < aXSize; k++)
+ U(k,i) *= scale;
+ }
+ }
+ anorm = MAX(anorm,(fabs(S(i,i))+fabs(aBuffer(i))));
+ }
+ for (i = aXSize-1; i >= 0; i--) {
+ if (i < aXSize-1) {
+ if (g != 0) {
+ for (j = l; j < aXSize; j++)
+ V(i,j) = U(j,i)/(U(l,i)*g);
+ for (j = l; j < aXSize; j++) {
+ for (s = 0.0, k = l; k < aXSize; k++)
+ s += U(k,i)*V(j,k);
+ for (k = l; k < aXSize; k++)
+ V(j,k) += s*V(i,k);
+ }
+ }
+ for (j = l; j < aXSize; j++)
+ V(j,i) = V(i,j) = 0.0;
+ }
+ V(i,i) = 1.0;
+ g = aBuffer(i);
+ l = i;
+ }
+ for (i = MIN(aYSize-1,aXSize-1); i >= 0; i--) {
+ l = i+1;
+ g = S(i,i);
+ for (j = l; j < aXSize; j++)
+ U(j,i) = 0.0;
+ if (g != 0) {
+ g = 1.0/g;
+ for (j = l; j < aXSize; j++) {
+ for (s = 0.0, k = l; k < aYSize; k++)
+ s += U(i,k)*U(j,k);
+ f = (s/U(i,i))*g;
+ for (k = i; k < aYSize; k++)
+ U(j,k) += f*U(i,k);
+ }
+ for (j = i; j < aYSize; j++)
+ U(i,j) *= g;
+ }
+ else {
+ for (j = i; j < aYSize; j++)
+ U(i,j) = 0.0;
+ }
+ ++U(i,i);
+ }
+ for (k = aXSize-1; k >= 0; k--) {
+ for (its = 1; its <= aIterations; its++) {
+ flag = 1;
+ for (l = k; l > 0; l--) {
+ nm = l - 1;
+ if (fabs(aBuffer(l))+anorm == anorm) {
+ flag = 0; break;
+ }
+ if (fabs(S(nm,nm))+anorm == anorm) break;
+ }
+ if (flag) {
+ c = 0.0;
+ s = 1.0;
+ for (i = l; i <= k; i++) {
+ f = s*aBuffer(i);
+ aBuffer(i) = c*aBuffer(i);
+ if (fabs(f)+anorm == anorm) break;
+ g = S(i,i);
+ h = PYTHAG(f,g);
+ S(i,i) = h;
+ h = 1.0/h;
+ c = g*h;
+ s=-f*h;
+ for (j = 0; j < aYSize; j++) {
+ y = U(nm,j);
+ z = U(i,j);
+ U(nm,j) = y*c + z*s;
+ U(i,j) = z*c - y*s;
+ }
+ }
+ }
+ z = S(k,k);
+ if (l == k) {
+ if (z < 0.0) {
+ S(k,k) = -z;
+ for (j = 0; j < aXSize; j++)
+ V(k,j) = -V(k,j);
+ }
+ break;
+ }
+ if (its == aIterations) std::cerr << "svd: No convergence in " << aIterations << " iterations" << std::endl;
+ x = S(l,l);
+ nm = k-1;
+ y = S(nm,nm);
+ g = aBuffer(nm);
+ h = aBuffer(k);
+ f = ((y-z)*(y+z)+(g-h)*(g+h))/(2.0*h*y);
+ g = PYTHAG(f,1.0);
+ f = ((x-z)*(x+z)+h*((y/(f+SIGN(g,f)))-h))/x;
+ c = s = 1.0;
+ for (j = l; j <= nm; j++) {
+ i = j+1;
+ g = aBuffer(i);
+ y = S(i,i);
+ h = s*g;
+ g = c*g;
+ z = PYTHAG(f,h);
+ aBuffer(j) = z;
+ float invZ = 1.0/z;
+ c = f*invZ;
+ s = h*invZ;
+ f = x*c+g*s;
+ g = g*c-x*s;
+ h = y*s;
+ y *= c;
+ for (jj = 0; jj < aXSize; jj++) {
+ x = V(j,jj);
+ z = V(i,jj);
+ V(j,jj) = x*c + z*s;
+ V(i,jj) = z*c - x*s;
+ }
+ z = PYTHAG(f,h);
+ S(j,j) = z;
+ if (z != 0) {
+ z = 1.0/z;
+ c = f*z;
+ s = h*z;
+ }
+ f = (c*g)+(s*y);
+ x = (c*y)-(s*g);
+ for (jj = 0; jj < aYSize; jj++) {
+ y = U(j,jj);
+ z = U(i,jj);
+ U(j,jj) = y*c + z*s;
+ U(i,jj) = z*c - y*s;
+ }
+ }
+ aBuffer(l) = 0.0;
+ aBuffer(k) = f;
+ S(k,k) = x;
+ }
+ }
+ // Order singular values
+ if (aOrdering) {
+ for (int i = 0; i < aXSize-1; i++) {
+ int k = i;
+ for (int j = i+1; j < aXSize; j++)
+ if (fabs(S(j,j)) > fabs(S(k,k))) k = j;
+ if (k != i) {
+ // Switch singular value i and k
+ float help = S(k,k);
+ S(k,k) = S(i,i);
+ S(i,i) = help;
+ // Switch columns i and k in U and V
+ for (int j = 0; j < aYSize; j++) {
+ help = U(k,j);
+ U(k,j) = U(i,j);
+ U(i,j) = help;
+ }
+ for (int j = 0; j < aXSize; j++) {
+ help = V(k,j);
+ V(k,j) = V(i,j);
+ V(i,j) = help;
+ }
+ }
+ }
+ }
+ }
+
+ #undef PYTHAG
+ #undef MAX
+ #undef MIN
+ #undef SIGN
+
+ // svdBack
+ void svdBack(CMatrix<float>& U, const CMatrix<float>& S, const CMatrix<float>& V) {
+ for (int y = 0; y < U.ySize(); y++)
+ for (int x = 0; x < U.xSize(); x++)
+ U(x,y) = S(x,x)*U(x,y);
+ U *= trans(V);
+ }
+
+ // Householder-Verfahren (nach Stoer), uebernommen von Bodo Rosenhahn
+ // Bei dem Verfahren wird die Matrix A (hier:*this und die rechte Seite (b)
+ // mit unitaeren Matrizen P multipliziert, so dass A in eine
+ // obere Dreiecksmatrix umgewandelt wird.
