A Bayesian approach to blind deconvolution based on Dirichlet distributions

This paper deals with the simultaneous identi cation of the blur and the restoration of a noisy and blurred image. We propose the use of Dirichlet distributions to model our prior knowledge about the blurring function together with smoothness constraints on the restored image to solve the blind deconvolution problem. We show that the use of Dirichlet distributions o ers a lot of exibility in incorporating vague or very precise knowledge about the blurring process into the blind deconvolution process. The proposed MAP estimator o ers additional exibility in modeling the original image. Experimental results demostrate the performance of the proposed algorithm.