搜索

x

留言板

尊敬的读者、作者、审稿人, 关于本刊的投稿、审稿、编辑和出版的任何问题, 您可以本页添加留言。我们将尽快给您答复。谢谢您的支持!

姓名
邮箱
手机号码
标题
留言内容
验证码

基于图像内容视觉感知的图像质量客观评价方法

姚军财 刘贵忠

引用本文:
Citation:

基于图像内容视觉感知的图像质量客观评价方法

姚军财, 刘贵忠

Objective assessment method of image quality based on visual perception of image content

Yao Jun-Cai, Liu Gui-Zhong
PDF
导出引用
  • 图像质量客观评价在图像和视频传输、编解码以及服务质量中起着非常重要的作用.然而现有的方法往往没有考虑图像内容特征及其视觉感知,使得其质量客观评价与主观感知结果存在一定的差距.基于此,本文结合图像内容的复杂性特征和人眼的掩蔽特性、对比敏感度特性以及亮度感知的非线性特性,提出了一种基于人眼对图像内容感知的图像质量客观评价方法.该方法首先结合亮度感知的非线性模型将图像进行转换,得到人眼感知强度图;再分别以人眼对比敏感度值和图像局部平均对比度值作为权重因子对强度求和,以求和的数据信息作为人眼感知图像的内容,并构建图像感知模型;最后以此模型分别模拟人眼感知参考图像和失真图像,并计算二者的强度差,以强度差为评价分数的基础构建图像质量客观评价模型.采用LIVE,TID2008和CSIQ三个数据库中的共47幅参考图像和1549幅测试图像进行仿真实验,且与SSIM,VSNR,FSIM和PSNRHVS等典型的图像质量客观评价模型进行对比分析,同时探讨影响图像质量评价的因素.结果表明:所提方法的评价分数与主观评价分数的Pearson线性相关性系数和Spearman秩相关系数值比SSIM的评价结果均有一定程度的提高,提高幅度分别平均为9.5402%和3.2852%,比PSNRHVS和VSNR提高幅度更大.综合以上表明:所提方法是一种有效可行的图像质量客观评价方法;同时,在图像质量客观评价中,考虑人眼对图像内容的感知和复杂度的分析有助于提高图像质量主客观评价的一致性,评价精度可得到进一步的提高.
    Objective image quality assessment (IQA) plays a very important role in transmission, encoding, and quality of service (QoS) of the image and video data. However, the existing IQA methods often do not consider image content features and their visual perception, so there is a certain gap between the objective IQA sores and the subjective perception. To solve this problem, in the study, we propose an objective IQA method based on the visual perception of image content, which combines the complexity characteristics of image content, and the properties of masking, contrast sensitivity and luminance perception nonlinearity of human visual system (HVS). In the proposed method, the image is first transformed using a nonlinear model of luminance perception to obtain the intensity perception image. Then, the intensity information is summed using the contrast sensitivity values of HVS and the average contrast values of the local image as a weighting factor of the intensity. The summed data information is taken as the content of human perceiving image, and an image perception model is constructed. Finally, the reference images and distorted images are perceived by simulating the HVS with this model. Moreover, the difference in intensity between two perceived images is calculated. Based on the intensity difference and peak signal-to -noise ratio model, an objective IQA model is constructed. Further, the simulation with 47 reference images and 1549 test images in the LIVE, TID2008, and CSIQ databases is conducted. Moreover, the experimental results are compared with those of four typical objective IQA models, namely SSIM, VSNR, FSIM, and PSNRHVS. In addition, we explore the factors that affect the IQA accuracy and a way to improve assessment accuracy by combining HVS characteristics, through analyzing the correlation between IQA results of the proposed model and the subjective mean opinion scores (MOSs) provided in the three image databases from the following two aspects. Namely, (1) all reference images in three image databases are distorted by multiple types, and the distorted images of each reference image are taken as a test sequence. Then, the proposed model is used to evaluate each test sequence to obtain the IQA scores. By analyzing the correlation between the IQA scores of each test sequence and the subjective MOSs and comparing them with the assessment results of SSIM, we explore the influence of the image content complexity on the objective IQA accuracy. (2) The test images which are distorted by each type and many distortion degrees are used as another sequence, and they are evaluated by the proposed IQA model. By analyzing the correlation between the subjective MOSs and the IQA results of each test sequence, and comparing them with assessment results of SSIM, we discuss the influence of image distortion mode on the IQA accuracy. The experimental results show that the coefficient values of Pearson linear correlation and Spearman rank order correlation between the objective IQA scores obtained by the proposed method and the subjective MOSs have been averagely improved by 9.5402% and 3.2852%, respectively, in comparison with IQA results from the SSIM method. Also, they are enhanced more significantly than those fom the PSNRHVS and VSNR methods. In summary, it is shown that the proposed IQA method is an effective and feasible method of objectively assessing the image quality; moreover, it is shown that in the objective assessment of image quality it is very helpful to improve the consistency of subjective and objective assessment of image quality by considering the content perception and complexity analysis of the images.
      通信作者: 刘贵忠, liugz@mail.xjtu.edu.cn
    • 基金项目: 国家自然科学基金(批准号:61301237)和陕西省科技新星计划(批准号:2015KJXX-42)资助的课题.
      Corresponding author: Liu Gui-Zhong, liugz@mail.xjtu.edu.cn
    • Funds: Project supported by the National Natural Science Foundation of China (Grant No. 61301237) and the Natural Science Foundation of Shaanxi Province, China (Grant No. 2015KJXX-42).
    [1]