+ // Dabei ist P = I + beta * u * uH
+ // Die Vektoren u werden bei jeder Transformation in den nicht
+ // benoetigten unteren Spalten von A gesichert.
+
+ void householder(CMatrix<float>& A, CVector<float>& b) {
+ int i,j,k;
+ float sigma,s,beta,sum;
+ CVector<float> d(A.xSize());
+ for (j = 0; j < A.xSize(); ++j) {
+ sigma = 0;
+ for (i = j; i < A.ySize(); ++i)
+ sigma += A(j,i)*A(j,i);
+ if (sigma == 0) {
+ std::cerr << "NMath::householder(): matrix is singular!" << std::endl;
+ break;
+ }
+ // Choose sign to avoid elimination
+ s = d(j) = A(j,j)<0 ? sqrt(sigma) : -sqrt(sigma);
+ beta = 1.0/(s*A(j,j)-sigma);
+ A(j,j) -= s;
+ // Transform submatrix of A with P
+ for (k = j+1; k < A.xSize(); ++k) {
+ sum = 0.0;
+ for (i = j; i < A.ySize(); ++i)
+ sum += (A(j,i)*A(k,i));
+ sum *= beta;
+ for (i = j; i < A.ySize(); ++i)
+ A(k,i) += A(j,i)*sum;
+ }
+ // Transform right hand side of linear system with P
+ sum = 0.0;
+ for (i = j; i < A.ySize(); ++i)
+ sum += A(j,i)*b(i);
+ sum *= beta;
+ for (i = j; i < A.ySize(); ++i)
+ b(i) += A(j,i)*sum;
+ }
+ for (i = 0; i < A.xSize(); ++i)
+ A(i,i) = d(i);
+ }
+
+ // leastSquares
+ CVector<float> leastSquares(CMatrix<float>& A, CVector<float>& b) {
+ CVector<float> aResult(A.xSize());
+ householder(A,b);
+ for (int i = A.xSize()-1; i >= 0; i--) {
+ float s = 0;
+ for (int k = i+1; k < A.xSize(); k++)
+ s += A(k,i)*aResult(k);
+ aResult(i) = (b(i)-s)/A(i,i);
+ }
+ return aResult;
+ }
+
+ // invRegularized
+ void invRegularized(CMatrix<float>& A, int aReg) {
+ if (A.xSize() != A.ySize()) throw ENonquadraticMatrix(A.xSize(),A.ySize());
+ CVector<float> eVals(A.xSize());
+ CMatrix<float> eVecs(A.xSize(),A.ySize());
+ PATransformation(A,eVals,eVecs);
+ for (int i = 0 ; i < A.xSize(); i++)
+ if (eVals(i) < aReg) eVals(i) = 1.0/aReg;
+ else eVals(i) = 1.0/eVals(i);
+ PABacktransformation(eVecs,eVals,A);
+ }
+
+ // cholesky
+ void cholesky(CMatrix<float>& A) {
+ if (A.xSize() != A.ySize()) throw ENonquadraticMatrix(A.xSize(),A.ySize());
+ CVector<float> d(A.xSize());
+ for (int i = 0; i < A.xSize(); i++)
+ for (int j = i; j < A.ySize(); j++) {
+ float sum = A(j,i);
+ for (int k = i-1; k >= 0; k--)
+ sum -= A(k,i)*A(k,j);
+ if (i == j) {
+ if (sum <= 0.0) return;//throw ENonPositiveDefinite();
+ d(i) = sqrt(sum);
+ }
+ else A(i,j) = sum/d(i);
+ }
+ for (int i = 0; i < A.xSize(); i++)
+ A(i,i) = d(i);
+ }
+
+ // triangularSolve
+ void triangularSolve(CMatrix<float>& L, CVector<float>& aIn, CVector<float>& aOut) {
+ for (int i = 0; i < aIn.size(); i++) {
+ float sum = aIn(i);
+ for (int j = 0; j < i; j++)
+ sum -= L(j,i)*aOut(j);
+ aOut(i) = sum/L(i,i);
+ }
+ }
+
+ void triangularSolve(CMatrix<float>& L, CMatrix<float>& aIn, CMatrix<float>& aOut) {
+ CVector<float> invLii(aIn.xSize());
+ for (int i = 0; i < aIn.xSize(); i++)
+ invLii(i) = 1.0/L(i,i);
+ for (int k = 0; k < aIn.ySize(); k++)
+ for (int i = 0; i < aIn.xSize(); i++) {
+ float sum = aIn(i,k);
+ for (int j = 0; j < i; j++)
+ sum -= L(j,i)*aOut(j,k);
+ aOut(i,k) = sum*invLii(i);
+ }
+ }
+
+ // triangularSolveTransposed
+ void triangularSolveTransposed(CMatrix<float>& L, CVector<float>& aIn, CVector<float>& aOut) {
+ for (int i = aIn.size()-1; i >= 0; i--) {
+ float sum = aIn(i);
+ for (int j = aIn.size()-1; j > i; j--)
+ sum -= L(i,j)*aOut(j);
+ aOut(i) = sum/L(i,i);
+ }
+ }
+
+ void triangularSolveTransposed(CMatrix<float>& L, CMatrix<float>& aIn, CMatrix<float>& aOut) {
+ CVector<float> invLii(aIn.xSize());
+ for (int i = 0; i < aIn.xSize(); i++)
