Sequential stack decoder for multichannel image restoration

Abstract

In this paper, we propose a novel scheme for image restoration (IR) employing a sequential decoding technique based on a tree search, known as Stack algorithm. The latter is a well-known method used for 1D signal decoding in wireless communication systems. The main idea is to extend the Stack algorithm for image restoration (2D) and to exploit the information diversity conveyed by the channels (Multichannel) in order to restore the original image. To deal with the noisy case, a regularization term is introduced using the total variation and the wavelet transform. This method was tested on artificially degraded images (blurred and noisy). Obtained results confirm the relevance of the proposed approach.