    Wang Y Q 2014 J. Nanjing Univ. (Nat. Sci. Ed.) 50 361 (in Chinese)[王元庆 2014 南京大学学报(自然科学版) 50 361]

    [2]

    Zhuang J Y, Chen Q, He W J, Mao T Y 2016 Acta Phys. Sin. 65 040501 (in Chinese)[庄佳衍, 陈钱, 何伟基, 冒添逸 2016 物理学报 65 040501]

    [3]

    Wang Z, Bovik A C, Sheikh H R, Simoncelli E P 2004 IEEE Trans. Image Process. 13 600

    [4]

    Chandler D M, Hemami S S 2007 IEEE Trans. Image Process. 16 2284

    [5]

    Zhang L, Zhang L, Mou X, Zhang D 2011 IEEE Trans. Image Process. 20 2378

    [6]

    Xue W, Zhang L, Mou X, Bovik A C 2014 IEEE Trans. Image Process. 23 684

    [7]

    Zhang L, Shen Y, Li H 2014 IEEE Trans. Image Process. 23 4270

    [8]

    Paudyal P, Battisti F, Sjöström M, Olsson R, Carli M 2017 IEEE Trans. Broadcast. 63 507

    [9]

    Bae S H, Kim M 2016 IEEE Trans. Image Process. 25 2392.

    [10]

    Gu K, Wang S, Zhai G, Ma S, Yang X 2016 Signal Image Video Process. 10 803

    [11]

    Wang X L, Wu D W, Li X, Zhu H N, Chen K, Fang G 2017 Acta Phys. Sin. 66 230302 (in Chinese)[王湘林, 吴德伟, 李响, 朱浩男, 陈坤, 方冠 2017 物理学报 66 230302]

    [12]

    Akamine W Y L, Farias M C Q 2014 J. Electron. Imaging 23 061107

    [13]

    Li C F, Bovik A C 2010 J. Electron. Imaging 19 143

    [14]

    Guo J, Hu G, Xu W, Huang L 2017 J. Vis. Commun. Image Represent. 43 50

    [15]

    Hou W, Mei F H, Chen G J, Deng X W 2015 Acta Phys. Sin. 64 024202 (in Chinese)[侯旺, 梅风华, 陈国军, 邓喜文 2015 物理学报 64 024202]

