Spatial-Temporal Adaptive Error Concealment Algorithm Based On H.264

Mankind in the process of awareness and understanding of the surrounding environment, more than70%of the information from the eyes, the image to give people more of the information, image communication service has always been the concern of the public. This visual information in the fields of computer vision and pattern recognition, is equal to analysis and processing the image and video. The rapid development of today’s society, has a higher demand for high-definition images and video, but the video quality enhancement requires tremendous network bandwidth support, so it is necessary to do image compression with few bits to represent. ISO and ITU has developed a series of standards for image and video coding, including H.261, H.263, H.264and MPEG-1, MPEG-2,MPEG-4, etc. H.264is the latest international standards, has been improved in many details as MPEG-4Part10, it’s recognized more promising than other coding standard and in order to enhance the compression ratio, removed a lot of redundant information in video transmission. But according to the information theory, we can know that because of the redundancy is removed, when the transmission data errors or packet loss, which will lead to the final decoded a serious decline of the quality, if we do not take the appropriate strategy. Therefore, this paper puts forward a highly adaptable spatial-temporal adaptive error concealment algorithm. This paper had a lot of data reading and analysis, proposed algorithm through simulation based on the H.264, final completion of the studies.This paper focuses on H.264error concealment code system to explain the spatial and temporal error concealment, put forward a new spatial-temporal adaptive error concealment algorithm and improved a spatial error concealment algorithm on this basis, the main research contents are the following:1. Detailed introduces the H.264coding standard and the key technology, and around the center of error concealment analysis the H.264coding system and design features2. On the basis of the information theory, video compression and transmission theory, introduce the classic error concealment techniques, especially, analyze the spatial-temporal error concealment, which adapt to the environment and the advantages and disadvantages. Describes the H.264’s code framework JM and carefully elaborate the error concealment module3. Analysis the shortage of traditional spatial error concealment algorithm, we put forward a recovery algorithm based on Markov random fields, because in restoring the lost macroblock used more the amount of information, the recovery of this algorithm is verified by simulation results of JM18.2is better than the traditional algorithm4. Since the spatial-temporal error concealment algorithm of their adaptability, this paper according to the video content and scene switching threshold processing method to realize the adaptive. Through analysis of the the damaged frame macroblock mode and more multi-threshold approach, using different algorithms for different scene types for damaged image reconstruction, have a good restore results.This paper through use JM18.2code framework at VS2005environment and MobilelP lost package software simulation analysis, the experimental results show that the proposed algorithm has better recovery effect and didn’t massive increase in the complexity of the algorithm, has high practical significance.