imgcompare

Calculates the difference between images in percent, checks equality with optional fuzzyness.

Algorithm

• Get diff image using Pillow's ImageChops.difference
• Convert the diff image to greyscale
• Sum up all diff pixels by summing up their histogram values
• Calculate a percentage based on a black and white image of the same size

Check the tests to see example diffs for different scenarios. Don't expect the diff of two jpg images be the same for the same images converted to png. Don't do interformat compares (e.g. JPG with PNG).

CAUTION: Due to a bugfix in the RGB -> L (Luminosity Greyscale) conversion in Pillow 7.0.0 the calculated outputs (percentages, etc.) differ between 6.2.2 and 7.0.0.

https://github.com/datenhahn/imgcompare/pull/3#issuecomment-576053494

Dependencies

• Pillow ( https://python-pillow.org/ )
  • imgcompare 0.1.0: no version checks
  • imgcompare 1.0.0: Python 2.7, Pillow <= 6.2.2
  • imgcompare 2.0.0: Python 3.x, Pillow >= 7.0.0, (calculated percentages will differ to Pillow 6.2.2)

Installation

pip install imgcompare

Usage

compare images

is_same = is_equal("image_a.jpg", "image_b.jpg")

use the tolerance parameter to allow a certain diff pass as same

is_same = is_equal("image_a.jpg", "image_b.jpg", tolerance=2.5)

get the diff percentage

percentage = image_diff_percent("image_a.jpg", "image_b.jpg")

or work directly with pillow image instances (also works for is_equal)

image_a = Image.open("image_a.jpg")
image_b = Image.open("image_b.jpg")
percentage = image_diff_percent(image_a, image_b)

Examples

On Image B the cat wears some fancy goggles which results in a small difference.

Slight Difference

Image A

Image B

imgcompare.image_diff_percent(JPG_CAT, JPG_CAT_SLIGHT_DIFF)

result => 0.344547385621

Difference between jpg and same jpg encoded again

Image B is Image A, but run again through the JPEG encoder, so Image B now has slightly more compression artefacts than Image A, which results in a small diff.

Image A

Image B

imgcompare.image_diff_percent(JPG_CAT, JPG_CAT_REENCODED)

result => 0.0149232026144

Black and White

When comparing a fully black image with a fully white one, the expected difference is 100%.

Image A

Image B

imgcompare.image_diff_percent(JPG_BLACK, JPG_WHITE)

result => 100.0

Black and Half-Black-White

Comparing a full black image with a half black, half white one gives a difference of 50%.

Image A

Image B

imgcompare.image_diff_percent(JPG_BLACK, JPG_HALF_BW)

result => 50.0

License

MIT License