Blind Image Restoration Using Improved APEX Method with Pre-denoising

Abstract

Loss of image quality due to degradation has seriously bad effect on the post-analyzing and postprocessing of image. Blind restoration of degraded image is usually required to improve the image quality, for the reason that the degradation is partially or completely unknown in many practical situations. Commonly, using the APEX method can effectively accomplish the blind restoration when a significant class of degradation (Levy stable processes) is considered, but noise is unexpectedly amplified in the process. In this paper, a novel method is proposed to both suppress the noise effect and improve the image quality. The performance of the method is investigated with experiments on real degraded images. The noniterative property of this method makes its application and implementation in real-time image processing systems be easy.