H.264中帧内预测模式选择算法的研究

H.264标准中的帧内预测模式在带来高效的压缩编码性能的同时,也带来了极大的运算量。针对这种情况,在对现有的优化思想进行研究的基础上,提出一种新算法,通过对各种预测模式的方向性进行分析,利用相邻模式之间的相关性,跳过一些小概率模式,从而降低模式选择的复杂度。通过仿真实验,对两种算法的编码时间、比特率以及PSNR进行了比较。实验结果表明,在采用新的模式选择算法后,运算复杂度大大降低,极大地减少了编码时间,有效地提高了视频压缩编码效率。

Research on Intra-predictio Mode Decision Algorithm Based on the Relativity of the Adjacent Models

In H. 264 /AVC,intra-prediction achieves significant coding performances but also involves extremely high computational complexity. A new algorithm is proposed after analyzing several optimization algorithms. For the intra mode prediction,the computational complexity of mode decision can be reduced by utilizing the correlation of adjacent mode and skipping some modes of low probability. Contrasts with the image vision quality,coding time,bits and changes of PSNR are made in the simulation before and after optimizing it. Finally,the experimental result shows that the improved fast prediction mode decision algorithm is able to reduce encoding time obviously and enhance the efficiency of coding with no decline in the quality of images.