Researches On Error Resilience Techniques In Wireless Video Transmission Based On H.264/AVC

The rapid progress of wireless communication technology, multimedia technology and digital signal processing technology has brought about the great development of wireless video communication. The third and the forth generation of mobile communication technology could provide high speed video services. However,due to random errors and burst errors brought by the multi-path effect and fading in wireless channels, the video quality decrease greatly. Additionally the compressed bit-streams is vulnerable to errors, which further deteriorates the video quality. This imposes more challenges on research of robust video transmission over wireless error-prone environment.The newest video coding standard H.264/AVC developed by ITU-T and ISO, compared with standard before, which has the characteristic of high compression rate and robust adaptability for different networks, which is very suitable for transport in wireless channels.Based on the advanced H.264/AVC video standard, this dissertation has conducted detailed research focusing on error resilience issues of video transmission over wireless channels. Firstly, the propagation and influence of channel errors is analyzed and the problems existing in wireless video transmission are pointed out and typical error control technologies are compared. Then, we introduced the basic video coding theories and video tools of H.264/AVC and the error resilience techniques are described based on wireless applications. At last, after deeply research on the error resilience techniques in the decoder, two error concealment methods are proposed.Against the problems of image blurring in current H.264 Joint Model concealment algorithms, an effective spatial error concealment method based on edge detection is proposed, which classified the error block into different categories by using edge information extracted from the restricted of surrounding blocks. A suitable method is chosen for different error block. Simulation results show that the proposed method yields better video quality than conventional approaches.Against the condition of missing frames which occur frequently in wireless low bit-rate packet-based video communications, a new improved algorithm based on motion vector extrapolation are proposed, which uses not only the extrapolated MVs of the pixels, but also the extrapolated MVs of the blocks, from which a more accurate MVs of the pixels in the missing frames is obtained. Then an adaptive method of error concealment for whole frame loss is brought in, and a suitable method is chosen by the motion activity. Experimental results show that this proposed algorithm is highly effective and significantly outperforms other existing frame recovery methods with low computation.