It is a method for segmenting an image based on the shape and size of specific features within the image.