Python Image Processing Cookbook

How it works...

If you run these code blocks, you will get the following output:

The random_walker() function from scikit-image segmentation was used to implement the random walker algorithm for segmentation from the markers. We used a colored input image, so we set the multichannel parameter to True.

The return_full_prob parameter to the same function was assigned to True in order to obtain all the probability values of a pixel belonging to each label. When this parameter is False (which is the default case), it returns only the most likely label.

The algorithm solves the infinite time diffusion equation, with the sources placed on the markers ...