+ invLii(i) = 1.0/L(i,i);
+ for (int k = 0; k < aIn.ySize(); k++)
+ for (int i = aIn.xSize()-1; i >= 0; i--) {
+ float sum = aIn(i,k);
+ for (int j = aIn.xSize()-1; j > i; j--)
+ sum -= L(i,j)*aOut(j,k);
+ aOut(i,k) = sum*invLii(i);
+ }
+ }
+
+ // choleskyInv
+ void choleskyInv(const CMatrix<float>& L, CMatrix<float>& aInv) {
+ aInv = 0;
+ // Compute the inverse of L
+ CMatrix<float> invL(L.xSize(),L.ySize());
+ for (int i = 0; i < L.xSize(); i++)
+ invL(i,i) = 1.0/L(i,i);
+ for (int i = 0; i < L.xSize(); i++)
+ for (int j = i+1; j < L.ySize(); j++) {
+ float sum = 0.0;
+ for (int k = i; k < j; k++)
+ sum -= L(k,j)*invL(i,k);
+ invL(i,j) = sum*invL(j,j);
+ }
+ // Compute lower triangle of aInv = invL^T * invL
+ for (int i = 0; i < aInv.xSize(); i++)
+ for (int j = i; j < aInv.ySize(); j++) {
+ float sum = 0.0;
+ for (int k = j; k < aInv.ySize(); k++)
+ sum += invL(j,k)*invL(i,k);
+ aInv(i,j) = sum;
+ }
+ // Complete aInv
+ for (int i = 0; i < aInv.xSize(); i++)
+ for (int j = i+1; j < aInv.ySize(); j++)
+ aInv(j,i) = aInv(i,j);
+ }
+
+ // eulerAngles
+ void eulerAngles(float rx, float ry, float rz, CMatrix<float>& A) {
+ CMatrix<float> Rx(4,4,0);
+ CMatrix<float> Ry(4,4,0);
+ CMatrix<float> Rz(4,4,0);
+ Rx(0,0)=1.0;Rx(1,1)=cos(rx);Rx(2,2)=cos(rx);Rx(3,3)=1.0;
+ Rx(2,1)=-sin(rx);Rx(1,2)=sin(rx);
+ Ry(1,1)=1.0;Ry(0,0)=cos(ry);Ry(2,2)=cos(ry);Ry(3,3)=1.0;
+ Ry(0,2)=-sin(ry);Ry(2,0)=sin(ry);
+ Rz(2,2)=1.0;Rz(0,0)=cos(rz);Rz(1,1)=cos(rz);Rz(3,3)=1.0;
+ Rz(1,0)=-sin(rz);Rz(0,1)=sin(rz);
+ A=Rz*Ry*Rx;
+ }
+
+ // RBM2Twist
+ void RBM2Twist(CVector<float>& T, CMatrix<float>& fRBM) {
+ T.setSize(6);
+ CMatrix<double> dRBM(4,4);
+ for (int i = 0; i < 16; i++)
+ dRBM.data()[i] = fRBM.data()[i];
+ CVector<double> omega;
+ double theta;
+ CVector<double> v;
+ CMatrix<double> R(3,3);
+ double sum = 0.0;
+ for (int i = 0; i < 3; i++)
+ for (int j = 0; j < 3; j++)
+ if (i != j) sum += dRBM(i,j)*dRBM(i,j);
+ else sum += (dRBM(i,i)-1.0)*(dRBM(i,i)-1.0);
+ if (sum < 0.0000001) {
+ T(0)=fRBM(3,0); T(1)=fRBM(3,1); T(2)=fRBM(3,2);
+ T(3)=0.0; T(4)=0.0; T(5)=0.0;
+ }
+ else {
+ double diag = (dRBM(0,0)+dRBM(1,1)+dRBM(2,2)-1.0)*0.5;
+ if (diag < -1.0) diag = -1.0;
+ else if (diag > 1.0) diag = 1.0;
+ theta = acos(diag);
+ if (sin(theta)==0) theta += 0.0000001;
+ omega.setSize(3);
+ omega(0)=(dRBM(1,2)-dRBM(2,1));
+ omega(1)=(dRBM(2,0)-dRBM(0,2));
+ omega(2)=(dRBM(0,1)-dRBM(1,0));
+ omega*=(1.0/(2.0*sin(theta)));
+ CMatrix<double> omegaHat(3,3);
+ omegaHat.data()[0] = 0.0; omegaHat.data()[1] = -omega(2); omegaHat.data()[2] = omega(1);
+ omegaHat.data()[3] = omega(2); omegaHat.data()[4] = 0.0; omegaHat.data()[5] = -omega(0);
+ omegaHat.data()[6] = -omega(1); omegaHat.data()[7] = omega(0); omegaHat.data()[8] = 0.0;
+ CMatrix<double> omegaT(3,3);
+ for (int j = 0; j < 3; j++)
+ for (int i = 0; i < 3; i++)
+ omegaT(i,j) = omega(i)*omega(j);
+ R = (omegaHat*(double)sin(theta))+((omegaHat*omegaHat)*(double)(1.0-cos(theta)));
+ R(0,0) += 1.0; R(1,1) += 1.0; R(2,2) += 1.0;
+ CMatrix<double> A(3,3);
+ A.fill(0.0);
+ A(0,0)=1.0; A(1,1)=1.0; A(2,2)=1.0;
+ A -= R; A*=omegaHat; A+=omegaT*theta;
+ CVector<double> p(3);
+ p(0)=dRBM(3,0);
+ p(1)=dRBM(3,1);
+ p(2)=dRBM(3,2);
+ A.inv();
+ v=A*p;
+ T(0) = (float)(v(0)*theta);
+ T(1) = (float)(v(1)*theta);
+ T(2) = (float)(v(2)*theta);
+ T(3) = (float)(theta*omega(0));
+ T(4) = (float)(theta*omega(1));
+ T(5) = (float)(theta*omega(2));
+ }
+ }
+
+}
+
