| With the development of video coding and communication technology, video network transmission becomes a reality, and widely development. But in video network transmission, video data loss, due to lost package or channel failure (wireless network), will reduce the quality of the reconstruction video. In order to ensure the quality of the reconstruction in error channels, the error resilience technology must be used. So the technique to combat channel disturbances has become a hotspot in current video transmission technology. Multiple description coding (MDC) has the ability to provide effective error resilience, while ensure real-time transmission, so becomes an effectively fight miscode technology.In order to enhance the robustness of the real-time video picture in wireless network, this paper makes a research in multiple descriptions coding (MDC). The main work in this paper includes:(1) Introduces H.264/AVC video compression coding standard and multiple description coding as the key point, and analyzes several most popular multiple description coding methods.(2) This paper mainly focuses on multiple state video coding (MSVC) which is compatible to H.264/AVC, then analyzes an improvement method multiple-state video coding with redundant picture (MSVC-RP). Through the experimental simulation under JM, analyzes the problems of MSVC-RP.(3) Aiming to the redundant problem in MSVC-RP, while considering the characters of the Humans Visual System, this paper designs an improved algorithm. For the picture with redundant slice coding mode, first determine the redundant coding mode via the determined threshold, in order to add redundancies more effectively and protect the region of interests (ROI).Through the comparison of the proposed algorithm and MSVC-RP methods, we can conclude that the improved algorithm proposed in this paper can provide better protection of the important area in the image than MSVC-RP, and obtain good rate distortion effect in 5% loss rate. |
PDF Full Text Request