Hidden Markov Random Fields and Swarm Particles: A Winning Combination in Image Segmentation

Segmenting an image, by splitting this latter into distinctive regions, is a crucial task in many nowadays ubiquitous applications. Several methods have been developed to perform segmentation. We present a method that combines Hidden Markov Random Fields (HMRF) and Particle Swarm Optimisation (PSO) to perform segmentation. HMRF is used for modelling the segmentation problem. This elegant model leads to an optimization problem. The latter is solved using PSO method whose parameters setting is a task in itself. We conduct a study for the choice of parameters that give a good segmentation. The quality of segmentation is evaluated on grounds truths images using Misclassification Error criterion. We use the NDT (Non Destructive Testing) image dataset to evaluate several segmentation methods. These results show a supremacy of the HMRF-PSO method over threshold based techniques.