    [16]

    Stephen W, Huw O, Vien C, Iain P S 2006 Color Res. Appl. 31 315

    [17]

    Nadenau M 2000 Ph. D. Dissertation (Lausanne:École Polytechnique Fédérale de Lausanne)

    [18]

    Yao J C, Shen J, Wang J H 2008 Acta Phys. Sin. 57 4034 (in Chinese)[姚军财, 申静, 王剑华 2008 物理学报 57 4034]

    [19]

    Sheikh H R, Sabir M F, Bovik A C 2006 IEEE Trans. Image Process. 15 3440

    [20]

    Nikolay P, Vladimir L, Alexander Z, Karen E, Jaakko A, Marco C, Federica B 2009 Adv. Modern Radioelectron. 10 30

    [21]

    Larson E C, Chandler D M 2010 J. Electron. Imaging 19 011006

    [22]

    Zhang F, Bull D R 2013 Proceedings of the 20th IEEE Interatinonal Conference on Image Processing (ICIP) Melbourne, Australia,September 15-18, 2013 p39

    [23]

    Gu K, Wang S, Zhai G, Lin W, Yang X, Zhang W 2016 IEEE Trans. Image Process. 62 446

  • [1]

    Wang Y Q 2014 J. Nanjing Univ. (Nat. Sci. Ed.) 50 361 (in Chinese)[王元庆 2014 南京大学学报(自然科学版) 50 361]

    [2]

    Zhuang J Y, Chen Q, He W J, Mao T Y 2016 Acta Phys. Sin. 65 040501 (in Chinese)[庄佳衍, 陈钱, 何伟基, 冒添逸 2016 物理学报 65 040501]

    [3]

    Wang Z, Bovik A C, Sheikh H R, Simoncelli E P 2004 IEEE Trans. Image Process. 13 600

    [4]

    Chandler D M, Hemami S S 2007 IEEE Trans. Image Process. 16 2284

    [5]

    Zhang L, Zhang L, Mou X, Zhang D 2011 IEEE Trans. Image Process. 20 2378

    [6]

    Xue W, Zhang L, Mou X, Bovik A C 2014 IEEE Trans. Image Process. 23 684

    [7]

    Zhang L, Shen Y, Li H 2014 IEEE Trans. Image Process. 23 4270

    [8]

    Paudyal P, Battisti F, Sjöström M, Olsson R, Carli M 2017 IEEE Trans. Broadcast. 63 507

    [9]

    Bae S H, Kim M 2016 IEEE Trans. Image Process. 25 2392.

    [10]

    Gu K, Wang S, Zhai G, Ma S, Yang X 2016 Signal Image Video Process. 10 803

    [11]

    Wang X L, Wu D W, Li X, Zhu H N, Chen K, Fang G 2017 Acta Phys. Sin. 66 230302 (in Chinese)[王湘林, 吴德伟, 李响, 朱浩男, 陈坤, 方冠 2017 物理学报 66 230302]

    [12]

    Akamine W Y L, Farias M C Q 2014 J. Electron. Imaging 23 061107

    [13]

    Li C F, Bovik A C 2010 J. Electron. Imaging 19 143

    [14]

    Guo J, Hu G, Xu W, Huang L 2017 J. Vis. Commun. Image Represent. 43 50

    [15]

    Hou W, Mei F H, Chen G J, Deng X W 2015 Acta Phys. Sin. 64 024202 (in Chinese)[侯旺, 梅风华, 陈国军, 邓喜文 2015 物理学报 64 024202]

    [16]

    Stephen W, Huw O, Vien C, Iain P S 2006 Color Res. Appl. 31 315

    [17]

    Nadenau M 2000 Ph. D. Dissertation (Lausanne:École Polytechnique Fédérale de Lausanne)

    [18]

