This code demonstrates how to compute gray level cooccurrence matrices for regions centered around each pixel of an image using the integral image principle.
This principle allow to reduce the time required to compute all matrices, but may use a lot more memory, since we need to keep in memory a cooccurrence matrix for each pixel of the image.
This tool is released under the terms of the MIT License. See the LICENSE.txt file for more details.