3D MR image restoration by combining local genetic algorithm with adaptive pre-conditioning

In this paper, we propose a novel efficient method by incorporating a local genetic algorithm and a new pre-conditioning technique into Markov random field model for image restoration. The role of genetic algorithm is to improve the quality of restoration and the pre-conditioning technique aims at accelerating the convergence. The remarkable advantage of our approach is that restoring corrupted images and preserving the shape transitions in the restored results have been orchestrated very well. The experiments on 3D MR image show that our method work very well