Segmenting color images using Markov random field models

We use Markov random field models for color images in conjunction with a three phase segmentation algorithm based on region splitting, conservative merging, and agglomerative clustering. At each step, the agglomerative clustering phase maximizes a global performance functional based on the conditional pseudo-likelihood of the image. A test for stopping the clustering is applied based on rapid changes in the pseudo-likelihood of the image. We provide experimental results to demonstrate the performance of the segmentation algorithm on color images of natural scenes. Most of the processing during segmentation is local making the algorithm amenable to high performance parallel implementation