diff --git a/video_input/consistencyChecker/NMath.h b/video_input/consistencyChecker/NMath.h new file mode 100644 index 0000000..5f31c3a --- /dev/null +++ b/video_input/consistencyChecker/NMath.h @@ -0,0 +1,170 @@ +// NMath
+// A collection of mathematical functions and numerical algorithms
+//
+// Author: Thomas Brox
+
+#ifndef NMathH
+#define NMathH
+
+#include <math.h>
+#include <stdlib.h>
+#include <CVector.h>
+#include <CMatrix.h>
+
+namespace NMath {
+ // Returns the faculty of a number
+ int faculty(int n);
+ // Computes the binomial coefficient of two numbers
+ int binCoeff(const int n, const int k);
+ // Returns the angle of the line connecting (x1,y1) with (y1,y2)
+ float tangent(const float x1, const float y1, const float x2, const float y2);
+ // Absolute for floating points
+ inline float abs(const float aValue);
+ // Computes min or max value of two numbers
+ inline float min(float aVal1, float aVal2);
+ inline float max(float aVal1, float aVal2);
+ inline int min(int aVal1, int aVal2);
+ inline int max(int aVal1, int aVal2);
+ // Computes the sign of a value
+ inline float sign(float aVal);
+ // minmod function (see description in implementation)
+ inline float minmod(float a, float b, float c);
+ // Computes the difference between two angles respecting the cyclic property of an angle
+ // The result is always between 0 and Pi
+ float absAngleDifference(const float aFirstAngle, const float aSecondAngle);
+ // Computes the difference between two angles aFirstAngle - aSecondAngle
+ // respecting the cyclic property of an angle
+ // The result ist between -Pi and Pi
+ float angleDifference(const float aFirstAngle, const float aSecondAngle);
+ // Computes the sum of two angles respecting the cyclic property of an angle
+ // The result is between -Pi and Pi
+ float angleSum(const float aFirstAngle, const float aSecondAngle);
+ // Rounds to the nearest integer
+ int round(const float aValue);
+ // Computes the arctan with results between 0 and 2*Pi
+ inline float arctan(float x, float y);
+
+ // Computes [0,1] uniformly distributed random number
+ inline float random();
+ // Computes N(0,1) distributed random number
+ inline float randomGauss();
+
+ extern const float Pi;
+
+ // Computes a principal axis transformation
+ // Eigenvectors are in the rows of aEigenvectors
+ void PATransformation(const CMatrix<float>& aMatrix, CVector<float>& aEigenvalues, CMatrix<float>& aEigenvectors, bool aOrdering = true);
+ // Computes the principal axis backtransformation
+ void PABacktransformation(const CMatrix<float>& aEigenVectors, const CVector<float>& aEigenValues, CMatrix<float>& aMatrix);
+ // Computes a singular value decomposition A=USV^T
+ // Input: U MxN matrix
+ // Output: U MxN matrix, S NxN diagonal matrix, V NxN diagonal matrix
+ void svd(CMatrix<float>& U, CMatrix<float>& S, CMatrix<float>& V, bool aOrdering = true, int aIterations = 20);
+ // Reassembles A = USV^T, Result in U
+ void svdBack(CMatrix<float>& U, const CMatrix<float>& S, const CMatrix<float>& V);
+ // Applies the Householder method to A and b, i.e., A is transformed into an upper triangular matrix
+ void householder(CMatrix<float>& A, CVector<float>& b);
+ // Computes least squares solution of an overdetermined linear system Ax=b using the Householder method
+ CVector<float> leastSquares(CMatrix<float>& A, CVector<float>& b);
+ // Inverts a square matrix by eigenvalue decomposition,
+ // eigenvalues smaller than aReg are replaced by aReg
+ void invRegularized(CMatrix<float>& A, int aReg);
+ // Given a positive-definite symmetric matrix A, this routine constructs A = LL^T.
