Research On The Error Control Model And Techniques For H.26L Video Transmission Over Lossy Channels

With the development of network, communication and multimedia technology, the network multimedia, especially the network video applications, such as network videophone based on H.323/H.324, videoconference and video communication based on wireless network, have developed rapidly. But present network environment cannot meet the need of video transmission for there still exist some lossy channels. As a new video coding standard, H.26L becomes a suitable choice for video transmission on lossy channel because it has the advantages of high compression ratio, network adaptability and error robustness. But it still has some disadvantages in error control.This paper mainly studies on the error control techniques in H.26L video coding standard, which contains error resilience and error concealment. The error resilience techniques include resynchronization, data partition and INTRA macroblock update, which can restrain error scope and reduce the influence of error propagation. And the error concealment techniques include weighted pixel averaging algorithm based on INTRA frame coding and motion compensation using predicted motion vectors based on INTER frame coding. The defects of error concealment algorithms in H.26L are exposed by the results of simulation. Then a new improved scheme is proposed, and it is verified by the simulation results.To overcome the disadvantages of error control in H.26L, this paper discusses a useful channel code, Reed-Solomon (RS) Code. The RS Codec is implemented by programming and is used for channel error control in the video transmission system. It provides quality insurances for the video transmission over lossy channels because it plays an important part in the random and burst error correction. This paper also implements a table lookup algorithm to improve the codec efficiency.Finally, referring to the H.263 video transmission system, this paper discusses the optimization of the INTRA rate and the channel code rate in the video transmission system based on H.26L to achieve the best overall performance of decoded video. The model proves out to be feasible at low error rate by the simulation experiments. At last, the optimized model parameters are found for given network environment.