H. 264 Simulation And Optimization Of Software Codecs

Entering 21 Century, the adoption of digital video has been fuelled by the development of many video coding standards in many applications. There are many video protocols such as JPEG, H.263, MPEG-1 and MPEG-2. which are targeted in different application areas. However the requirements for video transmission are increasing with the prevalence of Internet, so the video protocols have some limitation and localization nowadays. The H.264 is researched to be seasoned with low bitrate by VCEG and MPEG. As the latest standard in video codec field, H.264 has such advantage as high compression rate, strong error resilience capacity, highly adaptability and so on. In order to gain good coding result, H.264 adopts so many new algorithms that the amount of calculation increase drastically. As a result, its applications in PC environment are restricted. In order to realize H.264 real-time codec, many optimization works have to be done. But these optimization gains should not at the cost of obviously coding performance drop.This thesis first summarizes the international standards of video compression and codec'structure flow of H.264 and some new characters. It introduces intra prediction, inter prediction, transform, quantity, and code these modules' theory and character from H.264.And the codec process is realized on VC++6.0 environment. The paper Analyzes the relativity of motion estimation with emphasis in this foundation under the different stages between optimal modes relevance, and proposes improvement method of the block pattern choice on the foundation Baseline Profile of H.264, after improving it enable the speed of coding to have 10 ~ 20% enhancement. In order to further enhance the speed of codec for achieving real-time coding, this thesis introduces multimedia instruction collection such as MMX, SSE, SSE2 as well as SSE3 based on SIMD (Single Instruction Multi Data). And it further optimizes some time-consuming modules in the H.264 like SAD computation, the transformation, quantification and interpolation using these instruction collections. On the condition that quality of image is guaranteed it enhances the speed of encoder, the speed of encoder is enhanced 5-7 times comparing encoder that is not optimized , to the qcif sequence it may achieve real-time encoding an decoding.