    Yao J C, Shen J, Wang J H 2008 Acta Phys. Sin. 57 4034 (in Chinese)[姚军财, 申静, 王剑华 2008 物理学报 57 4034]

    [19]

    Sheikh H R, Sabir M F, Bovik A C 2006 IEEE Trans. Image Process. 15 3440

    [20]

    Nikolay P, Vladimir L, Alexander Z, Karen E, Jaakko A, Marco C, Federica B 2009 Adv. Modern Radioelectron. 10 30

    [21]

    Larson E C, Chandler D M 2010 J. Electron. Imaging 19 011006

    [22]

    Zhang F, Bull D R 2013 Proceedings of the 20th IEEE Interatinonal Conference on Image Processing (ICIP) Melbourne, Australia,September 15-18, 2013 p39

    [23]

    Gu K, Wang S, Zhai G, Lin W, Yang X, Zhang W 2016 IEEE Trans. Image Process. 62 446

  • [1] 贺芷椰, 张彦东, 唐春华, 李军利, 李四维, 于斌. 中继透镜分辨率在像素编码曝光成像中对图像重构质量的影响分析. 物理学报, 2023, 0(0): 0-0. doi: 10.7498/aps.72.20221588
    [2] 张海鹏, 赵昌哲, 鞠晓璐, 汤杰, 肖体乔. 基于迭代重构算法改进晶体衍射分光X射线鬼成像的图像质量研究. 物理学报, 2022, 71(7): 074201. doi: 10.7498/aps.71.20211978
    [3] 霍勇刚, 严江余, 张全虎. 缪子多模态成像图像质量分析. 物理学报, 2022, 71(2): 021401. doi: 10.7498/aps.71.20211083
    [4] 霍勇刚, 严江余, 张全虎. 缪子多模态成像图像质量分析. 物理学报, 2021, (): . doi: 10.7498/aps.70.20211083
    [5] 郎利影, 陆佳磊, 于娜娜, 席思星, 王雪光, 张雷, 焦小雪. 基于深度学习的联合变换相关器光学图像加密系统去噪方法. 物理学报, 2020, 69(24): 244204. doi: 10.7498/aps.69.20200805
    [6] 史晨阳, 林燕丹. 基于图像色貌和梯度特征的图像质量客观评价. 物理学报, 2020, 69(22): 228701. doi: 10.7498/aps.69.20200753
    [7] 姚军财, 申静. 基于图像内容对比感知的图像质量客观评价. 物理学报, 2020, 69(14): 148702. doi: 10.7498/aps.69.20200335
    [8] 田恒, 朱京平, 张云尧, 管今哥, 侯洵. 浑浊介质中图像对比度与成像方式的关系. 物理学报, 2016, 65(8): 084201. doi: 10.7498/aps.65.084201
    [9] 侯旺, 梅风华, 陈国军, 邓喜文. 基于背景最佳滤波尺度的红外图像复杂度评价准则. 物理学报, 2015, 64(23): 234202. doi: 10.7498/aps.64.234202
    [10] 王晓娟, 乔少博, 沈柏竹, 封国林. 东亚北部地区气温的冬季-冬季再现特征研究. 物理学报, 2014, 63(23): 239202. doi: 10.7498/aps.63.239202
    [11] 范洪义. 相干态在参数量子相空间的两维正态分布. 物理学报, 2014, 63(2): 020302. doi: 10.7498/aps.63.020302
    [12] 张太宁, 孟春宁, 刘润蓓, 常胜江. 基于暗瞳图像的人眼视线估计. 物理学报, 2013, 62(13): 134204. doi: 10.7498/aps.62.134204
    [13] 海凛, 张业荣, 潘灿林. 混合分集多天线系统基于相关性的分析信道建模. 物理学报, 2013, 62(23): 238402. doi: 10.7498/aps.62.238402
    [14] 王芳, 赵星, 杨勇, 方志良, 袁小聪. 基于人眼视觉的集成成像三维显示分辨率的比较. 物理学报, 2012, 61(8): 084212. doi: 10.7498/aps.61.084212
    [15] 章法强, 杨建伦, 李正宏, 叶凡, 徐荣昆. 厚闪烁体内次级中子对快中子图像质量的影响研究. 物理学报, 2009, 58(2): 1316-1320. doi: 10.7498/aps.58.1316
    [16] 杨朝羽, 唐国宁. 基于蜂拥控制算法思想的时空混沌耦合反馈控制. 物理学报, 2009, 58(1): 143-149. doi: 10.7498/aps.58.143
    [17] 姚军财, 申 静, 王剑华. 阴极射线管显示器亮度范围内对人眼视觉特性的实验研究. 物理学报, 2008, 57(7): 4034-4041. doi: 10.7498/aps.57.4034
    [18] 张 闯, 柏连发, 张 毅. 基于灰度空间相关性的双谱微光图像融合方法. 物理学报, 2007, 56(6): 3227-3233. doi: 10.7498/aps.56.3227
    [19] 张佃中. 非线性时间序列互信息与Lempel-Ziv复杂度的相关性研究. 物理学报, 2007, 56(6): 3152-3157. doi: 10.7498/aps.56.3152
    [20] 汪 敏, 胡小方, 伍小平. 物体内部三维位移场分析的数字图像相关方法. 物理学报, 2006, 55(10): 5135-5139. doi: 10.7498/aps.55.5135
计量
  • 文章访问数:  3781
  • PDF下载量:  205
  • 被引次数: 0
出版历程
  • 收稿日期:  2018-01-23
  • 修回日期:  2018-03-07
  • 刊出日期:  2019-05-20

