摘要
本文首先介绍了新一代视频编码VVC、AV1和AVS3的特点以及常用的视频质量主客观评价方法,然后使用PSNR、SSIM和基于机器学习的VMAF指标,对新一代视频编码进行客观指标评价,最后进行主观评价实验得出主观质量分,分析客观指标评价结果和主观评价结果的一致性。
This paper firstly introduces the characteristics of the new generation video coding VVC,AV1 and AVS3 and the common subjective and objective video quality evaluation methods,then uses PSNR,SSIM and VMAF parameters based on machine learning to evaluate the objective parameters of the new generation video coding,and finally conducts subjective evaluation experiments to obtain the subjective quality score,and analyses the consistency between the objective parameters evaluation results and the subjective evaluation results.
作者
孟昭旭
朱宏宣
Meng Zhaoxu;Zhu Hongxuan(China Film High-Tech Research Institute,Beijing Film Academy;Department of Film and Television Technology,Beijing Film Academy)
出处
《现代电影技术》
2021年第6期12-17,共6页
Advanced Motion Picture Technology
