基于有损率失真优化的AVS—M帧内更新算法

基于运动补偿混合编码的AVS—M低速率视频流，对信道误码十分敏感．网络丢包造成的错误信息不仅影响当前帧的重建质量，还会在其后续帧的时域和空域上迅速蔓延和扩散，使得视频质量急剧下降．为此，分析了AVS—M无损率失真优化和有损率失真优化模型，研究了受损MB的扩散失真和掩盖失真，通过比较在模拟丢包网络中MB在不同编码模式下的Lagrangian代价函数，实现了一种基于有损率失真优化的AVS—M帧内更新算法．同时，将该算法与AVS—M的随机帧内更新以及行更新算法进行了比较．实验表明，基于有损率失真优化的帧内更新算法，在计算复杂度增加不大的情况下，较大地提高了AVS—M视频的差错恢复性能．

Intra-Refreshing Algorithm of AVS-M Based on Rate-Distortion Optimization with Losses

The low bit-rate video stream compressed by motion-compensated hybrid codes such as AVS-M is vulnerable to error-prone networks. The errors caused by packet loss not only impair the reconstruction quality of current frame, but also lead to temporal and spatial error propagation to subsequent frames, which may result in visual drop- ping sharply. Models of rate-distortion optimization （RDO）with and without losses were analyzed, and the propagation distortion and concealment distortion of damaged macroblock（MB） were investigated. An intra-refreshing algorithm of AVS-M was realized based on RDO with losses by comparing values of Lagrangian cost function of MB in different coding modes. Meanwhile, random intra-refreshing and line refreshing algorithms were compared with the proposed algorithm. The experimental results show that superior performance of error resilient can be achieved using the proposed algorithm without increasing computational complexity remarkably.