+ // Only the upper triangle of A need be given. L is returned in the lower triangle.
+ void cholesky(CMatrix<float>& A);
+ // Solves L*aOut = aIn when L is a lower triangular matrix (e.g. result from cholesky)
+ void triangularSolve(CMatrix<float>& L, CVector<float>& aIn, CVector<float>& aOut);
+ void triangularSolve(CMatrix<float>& L, CMatrix<float>& aIn, CMatrix<float>& aOut);
+ // Solves L^T*aOut = aIn when L is a lower triangular matrix (e.g. result from cholesky)
+ void triangularSolveTransposed(CMatrix<float>& L, CVector<float>& aIn, CVector<float>& aOut);
+ void triangularSolveTransposed(CMatrix<float>& L, CMatrix<float>& aIn, CMatrix<float>& aOut);
+ // Computes the inverse of a matrix, given its cholesky decomposition L (lower triangle)
+ void choleskyInv(const CMatrix<float>& L, CMatrix<float>& aInv);
+ // Creates the rotation matrix RzRyRx and extends it to a 4x4 RBM matrix with translation 0
+ void eulerAngles(float rx, float ry, float rz, CMatrix<float>& A);
+ // Transforms a rigid body motion in matrix representation to a twist representation
+ void RBM2Twist(CVector<float> &T, CMatrix<float>& RBM);
+}
+
+// I M P L E M E N T A T I O N -------------------------------------------------
+// Inline functions have to be implemented directly in the header file
+
+namespace NMath {
+
+ // abs
+ inline float abs(const float aValue) {
+ if (aValue >= 0) return aValue;
+ else return -aValue;
+ }
+
+ // min
+ inline float min(float aVal1, float aVal2) {
+ if (aVal1 < aVal2) return aVal1;
+ else return aVal2;
+ }
+
+ // max
+ inline float max(float aVal1, float aVal2) {
+ if (aVal1 > aVal2) return aVal1;
+ else return aVal2;
+ }
+
+ // min
+ inline int min(int aVal1, int aVal2) {
+ if (aVal1 < aVal2) return aVal1;
+ else return aVal2;
+ }
+
+ // max
+ inline int max(int aVal1, int aVal2) {
+ if (aVal1 > aVal2) return aVal1;
+ else return aVal2;
+ }
+
+ // sign
+ inline float sign(float aVal) {
+ if (aVal > 0) return 1.0;
+ else return -1.0;
+ }
+
+ // minmod function:
+ // 0, if any of the a, b, c are 0 or of opposite sign
+ // sign(a) min(|a|,|b|,|c|) else
+ inline float minmod(float a, float b, float c) {
+ if ((sign(a) == sign(b)) && (sign(b) == sign(c)) && (a != 0.0)) {
+ float aMin = fabs(a);
+ if (fabs(b) < aMin) aMin = fabs(b);
+ if (fabs(c) < aMin) aMin = fabs(c);
+ return sign(a)*aMin;
+ }
+ else return 0.0;
+ }
+
+ // arctan
+ inline float arctan(float x, float y) {
+ if (x == 0.0)
+ if (y >= 0.0) return 0.5 * 3.1415926536;
+ else return 1.5 * 3.1415926536;
+ else if (x > 0.0)
+ if (y >= 0.0) return atan (y/x);
+ else return 2.0 * 3.1415926536 + atan (y/x);
+ else return 3.1415926536 + atan (y/x);
+ }
+
+ // random
+ inline float random() {
+ return (float)rand()/RAND_MAX;
+ }
+
+ // randomGauss
+ inline float randomGauss() {
+ // Draw two [0,1]-uniformly distributed numbers a and b
+ float a = random();
+ float b = random();
+ // assemble a N(0,1) number c according to Box-Muller */
+ if (a > 0.0) return sqrt(-2.0*log(a)) * cos(2.0*3.1415926536*b);
+ else return 0;
+ }
+
+}
+#endif
diff --git a/video_input/consistencyChecker/consistencyChecker b/video_input/consistencyChecker/consistencyChecker Binary files differnew file mode 100755 index 0000000..8ec0203 --- /dev/null +++ b/video_input/consistencyChecker/consistencyChecker diff --git a/video_input/consistencyChecker/consistencyChecker.cpp b/video_input/consistencyChecker/consistencyChecker.cpp new file mode 100644 index 0000000..03b0009 --- /dev/null +++ b/video_input/consistencyChecker/consistencyChecker.cpp @@ -0,0 +1,114 @@ +// consistencyChecker +// Check consistency of forward flow via backward flow. +// +// (c) Manuel Ruder, Alexey Dosovitskiy, Thomas