计算机工程与应用 ›› 2009, Vol. 45 ›› Issue (15): 196-199.DOI: 10.3778/j.issn.1002-8331.2009.15.057

• 图形、图像、模式识别 • 上一篇    下一篇

对视觉敏感区域进行加权的视频质量评价方法

卢国庆1,李均利1,陈 刚1,章 颖1,满家巨2   

  1. 1.宁波大学 信息学院,浙江 宁波 315211
    2.湖南师范大学,长沙 410083
  • 收稿日期:2008-03-25 修回日期:2008-06-11 出版日期:2009-05-21 发布日期:2009-05-21
  • 通讯作者: 卢国庆

Criterion of video quality assessment based on property of HVS

LU Guo-qing1,LI Jun-li1,CHEN Gang1,ZHANG Ying1,MAN Jia-ju2   

  1. 1.College of Information Science and Engineering,Ningbo University,Ningbo,Zhejiang 315211,China
    2.Hunan Normal University,Changsha 410083,China
  • Received:2008-03-25 Revised:2008-06-11 Online:2009-05-21 Published:2009-05-21
  • Contact: LU Guo-qing

摘要: 最近Guan-Hao Chen在Zhou Wang提出的MSSIM方法基础上,提出了一种基于边缘结构失真的测量方法(ESSIM),但是这种方法不能够很好的直接应用到视频领域。就人类视觉系统(HVS)做了一定的研究,提出了一种考虑人眼视觉特性,考虑结构信息和对运动场景分类的视频质量评价方法(HCSSIM)。实验结果表明,在压缩损失、模糊和部分椒盐噪声损失上,其评价结果要好于现有的几类通用的评价方法。

Abstract: Zhou Wang proposed the Structural Similarity theory(MSSIM).Based on it,Guan-Hao Chen proposed a marginal texture distorts theory.But the theory isn’t applied directly on the video field.In this dissertation,based on the HVS,the structural distortion measurement and motion information,this paper proposes a theory-HCSSIM.From lots of simulation results,HCSSIM outperforms MSSIM and the weighting SSIM based on variance proposed by Wang Zhou et al,in the compress distortion,smooth distortion and partial salt-and-pepper noise.