摘要
多功能视频编解码方式(H.266/VVC)在每一个编码环节上都采取了多项改进措施,诸多措施的总体效果使其编码效率比高效视频编码方式(HEVC)提高了一倍。该文对几种动态情况各异的视频序列(*.yuv)采用H.266编解码器进行压缩和解压,通过改变量化参数(QP)取值,研究在不同比特率条件下,视频压缩后编码效率和时间损耗的差异。实验表明,对于较低的比特率和4K视频常用的比特率(20 Mbps)附近,在编码效率和峰值信噪比上都有着很大的优势,但在超大比特率的情况下,两者表现近乎相同。对时间倍数和码率节省进行计算后发现,每个视频序列的结果差距较大,说明编码复杂度和编码性能与选取的视频序列本身有很大关系。H.266对复杂度较高的视频序列提升明显,编码复杂度有了显著提升,在某些比特率的情况下甚至可达到100倍之多,H.266同时对硬件的算力带来了很大挑战,需要在之后进行进一步优化。
The versatile video coding(H.266/VVC)takes several improvements in each coding step,and the overall effect of many measures does result in a doubling of the coding efficiency compared to high efficiency video coding(HEVC).In this paper,several video sequences with different dynamic conditions(*.yuv)are compressed and decompressed using the H.266 codec,and the differences in coding efficiency and time loss after video compression are investigated by varying the quantization parameters(QP)values at different bit rates.Experiments show that for lower bit rates and near the bit rate commonly used for 4K video(20 Mbps),there is a significant advantage in both bit coding rate and peak signal to noise radio(PSNR),but at very large bit rates,both perform nearly identically.Calculations of the time multiplier and bitrate savings reveal a wide range of results for each video sequence,suggesting that coding complexity and coding performance are highly dependent on the selected video sequence itself.H.266 improves significantly for more complex video sequences,with a significant increase in coding complexity,up to 100x at some bitrates.H.266 also poses a significant challenge to hardware arithmetic,which will need to be further optimized in the future.
作者
金伟正
宁明慧
王欣然
梁芳
JIN Weizheng;NING Minghui;WANG Xinran;LIANG Fang(School of Electronic Information,Wuhan University,Wuhan 430072,China)
出处
《实验科学与技术》
2024年第3期9-14,共6页
Experiment Science and Technology
基金
国家重点研发计划(2018YFB1201602-05)。
作者简介
金伟正(1966-),男,博士,副教授,主要从事射频系统、图像处理、环境感知等科研及教学工作。E-mail:jwz@whu.edu.cn。
