Image Restoration Based on Adaptive MCMC Particle Filter

In this paper, particle filter is applied in image restoration which can be posed as a recursive Bayesian estimation problem, in order to remove degeneracy phenomenon and alleviate the sample impoverishment problem, the convergence of Markov chain Monte Carlo (MCMC) method is introduced and used in resampling step, meanwhile a simple KLD sampling which separated from resampling step is combined to overcome the drawback of computational complexity by adapting the size of particle set. These improvements enhance the robustness, accuracy and flexibility of the particle filter, thus a new image restoration algorithm based on adaptive MCMC particle filter is proposed, the simulation results show the effectiveness of the proposed algorithm and present the superior performance over conventional particle filter.