视频编码失真估计综述
A Review of Distortion Estimation for Video Coding
陈志同 1王向文1
作者信息
- 1. 上海电力大学电子与信息工程学院,上海 201306
- 折叠
摘要
编码失真是评价视频算法性能的主要因素之一,如率失真优化、码率控制和最佳量化等算法设计中,在不进行复杂预编码的情况下精确估计编码失真,是达到良好率失真性能的必要途径.虽然部分编码失真估计方法已经被采用并作为编码算法的基础,然而,与对码率估计的关注度不同,失真估计并未引起人们的足够关注.系统地回顾总结了视频编码失真估计方法,实验并给出了一些比较结果.将失真估计方法分为四类:1)基于DCT系数分布的建模;2)ρ域;3)λ域和 4)基于失真传播的建模.概述了这些方法的基本原理,并且根据实验结果对模型的精度进行了简要对比分析.
Abstract
Coding distortion is one of the primary factors for evaluating the performance of video algorithms.In the design of algorithms such as rate-distortion optimization,rate control and optimal quantization.Precisely estimating the coding distortion without complex pre-coding is a necessary way to achieve good rate-distortion performance.Al-though some coding distortion estimation methods have been adopted and used as the basis of coding algorithms,how-ever,unlike the attention paid to rate estimation,distortion estimation has not received enough attention.This paper systematically reviews and summarizes video coding distortion estimation methods,and experiments and gives some comparative results.Specifically,this paper divides the distortion estimation methods into four categories:1)DCT co-efficient distribution based;2)the ρ-domain;3)the λ-domain and 4)Distortion propagation based modeling.This paper outlines the basic principles of these methods,and briefly compares the accuracy of the models based on the ex-perimental results.
关键词
视频编码/编码失真/失真传播/率失真优化Key words
Video coding/Coding distortion/Distortion propagation/Rate-distortion optimization引用本文复制引用
出版年
2024
NETL
NSTL
万方数据