Performance study of an image restoration algorithm for bursty mobile satellite channels

Abstract

Error-concealment techniques provide a simple framework to compensate transmission distortions without incurring additional delays and wasting bandwidth resources which are crucial for real-time applications over networks with limited resources. In [1], we have presented an error concealment technique utilizing discrete wavelet transform (DWT) for embedding macroblock-based best-neighborhood-matching (BNM) information into the original image in order to utilize spatial redundancy. In this paper, we investigate the performance of Best-neighborhood-matching and Wavelets based Error-Concealment (BWEC) algorithm in bursty mobile satellite channels. We show that this technique is specifically promising for bursty mobile satellite channels suffering a wide range of packet losses, at the expense of some degradation in the perceptual quality of the original image and computational burden. © Springer-Verlag Berlin Heidelberg 2005.