Research Of H.264/AVC Coder And Its Correlative Algorithm

With the explosive growth of the demand for video and audio information, the pursuing of interactivity of video and audio over long-distance has been fashioned.In recent years, based on the develop of computer science, communication and networking, multimedia information which integrated audio, video, images, texts and data, has facilitated more and more people by bringing the joys of cyber-life, tele-medical and telecommunications, which shortened the distance over time and space. Among this, digital TV(including HDTV), compressing and storage of DVDs, video conference and 3G is the focus of research and one of the trend of multimedia technologies.Therefore, ITU-T and ISO/IEC, with the tenet to design a simple and effective new codec technology, with the goal to get outstanding compress performance and excellent network adaptation, issued the H.264 video compress standard in May 2003. In most code schemes, H. 264 is the most effective standard of the code and decode at present. With the same quality of the refactoring image, H.264 can saving 50％of the data stream of the H.263 and MPEG-4 standards. Because H.264 can cover all the application of low broadband and high broadband and has the feature of strong strong error resilience. H.264 is especially suitable for the wireless video transmission which is low broadband and has high rate of losing package and serious interfering. But the enhancement of compressing is based on rapid increasing of compute complexity, which has been the main hindrance of massive application.Aiming at the status, many researches have been done and being done to decrease compute complexity and together improve running efficiency. Our work mainly focused on the following:1. Analyzed the standard in detail and compared with other standards comprehensively, to establish the emphases and orientation of further study. 2. we propose an efficient 4×4 intra prediction mode decision method for the complexity reduction of H.264 video coding. A good prediction could be achieved in transform domain when local edge direction of a block is in the same direction as the intra prediction mode, could filter out majority of intra prediction modes. By filtering out majority of intra prediction modes, we only have to consider a candidate set of most highly probable 4×4 intra prediction modes for which the rate distortion optimization process should be performed.3. To reduce the complexity of inter prediction,a fast mode decision algorithm using early SKIP mode decision and motion cost to reduce the candidate modes is proposed in this paper.4. To reduce the complexity of multiple reference frames, a fast selection algorithm of reference frame is presented. According to the actual characteristics of chosen reference frame, it utilizes the texture characteristic of sequence to estimate the possible number of reference frames, and further determine the best reference fiame of 16×16, 8×8 and 4×4 modes according to the motion information and the monotony of RDO among different reference frames. So the likely reference frames of other modes may be adaptively chosen according to the correlation of reference frame selection with the above modes. It can decrease the number of searched frames and avoid redundant reference frames.