Brox 2016 + +#include <algorithm> +#include <assert.h> +#include "CTensor.h" +#include "CFilter.h" + +// Which certainty value motion boundaries should get. Value between 0 (uncertain) and 255 (certain). +#define MOTION_BOUNDARIE_VALUE 0 + +// The amount of gaussian smoothing that sould be applied. Set 0 to disable smoothing. +#define SMOOTH_STRENGH 0.8 + +// readMiddlebury +bool readMiddlebury(const char* filename, CTensor<float>& flow) { + FILE *stream = fopen(filename, "rb"); + if (stream == 0) { + std::cout << "Could not open " << filename << std::endl; + return false; + } + float help; + int dummy; + dummy = fread(&help,sizeof(float),1,stream); + int aXSize,aYSize; + dummy = fread(&aXSize,sizeof(int),1,stream); + dummy = fread(&aYSize,sizeof(int),1,stream); + flow.setSize(aXSize,aYSize,2); + for (int y = 0; y < flow.ySize(); y++) + for (int x = 0; x < flow.xSize(); x++) { + dummy = fread(&flow(x,y,0),sizeof(float),1,stream); + dummy = fread(&flow(x,y,1),sizeof(float),1,stream); + } + fclose(stream); + return true; +} + +void checkConsistency(const CTensor<float>& flow1, const CTensor<float>& flow2, CMatrix<float>& reliable, int argc, char** args) { + int xSize = flow1.xSize(), ySize = flow1.ySize(); + int size = xSize * ySize; + CTensor<float> dx(xSize,ySize,2); + CTensor<float> dy(xSize,ySize,2); + CDerivative<float> derivative(3); + NFilter::filter(flow1,dx,derivative,1,1); + NFilter::filter(flow1,dy,1,derivative,1); + CMatrix<float> motionEdge(xSize,ySize,0); + for (int i = 0; i < size; i++) { + motionEdge.data()[i] += dx.data()[i]*dx.data()[i]; + motionEdge.data()[i] += dx.data()[size+i]*dx.data()[size+i]; + motionEdge.data()[i] += dy.data()[i]*dy.data()[i]; + motionEdge.data()[i] += dy.data()[size+i]*dy.data()[size+i]; + } + + for (int ay = 0; ay < flow1.ySize(); ay++) + for (int ax = 0; ax < flow1.xSize(); ax++) { + float bx = ax+flow1(ax, ay, 0); + float by = ay+flow1(ax, ay, 1); + int x1 = floor(bx); + int y1 = floor(by); + int x2 = x1 + 1; + int y2 = y1 + 1; + if (x1 < 0 || x2 >= xSize || y1 < 0 || y2 >= ySize) + { reliable(ax, ay) = 0.0f; continue; } + float alphaX = bx-x1; float alphaY = by-y1; + float a = (1.0-alphaX) * flow2(x1, y1, 0) + alphaX * flow2(x2, y1, 0); + float b = (1.0-alphaX) * flow2(x1, y2, 0) + alphaX * flow2(x2, y2, 0); + float u = (1.0-alphaY)*a+alphaY*b; + a = (1.0-alphaX) * flow2(x1, y1, 1) + alphaX * flow2(x2, y1, 1); + b = (1.0-alphaX) * flow2(x1, y2, 1) + alphaX * flow2(x2, y2, 1); + float v = (1.0-alphaY)*a+alphaY*b; + float cx = bx+u; + float cy = by+v; + float u2 = flow1(ax,ay,0); + float v2 = flow1(ax,ay,1); + if (((cx-ax) * (cx-ax) + (cy-ay) * (cy-ay)) >= 0.01*(u2*u2 + v2*v2 + u*u + v*v) + 0.5f) { + // Set to a negative value so that when smoothing is applied the smoothing goes "to the outside". + // Afterwards, we clip values below 0. + reliable(ax, ay) = -255.0f; + continue; + } + if (motionEdge(ax, ay) > 0.01 * (u2*u2+v2*v2) + 0.002f) { + reliable(ax, ay) = MOTION_BOUNDARIE_VALUE; + continue; + } + } +} + +int main(int argc, char** args) { + assert(argc >= 4); + + CTensor<float> flow1,flow2; + readMiddlebury(args[1], flow1); + readMiddlebury(args[2], flow2); + + assert(flow1.xSize() == flow2.xSize()); + assert(flow1.ySize() == flow2.ySize()); + + int xSize = flow1.xSize(), ySize = flow1.ySize(); + + // Check consistency of forward flow via backward flow and exclude motion boundaries + CMatrix<float> reliable(xSize, ySize, 255.0f); + checkConsistency(flow1, flow2, reliable, argc, args); + + if (SMOOTH_STRENGH > 0) { + CSmooth<float> smooth(SMOOTH_STRENGH, 2.0f); + NFilter::filter(reliable, smooth, smooth); + } + reliable.clip(0.0f, 