3 files changed, 51 insertions, 5 deletions
diff --git a/check/app/server/api.py b/check/app/server/api.py
index 1b17a1e..c4f9f80 100644
--- a/check/app/server/api.py
+++ b/check/app/server/api.py
@@ -5,7 +5,6 @@ import numpy as np
 from flask import Blueprint, request, jsonify
 from PIL import Image
 
-# from app.utils.im_utils import pil2np
 from app.models.sql_factory import search_by_phash, add_phash
 from app.utils.im_utils import pil2np
 
@@ -42,9 +41,10 @@ def upload():
     })
 
   im = Image.open(file.stream).convert('RGB')
-  im_np = pil2np(im)
+  phash = compute_phash_int(im)
 
-  res = search_by_phash(im_np)
+  threshold = request.args.get('threshold') || 6
+
+  res = search_by_phash(phash, threshold)
 
-  # get threshold
   return jsonify({ 'res': res })
diff --git a/check/cli_phash.py b/check/cli_phash.py
index c5df139..169dfa5 100644
--- a/check/cli_phash.py
+++ b/check/cli_phash.py
@@ -20,7 +20,7 @@ cc = ClickSimple.create(cfg.DIR_COMMANDS_PHASH)
   help='Verbosity: -v DEBUG, -vv INFO, -vvv WARN, -vvvv ERROR, -vvvvv CRITICAL')
 @click.pass_context
 def cli(ctx, **kwargs):
-  """\033[1m\033[94mMegaPixels: Dataset Image Scripts\033[0m                                                
+  """\033[1m\033[94mVFrame Check Image Deduplication API\033[0m                                                
   """
   ctx.opts = {}
   # init logger
diff --git a/check/commands/phash/dedupe.py b/check/commands/phash/dedupe.py
new file mode 100644
index 0000000..3cf60d4
--- /dev/null
+++ b/check/commands/phash/dedupe.py
@@ -0,0 +1,46 @@
+"""
+Dedupe a folder of images
+"""
+
+import click
+import os
+import glob
+
+from PIL import Image
+
+from app.utils.im_utils import compute_phash
+
+@click.command()
+@click.option('-i', '--input', 'opt_input_glob',
+  required=True,
+  help="Input glob to search -- e.g. '../docs/images/*.jpg'")
+@click.option('-t', '--threshold', 'opt_threshold',
+  required=True,
+  default=6,
+  type=int,
+  help="Threshold for PHash hamming distance comparison (0-64, default=6)")
+@click.pass_context
+def cli(ctx, opt_input_glob, opt_threshold):
+  """
+  Dedupe a folder of images
+  """
+  seen = []
+  total = 0
+  for fn in sorted(glob.iglob(os.path.expanduser(opt_input_glob))):
+    total += 1
+    im = Image.open(fn).convert('RGB')
+    phash = compute_phash(im)
+    if is_phash_new(fn, phash, seen, opt_threshold):
+      seen.append({
+        'phash': phash,
+        'fn': fn,
+      })
+  print("checked {} files, found {} unique".format(total, len(seen)))
+
+def is_phash_new(fn, phash, seen, opt_threshold):
+  for item in seen:
+    diff = item['phash'] - phash
+    if diff < opt_threshold:
+      print("{} === {} (diff: {})".format(fn, item['fn'], diff))
+      return False
+  return True