Unsupervised Image Segmentation using Markov Random Field Models

. We present an unsupervised segmentation algorithm based on a Markov Random Field model for noisy images. The algorithm finds the the most likely number of classes, their associated model parameters and generates a corresponding segmentation of the image into these classes. This is achieved according to the MAP criterion. To facilitate this, an MCMC algorithm is formulated to allow the direct sampling of all the above parameters from the posterior distribution of the image. To allow the number of classes to be sampled, a reversible jump is incorporated into the Markov Chain. The jump enables the possible splitting and combining of classes and consequently, their associated regions within the image. Experimental results are presented showing rapid convergence of the algorithm to accurate solutions. 1 Introduction The segmentation of noisy or textured images into a number of different regions comprises a difficult optimisation problem. This is compounded when the number of regions into..