论文部分内容阅读
率失真优化(RDO)是视频编码压缩中的关键技术之一。传统的率失真优化技术使用误差平方和(SSE)或绝对差和(SAD)度量失真,不能获得较好的视觉感知质量。本文针对新一代视频编码标准高效视频编码(HEVC)帧内编码RDO,提出了一种基于结构相似度(SSIM)度量失真的RDO方法。首先,根据HEVC编码树形单元(CTU)结构设计了基于SSIM的失真计算方法;然后,提出了一种基于量化器推导的码率-量化步长(R-Δ)模型和一种基于统计分析的失真-量化参数(DSSIM-QP)模型,用于求取拉格朗日乘数;最终,使用多QP优化方法求取模型参数。实验表明,相对于HEVC传统的RDO,针对全I帧(即所有帧都为帧内编码)固定QP编码和多QP优化编码,在相同的SSIM条件下,码率分别平均降低8.4%和13.9%左右,同时编码复杂度分别平均增加约3%和2%。
Rate-distortion optimization (RDO) is one of the key techniques in video coding compression. Traditional rate-distortion optimization techniques use error square sum (SSE) or absolute difference (SAD) metric distortion to achieve better visual perception quality. In this paper, an RDO method based on Structural Similarity (SSIM) measure distortion is proposed for the intra-coding RDO of next generation video coding standard high efficiency video coding (HEVC). First, a method of calculating the distortion based on SSIM is designed according to the HEVC coding tree unit (CTU) structure. Then, a rate-quantization step (R-Δ) model based on the quantizer and a statistic analysis The distortion-quantization parameter (DSSIM-QP) model is used to determine the Lagrange multiplier. Finally, the model parameters are obtained using the multi-QP optimization method. Experiments show that compared with HEVC traditional RDO, the code rate is reduced by 8.4% and 13.9% respectively on the same SSIM condition for both fixed QP coding and multi-QP coding for all I-frames (ie all frames are intra-coded) While the coding complexity increases about 3% and 2% respectively on average.