Research On Coding Block Mode Selection Algorithm In H.264

H.264/AVC is the latest international video coding standard developed by ITU-T and ISO/IEC. With the previous video coding standards, H.264 is still uses motion compensated hybrid video coding framework which based on previous video coding standard, it also uses a number of new technologies such as: enhanced motion prediction capacity, accurate matching of the smaller block transform, adaptive loop filter, entropy coding and so on. They improved the coding efficiency greatly. H.264 has enormous advantages: high compress efficiency, better image quality and excellent network performance, which caters to various applications, attracting much attention from research institution and multimedia industry. Rate control is an important technology in video coding, which should be researched further. It aims to achieve high video quality by allocating bit-rate fairly.However, because of the use of a wide range of advanced coding techniques,H.264 encoding complexity is also much higher than the previous standards, and the codec complexity are 3 times and 2 times of H.263 respectively. Therefore, the research and development of fast algorithm based on H.264, thereby reducing code complexity, and improving encoding speed, will have important theoretical and application value for the wide range applications of H.264, as well as further study and development of video compression and multimedia communication technology.In this paper, the key technologies used in H.264 encoding standard and rate-distortion optimization algorithm principle were introduced. The mode decision algorithm was briefly introduced in H.264 application based on JM. Research an analysis focusing on the intra/inter-frame mode decision algorithm and describe the utilization of mode decision algorithm. Combination of early termination and judging and choosing strategies, depending on the nature of video sequences encoding the different blocks to determine the image is the smooth movement or violent movement sequences. In the relatively complex texture, there is a relatively large proportion occupied by large block mode, in the relatively gentle texture, there is a relatively large proportion occupied by small block mode. This algorithm combined the strategies of early termination and judgment selection and other algorithms according to the threshold, the coded frames and the neighboring ones. It uses threshold information to select skip mode effectively, and skipping unnecessary block mode selection according to coded frame and coded block information, then selecting best coding block mode for current block adaptively. Experimental results show that under the cost that bit rate lowly increased, and the image peak signal to noise ratio reduces slightly, the encoding speed increases notably.In this paper, for the optimization of mode selection algorithm-level, reduce the complexity of model selection algorithm to improve encoding speed. The study on the coordination of the peak signal to noise ratio in images, the average bit rate, computational complexity and coding speed and other issues useful exploration.