frodon1/pdiffer

A Python interface to the PerceptualDiff tool.

pdiffer

A Python interface to the perceptualdiff tool, plus helpers for making pdiff-based assertions in tests. By Zack.

Install using pip:

pip install pdiffer

You need to have perceptualdiff installed somewhere on your computer for this library to work.

Usage

Create your PDiffer instance, optionally specifying a path to the binary:

from pdiffer import PDiffer
pdiff = PDiffer(bin='/usr/local/bin/perceptualdiff')
# or, just
from pdiffer import pdiff

Diff two images by pathname:

result = pdiff('image1.png', 'image2.png')

If the images are similar enough, result will be falsy; that is, the following code will print 'SAME':

>>> if result:
...     print 'DIFFERENT'
... else:
...     print 'SAME'
SAME

You can get the string output of the command by just calling unicode/str (or even print) on the result:

>>> print(result)
PASS: Images are binary identical

If the images are different, you'll see something else:

>>> result = pdiff('image1.png', 'image3.png')
>>> if result:
...     print(result)
FAIL: Image dimensions do not match

Parameters

There are a few parameters you can customize in the pdiff invocation:

pdiff('image1.png', 'image2.png',
    fov=65,
    threshold=40,
    gamma=2.2,
    luminance=100,
    luminanceonly=True,
    colorfactor=1.0,
    downsample=0,
    output='o.ppm')

Documentation on all of these options can be found by running perceptualdiff -help.

Test Assertions

PerceptualDiff is mostly useful in automated testing, and therefore some basic assertions are provided for checking image similarity (under both PEP8 and camelCase names):

from pdiffer import assertImagesSimilar, assertImagesDifferent
from pdiffer import assert_images_similar, assert_images_different


def test_something():
    assert_images_different('image1.png', 'image2.png')
    assert_images_similar('image1.png', 'image3.png')

These assertions take parameters just like pdiff(), e.g.:

assert_images_different('image1.png', 'image2.png', fov=89.9, threshold=40)

Unlicense

This is free and unencumbered software released into the public domain.

Anyone is free to copy, modify, publish, use, compile, sell, or distribute this software, either in source code form or as a compiled binary, for any purpose, commercial or non-commercial, and by any means.

In jurisdictions that recognize copyright laws, the author or authors of this software dedicate any and all copyright interest in the software to the public domain. We make this dedication for the benefit of the public at large and to the detriment of our heirs and successors. We intend this dedication to be an overt act of relinquishment in perpetuity of all present and future rights to this software under copyright law.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.

For more information, please refer to http://unlicense.org/