Research On Fast Algorithms Of Intra-prediction Mode Decision In H.264

With the requirement of consumers, now High Definition video has been used in variable recreation facilities as the video compression standard. In order to meet the requirements, H.264is a new video compression standard with higher compression rate compared with other compression standards.Although H.264standard has very high compression rate, however, its complexity increases highly. Therefore, intra prediction mode decision has become a bottleneck of the H.264compression standard. So it is important to consider about the fast algorithms.Now in H.264standard, RDO is used for intra-prediction mode decision. Although RD-OPT can increase SNR and bit rate, but it may also lead to computation complexity increase at the same time. From the point of coding time, applying RDO also increase the processing time. How to apply fast algorithm has been a heated topic in the whole world. But most of the algorithms can only decrease the complexity with low quality of frame. Therefore, we need to pay attention to two points, one is based on current algorithms, how to reduce the time complexity. Avoiding to decrease the quality of reference frame, fast algorithms for intra-prediction should be considered. The other one is that in current JM, only SSIM is used as the judgment measurement. In this paper, I proposed an algorithm for introducing SSIM to intra-prediction mode decision. We proposed SSRDO, which combines SSE with SSIM. Instead of traditional SSE, we can get better frame quality. This method is more suitable for video compression.Finally, the simple experimental results show that the proposed fast algorithms decrease the time complexity, under almost the same PSNR.The result shows that the fast algorithm can reduce the complexity by about70%, but the bit rate increases less than5%and PSNR is almost the same. The performance is better than current H.264.