MAP-MRF approach for image denoising

Abstract

Digital images gets degraded due to several reasons. The aim of denoising involves the restoration of signal to yield visually good quality representation. In this paper, we are referring to an image corrupted by additive Gaussian noise. The image is modeled as MRF and maximum-a-posteriori (MAP) estimate is obtained using graduated non-convexity technique. Results suggest that the proposed technique yields better compared to other techniques.