A Markov Chain Monte Carlo based Rigid Image Registration Method

We propose a Monte Carlo Markov Chain (MCMC) based method for image registration. We formulate the image registration problem within a Bayesian framework and generate samples from the resulting posterior density of the registration parameters using MCMC. Thus, posterior density is characterized through the samples that are drawn with the MCMC principle. When the posterior density is multimodal, samples from different modes of the posterior lead to different and meaningful solutions for the image registration problem. We perform experiments on pairs of test images which may admit multiple registration solutions. Preliminary results demonstrate the potential of the proposed approach.