255.0f); + + reliable.writeToPGM(args[3]); + reliable.writeToTXT(args[3], true); +} diff --git a/video_input/consistencyChecker/consistencyChecker.cpp~ b/video_input/consistencyChecker/consistencyChecker.cpp~ new file mode 100644 index 0000000..0a8ea11 --- /dev/null +++ b/video_input/consistencyChecker/consistencyChecker.cpp~ @@ -0,0 +1,113 @@ +// consistencyChecker +// Check consistency of forward flow via backward flow. +// +// (c) Manuel Ruder, Alexey Dosovitskiy, Thomas Brox 2016 + +#include <algorithm> +#include <assert.h> +#include "CTensor.h" +#include "CFilter.h" + +// Which certainty value motion boundaries should get. Value between 0 (uncertain) and 255 (certain). +#define MOTION_BOUNDARIE_VALUE 0 + +// The amount of gaussian smoothing that sould be applied. Set 0 to disable smoothing. +#define SMOOTH_STRENGH 0.8 + +// readMiddlebury +bool readMiddlebury(const char* filename, CTensor<float>& flow) { + FILE *stream = fopen(filename, "rb"); + if (stream == 0) { + std::cout << "Could not open " << filename << std::endl; + return false; + } + float help; + int dummy; + dummy = fread(&help,sizeof(float),1,stream); + int aXSize,aYSize; + dummy = fread(&aXSize,sizeof(int),1,stream); + dummy = fread(&aYSize,sizeof(int),1,stream); + flow.setSize(aXSize,aYSize,2); + for (int y = 0; y < flow.ySize(); y++) + for (int x = 0; x < flow.xSize(); x++) { + dummy = fread(&flow(x,y,0),sizeof(float),1,stream); + dummy = fread(&flow(x,y,1),sizeof(float),1,stream); + } + fclose(stream); + return true; +} + +void checkConsistency(const CTensor<float>& flow1, const CTensor<float>& flow2, CMatrix<float>& reliable, int argc, char** args) { + int xSize = flow1.xSize(), ySize = flow1.ySize(); + int size = xSize * ySize; + CTensor<float> dx(xSize,ySize,2); + CTensor<float> dy(xSize,ySize,2); + CDerivative<float> derivative(3); + NFilter::filter(flow1,dx,derivative,1,1); + NFilter::filter(flow1,dy,1,derivative,1); + CMatrix<float> motionEdge(xSize,ySize,0); + for (int i = 0; i < size; i++) { + motionEdge.data()[i] += dx.data()[i]*dx.data()[i]; + motionEdge.data()[i] += dx.data()[size+i]*dx.data()[size+i]; + motionEdge.data()[i] += dy.data()[i]*dy.data()[i]; + motionEdge.data()[i] += dy.data()[size+i]*dy.data()[size+i]; + } + + for (int ay = 0; ay < flow1.ySize(); ay++) + for (int ax = 0; ax < flow1.xSize(); ax++) { + float bx = ax+flow1(ax, ay, 0); + float by = ay+flow1(ax, ay, 1); + int x1 = floor(bx); + int y1 = floor(by); + int x2 = x1 + 1; + int y2 = y1 + 1; + if (x1 < 0 || x2 >= xSize || y1 < 0 || y2 >= ySize) + { reliable(ax, ay) = 0.0f; continue; } + float alphaX = bx-x1; float alphaY = by-y1; + float a = (1.0-alphaX) * flow2(x1, y1, 0) + alphaX * flow2(x2, y1, 0); + float b = (1.0-alphaX) * flow2(x1, y2, 0) + alphaX * flow2(x2, y2, 0); + float u = (1.0-alphaY)*a+alphaY*b; + a = (1.0-alphaX) * flow2(x1, y1, 1) + alphaX * flow2(x2, y1, 1); + b = (1.0-alphaX) * flow2(x1, y2, 1) + alphaX * flow2(x2, y2, 1); + float v = (1.0-alphaY)*a+alphaY*b; + float cx = bx+u; + float cy = by+v; + float u2 = flow1(ax,ay,0); + float v2 = flow1(ax,ay,1); + if (((cx-ax) * (cx-ax) + (cy-ay) * (cy-ay)) >= 0.01*(u2*u2 + v2*v2 + u*u + v*v) + 0.5f) { + // Set to a negative value so that when smoothing is applied the smoothing goes "to the outside". + // Afterwards, we clip values below 0. + reliable(ax, ay) = -255.0f; + continue; + } + if (motionEdge(ax, ay) > 0.01 * (u2*u2+v2*v2) + 0.002f) { + reliable(ax, ay) = MOTION_BOUNDARIE_VALUE; + continue; + } + } +} + +int main(int argc, char** args) { + assert(argc >= 4); + + CTensor<float> flow1,flow2; + readMiddlebury(args[1], flow1); + readMiddlebury(args[2], flow2); + + assert(flow1.xSize() == flow2.xSize()); + assert(flow1.ySize() == flow2.ySize()); + + int xSize = flow1.xSize(), ySize = flow1.ySize(); + + // Check consistency of forward flow via backward flow and exlucde motion boundaries + CMatrix<float> reliable(xSize, ySize, 255.0f); + checkConsistency(flow1, flow2, reliable, argc, args); + + if (SMOOTH_STRENGH > 0) { + CSmooth<float> smooth(SMOOTH_STRENGH, 2.0f); + NFilter::filter(reliable, smooth, smooth); + } + reliable.clip(0.0f, 255.0f); + + reliable.writeToPGM(args[3]); +}
