Based On The H.264 Transform Coding And Entropy Coding

With the rapid development of computer network and multimedia techniques, the research about telemedicine based on Internet has a high-speed improvement. Video coding is one of the key techniques of telemedicine system. The research of this thesis is on the video coding techniques of international video coding Standard H.264. After the introduction of the basic theory of video coding, the development and technical characteristics of two series of international video coding standards, the thesis discusses the system framework, encoding process and several core techniques of video coding section of H.264,such as motion estimation and compensation, intra prediction, transform and quantization, deblocking filter and entropy coding.Then the major research work of this thesis on the transform, and entropy coding modules of H.264 was presented, including:Design a fast algorithm for the one-dimensional discrete cosine transform(DCT) though extracting multiplication factor from the transform matrix. This fast algorithm can effectively reduce computation of one-dimensional discrete cosine transform.Design integer discrete cosine transform according to the Standard H.264,and compare integer discrete cosine transform with general discrete cosine transform in two aspects: energy compaction and decorrelation efficiency.Research Context-based Adaptive Variable Length Coding(CAVLC) and Context-Based Adaptive Binary Arithmetic Coding(CABAC) which are adopted by H.264 as two efficient entropy coding. Describe the coding process of the two entropy coding algorithms in detail and compare the two entropy coding algorithms in complexity and compression efficiency.