Research On Error Concealment Technique In H.264/AVC

With the technological development of communications and networks, video streaming services are playing a more and more important role in the field of network communications. The Moving Picture Expert Group (MPEG) and VCEG worked together as a Joint Video Team (JVT). They created a single technical design called H.264/AVC for an ITU-T Recommendation H.264/AVC and for a new part of the MPEG-4 standard called AVC. Compared with H.263+ or MPEG-4, H.264 permits a reduction bit rate by 50% for a similar degree of encoder optimization at most bit rates. H.264 bit streams can be easily transported over different networks through the Network Adaptation Layer (NAL). H.264 offers good video quality and low bit rates because it has more creative algorithms. Meanwhile, more efficient algorithms lead to more complexity and more correlation. So any bit error or package loss in transmission channel may be a significant error in decoder or an acute video quality loss. So error control over unreliable network is the most important thing in a robust system.First, this paper introduced the background, application and overview of video coding. It elaborated all error resilience related techniques generally and classifies them. Deeper research was taken on error concealment for H.264. In order to improve the error concealing performance of H.264/AVC, a new algorithm was presented. At the encoder, the proposed scheme first extracted the important data from each Macro Block (MB), which can be used for error concealment at the decoder. These data were embedded into the next frame using the odd-even data embedding scheme in an MB-interleaving slice-based manner. At the decoder, when the decoder could not receive the right MB, it would extract the embedded data to conceal the corrupted MB. This technique has following merits: (1) easy to implement, (2) obviously improve the efficiency of video error-recovery, and (3) needs little redundancy bits. Experimental results show that the error concealing performance has improved greatly than that of the JM model, both subjectively and objectively.Finally this paper gave the conclusion and future work.