AVS 中熵编码算法的研究与实现

AVS+(audio and video coding standard of china)是2012年颁发的新一代视频编码标准。CAVLC(基于上下文的自适应变长编码)和CABAC(基于上下文的自适应二进制算术编码)是该标准在熵编码部分采用的两种算法,相关领域对其研究甚少。分别介绍这两种算法的原理,并将其用C语言在x AVS中实现,最后通过实验对两种算法深入分析与测试。结果表明,CABAC算法复杂度较CAVLC稍有增加,但其获得的PSNR提高0.37 d B,码率降低11.4%,即编码压缩效果好,图像失真小,这种优势在低码率情况下尤其明显。综合两种算法利弊来看,CABAC算法的应用性和发展前景更广。

Research and Implementation of entropy coding algorithm in AVS

AVS , Advanced Audio and Video Coding in Broadcasting- Part 1: Video, is a new generation of video coding standard published in 2012.Two methods of entropy coding are adopted in this standard. One is Context-based Adaptive Variable Length Coding (CAVLC). Another one is Context-based Adaptive Binary Arithmetic Coding (CABAC). CAVLC is commonly used in H264 and AVS and it has already been implemented in C language in xavs. Since CABAC of AVS has rarely been studied in related field, the paper focuses on the algorithm principles of it and implements it in C language in xavs. Then a series of experiments are conducted on the two algorithms. The results of software test indicate that when compared with CAVLC, CABAC costs more time but it has better compression effect and less image distortion. And its advantages are more significant in the case of low bit rate. To sum up, CABAC has wider application and broader developing prospect, and it makes higher quality and lower bit rate image coding possible.