论文部分内容阅读
帧内更新是在视频编码中常采用的容错手段,它能够有效的阻止差错传播。帧内更新率高,能够有效的抑制差错扩散,但是会降低编码效率,在码率一定时,会增大信源失真。只有选择合适的更新率,才能更好的达到抗差错的目的。本文提出了一种基于端到端模型,对自适应更新和码率控制进行联合优化的算法,在满足码率的要求下,自适应的调整帧内更新位置和更新数量,实现端到端失真最小。实验结果表明,本文提出的算法与随机更新算法相比,PSNR提高了0.5～1.5dB。
Intra-frame update is often used in video coding fault-tolerant means, it can effectively prevent the error propagation. Intra-frame update rate is high, which can effectively suppress the error diffusion, but will reduce the coding efficiency, at a fixed code rate, will increase the source distortion. Only by choosing the appropriate update rate, in order to better achieve the purpose of anti-error. This paper presents a joint end-to-end model based on adaptive updating and rate control joint optimization algorithm to meet the bit rate requirements, adaptive adjustment of the update location and update the number of frames to achieve end-to-end distortion Min. The experimental results show that compared with the random update algorithm, the proposed algorithm improves the PSNR by 0.5 ~ 1.5dB.