Counters chickpeas and lentils in a BMP image. Works with two classes. Tested with 768x728 image. May not work on larger images.
Program itself uses K-Means algorithm to convert intensity image to binary representation.
Edge detection or boundary extraction is implementing with erosion and dilation features.
All objects in image are detecting with CCA which is Connected Component Analysis algorithm that searching for a neighborhood between pixels. Colorization is implementing according to tag array returned from CCA function.
