An iterative method for Bayesian Gauss–Markov image restoration

The purpose of this paper is to introduce a new method for the restoration of images that have been degraded by a blur and an additive white Gaussian noise. The model adopted here is assumed to be Bayesian Gauss–Markov linear model. By exploiting the structure of the blurring matrix and by using Kronecker product approximations, the image restoration problem is formulated as matrix equations which will be solved iteratively by projection methods onto Krylov subspaces. We give some theoretical and experimental results with applications to image restoration.