摘要
通过介绍几种持续出席模式的视频合成方法,以及分析H.264不能采用快速编码域合成方法的原因,深入研究了一种像素域和编码域相结合的混合合成方法,利用H.264标准的新特性,对解码图像帧的选择、合成图像帧的条带组分割、宏块编码参数的设置以及I集合的构造等问题进行了分析和改进,实验结果表明改进方法在降低计算复杂度和码率的同时视频质量约提高0.4-0.7dB.
Several traditional approaches have been introduced to perform video combining.Through analysis of problems with coded-domain approach in H.264,this paper investigates an efficient hybrid approach that blends pixel-domain and coded-domain tools.Taking advantage of the significant coding new features of the H.264,several issues including the selection of the decoded pictures in combining,slice group partitioning of the combined picture,setting of macroblock coding parameters,and constructing of I-set are analyzed and improved.Simulation results demonstrate that the performance of the improved approach is about 0.4 to 0.7 dB better than the original approach as well as low computational complexity and low bit-rate.
出处
《小型微型计算机系统》
CSCD
北大核心
2012年第5期1062-1067,共6页
Journal of Chinese Computer Systems
关键词
视频会议
视频合成
持续出席
H.264
像素域
编码域
video conference
video combing
continuous presence
H.264
pixel-domain
coded-domain
