Add rudimentary ICC color profile support to colorhash

.travis.yml

@@ -38,12 +38,17 @@ install:
   - conda install --quiet --file conda-requirements.txt
   - pip install coveralls
 
+  # Work around conda's Pillow package lacking ImageCms
+  - conda uninstall pillow
+  - pip install pillow
+
   # Summerise environment
   # ---------------------
   - conda list
   - conda info -a
 
   # Install and test imagehash
+  # --------------------------
   - python setup.py install
 
 script:

imagehash.py

@@ -31,10 +31,11 @@
 
 from __future__ import (absolute_import, division, print_function)
 
-from PIL import Image
 import numpy
-#import scipy.fftpack
 #import pywt
+#import scipy.fftpack
+from PIL import Image, ImageCms
+
 __version__ = "4.1.0"
 
 """
@@ -270,7 +271,7 @@ def dhash_vertical(image, hash_size=8):
 	return ImageHash(diff)
 
 
-def whash(image, hash_size = 8, image_scale = None, mode = 'haar', remove_max_haar_ll = True):
+def whash(image, hash_size=8, image_scale=None, mode='haar', remove_max_haar_ll=True):
 	"""
 	Wavelet Hash computation.
 
@@ -320,7 +321,7 @@ def whash(image, hash_size = 8, image_scale = None, mode = 'haar', remove_max_ha
 
 
 
-def colorhash(image, binbits=3):
+def colorhash(image, binbits=3, ignore_icc=False):
 	"""
 	Color Hash computation.
 
@@ -332,11 +333,23 @@ def colorhash(image, binbits=3):
 	* the next 6*binbits encode the fraction in 6 bins of saturation, for mildly saturated parts of the remaining image
 
 	@binbits number of bits to use to encode each pixel fractions
+	@ignore_icc use raw color values, ignoring embedded ICC profiles
 	"""
 
+	if not ignore_icc:
+		image_profile = image.info.get("icc_profile")
+		if image_profile:
+			from io import BytesIO
+			# standardize color space to sRGB and preserve relative
+			# color values by using perceptual rendering intent
+			srgb_profile = ImageCms.createProfile("sRGB")
+			image_profile = ImageCms.ImageCmsProfile(BytesIO(image_profile))
+			ImageCms.profileToProfile(image, image_profile, srgb_profile,
+				renderingIntent=ImageCms.INTENT_PERCEPTUAL, inPlace=True)
+
 	# bin in hsv space:
 	intensity = numpy.asarray(image.convert("L")).flatten()
-	h, s, v = [numpy.asarray(v).flatten() for v in image.convert("HSV").split()]
+	h, s, _ = [numpy.asarray(v).flatten() for v in image.convert("HSV").split()]
 	# black bin
 	mask_black = intensity < 256 // 8
 	frac_black = mask_black.mean()

tests/test_colorhash.py

@@ -11,6 +11,9 @@ def setUp(self):
         self.image = self.get_data_image()
         self.func = imagehash.colorhash
 
+    def tearDown(self):
+        self.image.close()
+
     def test_colorhash(self):
         self.check_hash_algorithm(self.func, self.image)
 
@@ -63,6 +66,33 @@ def check_hash_size(self, func, image, binbits=range(-1,1)):
             with self.assertRaises(ValueError):
                 func(image, bit)
 
+    def test_colorhash_icc_support(self):
+        image = self.get_data_image('bigbuckbunny.png')
+        image_with_icc = self.get_data_image('bigbuckbunny-with-diff-icc.png')
+        with image, image_with_icc:
+            self.check_icc_support(self.func, image, image_with_icc)
+
+    def check_icc_support(self, func, orig_image, image_with_diff_icc):
+        orig_data_hash = func(orig_image)
+        diff_icc_hash = func(image_with_diff_icc)
+        emsg = ('same image data but with a different ICC profile should '
+                'result in different hashes: {} == {}'.format(orig_data_hash,
+                                                              diff_icc_hash))
+        self.assertNotEqual(orig_data_hash, diff_icc_hash, emsg)
+
+    def test_colorhash_icc_ignored(self):
+        image = self.get_data_image('bigbuckbunny.png')
+        image_with_icc = self.get_data_image('bigbuckbunny-with-diff-icc.png')
+        with image, image_with_icc:
+            self.check_icc_ignored(self.func, image, image_with_icc)
+
+    def check_icc_ignored(self, func, orig_image, image_with_diff_icc):
+        orig_data_hash = func(orig_image)
+        diff_icc_hash = func(image_with_diff_icc, ignore_icc=True)
+        emsg = ('ignoring the ICC profile in an image with the same image '
+                'data as one without a profile should result in the same hash: '
+                '{} != {}'.format(orig_data_hash, diff_icc_hash))
+        self.assertEqual(orig_data_hash, diff_icc_hash, emsg)
 
 if __name__ == '__main__':
     unittest.main()