PerceptualDiff is an image comparison utility that makes use of a computational model of the human visual system to compare two images.
Added a few features at the request of the O3D team.
Added colorfactor and downsample options.
Also always output difference file if requested.
Always print out differing pixels even if the test passes.
Fixed bug in low light levels.
colorfactor - a float between 0 to 1. 0 means don't use color in determining differences.... read more
PerceptualDiff is an image comparison utility that compares two images using a perceptual metric. That is, it uses a computational model of the human visual system to determine if two images are visually different, so minor changes in pixels are ignored.
This version of perceptual diff has the file IO changed to use FreeImage so it supports a lot more file formats than before. Thanks for Jim Tilander for the patch.
PerceptualDiff, a command line utility that uses a computational model of the human visual system to compare images is now at version 1.0.1. It was recently adopted by the Cairo project for testing the Cairo graphics renderer (see http://cairographics.org/news/cairo-1.3.8\). This version includes a memory leak fix contributed by the Cairo team and fixes a linking error with libpng in order to get Fedora ext compliance.
Perceptualdiff, a command line utility that models the human visual system in comparing images, is released in version 1.0. Source code is provided, as are Intel compiled OSX and GNU Win 32 executables. This version supports TIFF and PNG files for input and an optional PPM file for output of the differences.
PerceptualDiff now supports PNG as well as TIF for input for perceptual difference and also outputs the difference in PPM format.
Perceptualdiff is a utility for comparing two images that uses a model of the human visual system for comparison. This version includes a color space test for pixels that pass the luminance test for stricter compares.
The simplest perceptual metric, which uses the Contrast Sensitivity Function and Threshold vs Intensity function on the luminance channel has been implemented.