Optimization Algorithm Of Fast Block Partition And Mode Decision For HEVC Intra Prediction

With the development of high-definition video, the requirement of channel become higher and higher in the process of image transmission, so the requirement of image compression algorithms become more sophisticated. As a popular image compression algorithm, H.264 has prevailed at home and abroad for many years, but the compression effect is not very good, especially for the high-definition video. That is to say, H.264 cannot meet the increasing demands. So H.265 emerges at the right moment, H.265 / HEVC(High efficient video coding), serving as a new generation of video coding technology, has many advantages. Compared with H.264, the compression efficiency of H.265 increases a lot, especially for high-definition video.Although HEVC improves the compression efficiency greatly, its complexity is also improved a lot. A huge amount of calculation in intra prediction has to be done, the process of CU block partition and PU mode decision is very complex. Thus it is quite important for decreasing the complexity of intra prediction and reducing coding time.Aiming at the CU block partition and PU mode decision in intra prediction, this paper proposes some optimization algorithms for CU block partition, PU mode decision and merging version based on the theoretical analysis. These algorithms reduce coding time and the complexity of intra prediction without affecting the performance of coding. In this paper, these algorithms are verified respectively from the perspective of objective quality and subjective quality. Consequently, the effectiveness of these algorithms is demonstrated.