联合率失真模型的帧级码率控制方法  

Joint rate distortion models rate control method for frame level

作　　者：何书前[1,2] 邓正杰[2] 石春[2] 

机构地区：[1]中山大学信息科学与技术学院,广州510275 [2]海南师范大学信息科学技术学院,海口571158

出　　处：《计算机工程与应用》2013年第23期19-23,共5页Computer Engineering and Applications

基　　金：海南省自然科学基金项目(No.611128;No.612122;No.613164;No.613163)

摘　　要：在H.264/AVC视频编码框架下,基于联合率失真模型,提出了一种新的帧级码率控制方法。利用分块数量和平均运动矢量信息,发展了一种精确的头信息估计模型;联合头信息与残差信息模型,并结合残差失真模型,提出新的联合头信息与残差率失真模型;利用精确的估计方法进一步提高率失真性能。相对于最新的JVT H.264/AVC参考软件JM10.2中采用的JVT-G012方法,该方法提高了实际码率与目标码率之间的匹配率达到了98.06%,重构视频的平均亮度PSNR值增加了0.27 dB。This paper presents a new frame-level rate control scheme for H.264/AVC with joint rate and distortion models. A rate model for the header information is developed to estimate header bits more accurately. The number of header bits is mod- eled as a function of the number of blocks and the average bits of MVs. A new Rate-Distortion （R-D） model for header informa- tion and texture is proposed to joint header-texture rate model and a distortion model. An accurate estimation method is pro- posed to improve rate distortion performance. When compared with the rate control scheme G012 which is adopted by the latest JVT H.264/AVC reference model JM10.2, the proposed rate control algorithm could improve the matching rate between actual bits and targets ones by up to 98.06%. The average luminance PSNR of decoded video is increased by up to 0.27 dB.

关 键 词：率失真 码率控制 头信息 目标码率 

分 类 号：TN919.8[电子电信—通信与信息系统]