H.264预测编码算法的优化  

On Optimization of H.264 Predictive Coding Algorithm

作　　者：朱珉[1] 葛万成[1] 

机构地区：[1]同济大学中德学院,上海200092

出　　处：《通信技术》2013年第3期91-94,共4页Communications Technology

基　　金：德国罗德与施瓦茨公司资助

摘　　要：在相同的重构图像质量下,视频编码标准H.264与另两个标准H.263、MPEG-4相比,分别能节约50%、20%的码流,而且加强了对网络的支持,但其复杂度高。因此,这里对一种基于绝对变换差和(SATD)为预判断标准的快速帧内预测编码模式进行了改进,同时通过对前一帧的训练得到下一帧的阈值,从而对下一帧的宏块模式进行选择来减少帧间模式选择的计算量。仿真结果表明,在码率和PSNR基本不变的情况下,使帧内编码时间比原来下降50%左右,帧间预测模式编码时间减少到原来的40%多。Compared separately with H. 263 and MPEG-4, H. 264/AVC code could save 50% and 20% bit-streams with the same reconstruction video quality. In addition, it could improve its ability to support network applications, but however, at the cost of high complexity. The intra-frame coding is improved through pre-judgment based on SATD, while the calculation workload in inter-frame coding reduced via the training on previous frame in choosing the threshold of the next frame. The simulation result shows that the time in intra-frame coding declines to 50%, and the time in inter-frame coding to 40%.

关 键 词：H.264 算法优化 帧内 帧间 预测编码 

分 类 号：TN919.81[电子电信—通信与信息系统]