Generalized adaptive edge-preserving image restoration algorithm

Discontinuities present serious difficulties to standard regularization, since standard regularization theory imposes global smoothness constraints on possible solution. We propose a noise-adaptive edge-preserving image restoration algorithm based on the Markov random field image model. Our potential function is controlled by the weighting function for providing the capability of adaptively introducing the discontinuities into the solution. Moreover a new parameter is adopted to prevent the undesirable amplification of strong noise. Extending our previous work, we propose a nonlinear formulation of the regularization functional and derive an iterative algorithm for ensuring the global minimum. The effectiveness of the proposed algorithm is demonstrated experimentally.