A new, efficient and flexible error detection approach for compressed visual contents

Abstract

In this paper, we propose an efficient and flexible error detection (ED) approach for compressed visual contents. We noticed that the major influence of channel errors is to make the subsequent symbols lose synchronization so that wrong symbols are extracted from the bitstream. Hence, checking the start position of a certain symbol can tell correctness of all of its previous dependent symbols. By utilizing this intrinsic feature in compressed image/video bitstreams, our proposed scheme achieves more efficient error detection than conventional approaches. Moreover, the generated checking information can also be adjusted flexibly without affecting the original bitstreams, thus making our scheme more suitable for practical streaming systems.