H.264 Video Coding Algorithm Research And DSP Optimization

With the popularity of multimedia products and the development of digital technology, video encoding and transmission technology is widespread concern. In recent years,a lot of video encoding standards such as MPEG-1, MPEG-2, H.261, H.263. And H.264 is a new generation of video coding standard proposed by Joint Video Team (JVT) which is united by International Telecommunication Union (ITU) and International Organization for Standardization (ISO). Compared to previous generations have a better standard of the encoding efficiency and network compatibility and broader prospects. H.264 compression performances efficiently, but the complexity of encoder is much higher, how to improve the H.264 encoding speed, real-time encoding and transmission, has become a very challenging area of video coding issue.This paper studied the H.264 coding algorithm to optimize and improve. Firstly analysised of the key technologies of h.264, such as intra-coding, frame coding, transform and quantified, entropy coding, SI/SP frame technology. This paper did a deeply analysis and in-depth study on the intra-coding, and proposed a new intra prediction algorithm. According to the algorithm of the relevance of adjacent block and enhanced Kirsch operator, On one hand, it reduced unnecessary computation effectively; On the other hand, it saved substantial course of Conditions to determine of the Traditional Algorithm based on edge direction and improve the efficiency of the code in the DSP implementation.Experimental results showed that the algorithm can not affect the basic quality of the image and greatly enhanced the intra-coding efficiency, reduced the encoding time. This article also analyzed the CW5521 chip structure and introduced a optimization method, based on DSP chip and gave a briefing coding optimization approach. For the part of intra-coding, this article make structural adjustment of some key functions in order to make full use of chip resources, and enhance the ability of parallel execution pipeline through code optimization and other methods.