H.264/AVC可伸缩性扩展视频编码系统(JSVM)提供了一种可伸缩视频编码(SVC)的解决方案,然而它本身并没有提供一种有效的码率控制算法.文中基于JSVM的分层预测结构,提出了一种全新的码率控制算法.首先在码率分配方面,考虑到分层B帧预测(或运动补偿时间域滤波(MCTF))结构,给出一种分层的码率分配方案;然后,针对不同类型和不同时间分解层各自的统计特性,分别为它们设计了不同的率失真(R-D)模型.实验结果表明,文中算法能够有效地控制码率,使得目标码率跟实际产生码率之间的偏差最大不超过2%;同时文中算法较大地提高了解码图像的质量,使得峰值信噪比(PSNR)在低码率端可提高1dB;另外,JSVM是通过不断调整量化参数(QP)使得实际产生的码率逐次逼近目标码率,较之这种尝试型的码率控制算法,文中基于模型的码率控制是在一次编码中产生最终的目标码率,从而大大降低了计算复杂度.
The joint scalable video model (JSVM) as a scalable extension of H. 264/AVC prorides a high efficient solution for scalable video coding (SVC). However, there is not efficient rate control scheme is specified. This paper presents a new rate control scheme for the JSVM based on the special structure of hierarchical B decomposition (or motion compensation temporal filtering (MCTF)). In a sense of bit allocation, a novel hierarchical bit allocation strategy is proposed. Meanwhile, the different rate-distortion (R-D) model is provided for each kind picture with different picture type or temporal decomposition level. Experiment results show that the mismatch of target bit rate and real bit rate generated for coding doesn't exceed 2% ; meanwhile, the PSNR improvement can be up to ldB in low bit rate; furthermore, comparing to the original rate control method which achieves the target bit rate through regulating quantization parameter (QP) step by step, the computation complexity of the proposed rate control algorithm base on experimental model is reduced drastically.