\ No newline at end of file diff --git a/video_input/deepflow2-static b/video_input/deepflow2-static Binary files differnew file mode 100755 index 0000000..8244834 --- /dev/null +++ b/video_input/deepflow2-static diff --git a/video_input/deepmatching-static b/video_input/deepmatching-static Binary files differnew file mode 100755 index 0000000..af0e4fc --- /dev/null +++ b/video_input/deepmatching-static diff --git a/video_input/make-opt-flow.sh b/video_input/make-opt-flow.sh new file mode 100755 index 0000000..e9ad60a --- /dev/null +++ b/video_input/make-opt-flow.sh @@ -0,0 +1,58 @@ +# Specify the path to the optical flow utility here. +# Also check line 44 and 47 whether the arguments are in the correct order. + +# deepflow and deepmatching optical flow binaries +flowCommandLine="bash run-deepflow.sh" + +if [ -z "$flowCommandLine" ]; then + echo "Please open make-opt-flow.sh and specify the command line for computing the optical flow." + exit 1 +fi + +if [ ! -f ./consistencyChecker/consistencyChecker ]; then + if [ ! -f ./consistencyChecker/Makefile ]; then + echo "Consistency checker makefile not found." + exit 1 + fi + cd consistencyChecker/ + make + cd .. +fi + +filePattern=$1 +folderName=$2 +startFrame=${3:-1} +stepSize=${4:-1} + +if [ "$#" -le 1 ]; then + echo "Usage: ./make-opt-flow <filePattern> <outputFolder> [<startNumber> [<stepSize>]]" + echo -e "\tfilePattern:\tFilename pattern of the frames of the videos." + echo -e "\toutputFolder:\tOutput folder." + echo -e "\tstartNumber:\tThe index of the first frame. Default: 1" + echo -e "\tstepSize:\tThe step size to create long-term flow. Default: 1" + exit 1 +fi + +i=$[$startFrame] +j=$[$startFrame + $stepSize] + +mkdir -p "${folderName}" + +while true; do + file1=$(printf "$filePattern" "$i") + file2=$(printf "$filePattern" "$j") + if [ -a $file2 ]; then + if [ ! -f ${folderName}/forward_${i}_${j}.flo ]; then + eval $flowCommandLine "$file1" "$file2" "${folderName}/forward_${i}_${j}.flo" + fi + if [ ! -f ${folderName}/backward_${j}_${i}.flo ]; then + eval $flowCommandLine "$file2" "$file1" "${folderName}/backward_${j}_${i}.flo" + fi + ./consistencyChecker/consistencyChecker "${folderName}/backward_${j}_${i}.flo" "${folderName}/forward_${i}_${j}.flo" "${folderName}/reliable_${j}_${i}.txt" + ./consistencyChecker/consistencyChecker "${folderName}/forward_${i}_${j}.flo" "${folderName}/backward_${j}_${i}.flo" "${folderName}/reliable_${i}_${j}.txt" + else + break + fi + i=$[$i +1] + j=$[$j +1] +done diff --git a/video_input/run-deepflow.sh b/video_input/run-deepflow.sh new file mode 100644 index 0000000..68fd593 --- /dev/null +++ b/video_input/run-deepflow.sh @@ -0,0 +1,10 @@ +if [ "$#" -ne 3 ]; then + echo "This is an auxiliary script for makeOptFlow.sh. No need to call this script directly." + exit 1 +fi +if [ ! -f deepmatching-static ] && [ ! -f deepflow2-static ]; then + echo "Place deepflow2-static and deepmatching-static in this directory." + exit 1 +fi + +./deepmatching-static $1 $2 -nt 0 | ./deepflow2-static $1 $2 $3 -match
\ No newline at end of file diff --git a/video_input/video.mp4 b/video_input/video.mp4 Binary files differnew file mode 100644 index 0000000..331cefb --- /dev/null +++ b/video_input/video.mp4 |