Error detection For H.264/AVC coded video based on artifact characteristics

This paper proposed a novel error detection algorithm of the H.264/AVC coded video stream based on the analysis of artifact characteristics. Firstly, the visual distortions caused by transmission are classified and construed. Then, two methods are implemented according to different kinds of visual artifact. Comparing with conventional ways, proposed algorithm not only considers the error happening on the boundary of the macroblocks(MBs), but also considers the error appears inside the MBs. Simulation results demonstrate that our algorithm shows a good performance for helping concealing the erroneous MBs to gain a high video quality.