Rate control plays an important role in video coding, which can minimize the distortion of compressed video with bit rate constraints. Rate control algorithm based on $R-\lambda$ model is capable of controlling the bit rate precisely for high efficiency video coding (HEVC). However, quality dependency is not considered very well in $\lambda$-domain rate control algorithm when allocating bits and updating model parameters. This paper proposes a rate control algorithm based on group of pictures (GOP) level quality dependency for low-delay hierarchical coding structure. Firstly, the quality dependency between the current GOP and encoded GOPs mainly comes from the crucial frame (C-Frame). C-Frame guided GOP level quality dependency (CF-GOP-QD) model can be expressed by a CF-GOP-QD factor. Secondly, a CF-GOP-QD model based GOP level rate–distortion optimization (RDO) under the total bit constraints is proposed to guide the bit allocation processing, which is better than traditional bit allocation method. Experimental results show that the proposed rate control algorithm can achieve better rate–distortion performance, better visual quality and better rate control accuracy. Specifically, the proposed method shows 5.801% and 5.068% BD-rate saving on average compared against HM16.7 frame level rate control method under low-delay B and low-delay P configurations.

速率控制在视频编码中起着重要作用，它可以将具有比特率约束的压缩视频的失真降至最低。基于速率控制算法$\mathrm{[R}-\lambda$该模型能够针对高效视频编码（HEVC）精确控制比特率。但是，质量依赖性在$\lambda$分配比特和更新模型参数时的域速率控制算法。针对低延迟分层编码结构，提出了一种基于图像组（GOP）级质量依赖性的码率控制算法。首先，当前GOP与编码GOP之间的质量相关性主要来自关键帧（C-Frame）。C帧引导的GOP级别质量依存关系（CF-GOP-QD）模型可以由CF-GOP-QD因子表示。其次，提出了一种基于CF-GOP-QD模型的总比特约束下的GOP电平速率-失真优化（RDO），以指导比特分配处理，优于传统的比特分配方法。实验结果表明，提出的码率控制算法可以实现更好的码率失真性能，更好的视觉质量和更好的码率控制精度。特别，