基于图像内容视觉感知的图像质量客观评价方法

  • 1. 西安交通大学电子与信息工程学院, 西安 710049;
  • 2. 陕西理工大学物理与电信工程学院, 汉中 723000
  • 通信作者: 刘贵忠, liugz@mail.xjtu.edu.cn
    基金项目: 国家自然科学基金(批准号:61301237)和陕西省科技新星计划(批准号:2015KJXX-42)资助的课题.

摘要: 图像质量客观评价在图像和视频传输、编解码以及服务质量中起着非常重要的作用.然而现有的方法往往没有考虑图像内容特征及其视觉感知,使得其质量客观评价与主观感知结果存在一定的差距.基于此,本文结合图像内容的复杂性特征和人眼的掩蔽特性、对比敏感度特性以及亮度感知的非线性特性,提出了一种基于人眼对图像内容感知的图像质量客观评价方法.该方法首先结合亮度感知的非线性模型将图像进行转换,得到人眼感知强度图;再分别以人眼对比敏感度值和图像局部平均对比度值作为权重因子对强度求和,以求和的数据信息作为人眼感知图像的内容,并构建图像感知模型;最后以此模型分别模拟人眼感知参考图像和失真图像,并计算二者的强度差,以强度差为评价分数的基础构建图像质量客观评价模型.采用LIVE,TID2008和CSIQ三个数据库中的共47幅参考图像和1549幅测试图像进行仿真实验,且与SSIM,VSNR,FSIM和PSNRHVS等典型的图像质量客观评价模型进行对比分析,同时探讨影响图像质量评价的因素.结果表明:所提方法的评价分数与主观评价分数的Pearson线性相关性系数和Spearman秩相关系数值比SSIM的评价结果均有一定程度的提高,提高幅度分别平均为9.5402%和3.2852%,比PSNRHVS和VSNR提高幅度更大.综合以上表明:所提方法是一种有效可行的图像质量客观评价方法;同时,在图像质量客观评价中,考虑人眼对图像内容的感知和复杂度的分析有助于提高图像质量主客观评价的一致性,评价精度可得到进一步的提高.

English Abstract

参考文献 (23)

目录

    /

    返回文章
    返回