Adaptive Intra Prediction Method Based On Intra Coding In H.264

This thesis focuses on the research of coding process of the video coding standard H.264/AVC, which is quite popular recently and put forward an original algorithm, called adaptive intra prediction method (AIP), in the field of intra coding process. In this thesis, the algorithm, which is based on an analysis of the rate-distortion of the prediction modes, is proposed to improve the coding efficiency and reduce the encoding complexity in the H.264/AVC coding. Tested by series of experiments, the proposed algorithm can achieve our purpose satisfactorily.In the procedure of video coding, there are two prediction methods, i.e. Intra prediction and Inter prediction. The former is used to remove the spatial redundancy in a frame, while the later removes the inter frame temporal redundancy in a sequence. Generally, inter coding is superior to intra coding in terms of the coding efficiency, because it offers a better prediction result. However, intra coding is essential for random access, abrupt scene changes, and error recovery. In the standard of H.264, it allows three block types for intra coding, i.e. Intra4×4, Intra8×8, and Intra16×16. Each of these Intra block types has several directional predictions modes, which use the neighboring decoded pixels for prediction. The best mode of these predictions is decided through the Rate-Distortion Optimization (RDO) process. The directional prediction modes need a lot of overhead bits to represent the modes for each block, which may reduce the coding efficiency. In the thesis, we gave an original adaptive intra prediction algorithm to improve the coding efficiency and reduce the coding complexity by means of reducing the number of bits representing the prediction modes of the intra block.After theoretic analysis, we gave out our original algorithm and henceforth we tested series of pictures by this algorithm. From the experiment results, the effect of our algorithm to improve coding efficiency was obvious, especially, in the field of reducing coding time, which will be positive power for web-video field.