H.264 Error Concealment Based On Research And Application Of The Algorithm

With the advent of information age, the video service quickly accepted by people rely on its intuitive, certainty, efficiency and extensive quickly, and it has become a part of people’s daily lives. With the video development from SD to HD, the amount of video-information is rapidly increasing. but the network bandwidth is limited, which requires the high-efficiency to compress video data. the new generation video coding standard h.264has good compression performance and better network affinity perfect to solve the above problem, so this paper will adopt the h.264standard as a research object.Due to channel noise interference, when the video code stream transmission in communication channel, the stream data can easily cause data missed or error, and forming error code stream in the end. If used the regular way to decoded the error code stream, it will inevitably reduce the decoded image quality greatly. How to recover the missed or error data, has become a hot research. In this paper, the decoded side of error concealment techniques is the content for the study, the inter-frame error concealment is the direction of research, the classical boundary matching algorithm is the entry point for the research, the ultimate purpose of paper is engineering application.In this paper, first of all, it briefly introduces the video compression technology, H.264codec features and principles, and the key technologies which used in H.264. then it In-depth analyzed the common error concealment algorithm which used in H.264. Finally, aiming at the deficiency of the boundary matching algorithm in image detail piece, the edges of objects and moving slice in picture have poor hidden ineffective. the main Optimized of the proposed algorithm is reflected in the following four aspects:1) Expand and modify the set of candidate motion vectors, the expansion is based on multi-reference prediction code technique as theoretical basis, the modified is to prevent the introduced of unnecessary quadric error.2) Modify the constraint conditions to obtain the best motion vector and the optimal motion compensated block. 3) Selective introducted a linear interpolation to obtain damaged macro-block residual estimate, better hide lost macro-block is filtered by set a threshold, then focus on processed contains moving object in lost macro-block.4) Using residual recovery in the motion compensation block to obtain more accurate reconstruction macro-blocks, and then in replaced of damaged macro-block,realize the error concealment.Finally, the improved algorithm evaluated in that the quality of image restoration and the complexity of the algorithm. The improved decoder will be get by embedding and transplanting the cropped improved algorithm. Through before and after improvement real-time-play comparison to verify error concealment performance for the improved decoder.