| In recent years, the excellent compression efficiency of H.264 has attracted plenty of attention, and its advantages, such as compression ratio, fidelity and network adaptability, make video experts more willing to use it. Currently, H.264 mainly used in streaming media and video broadcasting, however, as increasing demand of video compression, the contradiction between high complexity and limited processing speed of traditional H.264 has been more and more obvious. How to reduce the complexity of H.264 and achieve application of high-speed and high-fidelity video compression has become an urgent problem.This paper introduces a method that can resolve the High-Speed and High-Fidelity compression encoding problem by using H.264 encoding technology. Firstly, based on the study of x264, which is an open source encoder, this paper completed the extraction of I frame encoder through the reorganization, the deletion and improvement of a few modules in the x264 encoder. Secondly, Because of the redundant computing that original H.264 standard has brought by using the overall search method in High-Speed and High-Fidelity application, HSHF Intra Improved Algorithm has been proposed in this paper. This algorithm uses a series of fixed prediction models to compute scheduled by fixed regions, which make a significant reduction in redundancy in I frame encoding, compared to the I frame only encoder, the proposed encoder gain a more than 1-2 speed increase and bring no more than 0.1 dB decrease in PSNR; Finally, the designed encoder has been transplanted to TMS320DM6446 affiliate to DaVinci platform, and complete the platform-independent encoding function.Through the encoding compression comparison made by optimized x264, non-optimized x264 and only I frame encoder, the proposed encoder has more than 6 times gain in encoding speed,41dB or more in PSNR,8:1 or more in compression ratio, and especially effective in large image, little content and intensive sequences. |
PDF Full Text Request