视频质量平滑中恒定SNR比特分配

Constant-SNR Bit Allocation in Video Quality Control

该文提出了一种基于恒定SNR的单次扫描帧级比特率分配的视频质量平滑方案(CSNRBA)。采用之前所有编码帧的SNR的递减几何级数加权平均值作为当前帧的目标SNR,再根据国际上最近提出的闭合形式D-Q分析模型和线性码率模型,准确计算当前帧的比特预算。实验结果表明,该文提出的CSNRBA算法提供了比保持PSNR恒定算法和目前广泛采用的恒定比特率分配算法都要平滑得多的视频质量,并在物体清晰度、轮廓和自然色彩方面明显改善了传输视频的主观质量。

A single-pass frame-level Constant-SNR Bit Allocation （CSNRBA） scheme is developed for video quality smoothing throughout the video sequence in this paper. A power series weighted actual SNR average of previous coded frames is adopted as the target SNR for the current frame. According to the newly developed analytic close-form D-Q model and the linear rate control algorithm, the bit budget can be estimated. Experimental results show that the proposed CSNRBA scheme provides much smoother video quality and achieves much better subjective video quality in terms of natural color, sharp objects and silhouette significantly on all testing video sequences than both constant bit allocation and constant PSNR bit allocation schemes.