电信科学 ›› 2017, Vol. 33 ›› Issue (10): 124-133.doi: 10.11959/j.issn.1000-0801.2017250

• 研究与开发 • 上一篇    下一篇

基于显著特征的HEVC帧内率失真优化

李金龙,陈芬,彭宗举,蒋刚毅,郁梅   

  1. 宁波大学信息科学与工程学院,浙江 宁波 315211
  • 修回日期:2017-08-18 出版日期:2017-10-01 发布日期:2017-11-13
  • 作者简介:李金龙(1993-),男,宁波大学信息科学与工程学院硕士生,主要研究方向为视频编码。|陈芬(1973-),女,博士,宁波大学信息科学与工程学院副教授,主要研究方向为三维视频信号处理与编码。|彭宗举(1973-),男,博士,宁波大学信息科学与工程学院教授,主要研究方向为三维视频信号处理与编码。|蒋刚毅(1964-),男,博士,宁波大学信息科学与工程学院教授,主要研究方向为三维视频信号处理与编码。|郁梅(1968-),女,博士,宁波大学信息科学与工程学院教授,主要研究方向为三维视频信号处理与编码。
  • 基金资助:
    国家自然科学基金重点国际(地区)合作研究项目(61620106012);国家自然科学基金资助项目(61771269);国家自然科学基金联合基金(NSFC-广东联合基金)资助项目(U1301257);浙江省自然科学基金资助项目(LY16F010002);浙江省自然科学基金资助项目(LY15F010005);浙江省自然科学基金资助项目(LY17F010005)

Rate distortion optimization based on saliency characteristic for HEVC intra coding

Jinlong LI,Fen CHEN,Zongju PENG,Gangyi JIANG,Mei YU   

  1. College of Information Science and Engineering,Ningbo University,Ningbo 315211,China
  • Revised:2017-08-18 Online:2017-10-01 Published:2017-11-13
  • Supported by:
    The National Natural Science Foundation of Key International(Regional) Cooperative Research Project of China(61620106012);The National Natural Science Foundation of China(61771269);The National Natural Science Foundation of China (NSFC-Guangdong Joint Foundation)(U1301257);Zhejiang Provincial Natural Science Foundation of China(LY16F010002);Zhejiang Provincial Natural Science Foundation of China(LY15F010005);Zhejiang Provincial Natural Science Foundation of China(LY17F010005)

摘要:

引入感知因素的RDO算法和感知编码可以提高编码视频主观质量,但也产生了明显的率失真性能损失。针对上述问题,提出了一种基于显著特征的 RDO 方法。首先探索了 HEVC 编码树单元的码率分布和显著特征之间的相关性,提出显著码率先验模型并确定了模型的初始参数;然后建立了码率差异最小化的率失真代价模型,并根据该模型推导得到自适应拉格朗日乘子计算方法;最后设计了模型参数更新规则。实验结果表明,算法在率失真性能基本不变的情况下,提高了视频的主观质量。

关键词: 高性能视频编码, 率失真优化, 拉格朗日乘子, 显著特性

Abstract:

Although the perception based rate distortion optimization (RDO) and coding algorithms can improve subjective quality,they significantly decrease rate distortion performance.For the above questions,a saliency-based RDO method was proposed.Firstly,according to the relation between rate distribution and saliency features of coding tree unit in HEVC,a priori saliency-rate model was designed.Furthermore,the initial parameters of the model were determined.Secondly,the rate difference minimization rate distortion model was proposed,and then a calculation method of the adaptive Lagrangian multiplier was derived from the model.Finally,update mechanism of model parameters was designed.Experimental results show that the proposed algorithm can improve the subjective quality while it maintains rate distortion performance.

Key words: high efficiency video coding, rate distortion optimization, Lagrangian multiplier, saliency characteristic

中图分类号: 

No Suggested Reading articles found!