视频多描述编码与差错隐藏的联合抗误码算法

Combat Channel Errors Algorithm Combining Multiple Description Coding and Error Concealment for Video

将MD SPIHT方法推广应用于视频序列 ,提出了视频多描述编码与差错隐藏的联合抗误码算法 ,以提高视频传输的鲁棒性 .该算法根据视频小波变换后的小波系数的时 -空关系 ,先对小波系数进行分组 ,再分别对各组进行 3D SPIHT编码 ,即产生多个描述 .多个描述通过多个独立的信道传输 .每一描述的码流是完全嵌入式的 ,易于实现质量可分级编码 .在接收端接收到的描述较少时 ,首先根据接收到的若干描述重构低质量的视频 ,然后采用差错隐藏方法恢复丢失的子波树 .对多个序列进行仿真实验 ,结果证实本文算法有效地提高了视频传输的鲁棒性 .

In this paper,the MD-SPIHT algorithm was extende d to video sequence,and a combat channel errors algorithm combining MDC(Multiple Description Coding) with ECN(Error Concealment) was presented for vi-deo in order to improve the robustness of the video transmission.In this new algorithm,the wavelet coefficients are divided into different groups according to their spatial-temporal relationships after the wavelet transformation,and each group was independently encoded by using the 3D-SPIHT algorithm,thus creation multiple description is which are transmitted through multiple independent channels .In addition,every description embedded bit-streams,and this makes it feasible to encode with quality scalability.When the end-user receives few descriptions,a low-quality image can be reconstructed according to the received descriptions,followed by the recovery of the lost wavelet tree by error concealment method.Simulation tests were finally carried out aiming at multiple sequences,with the conclusion that the proposed algorithm can effectively increase the robustness in video transmission.