广东工业大学学报 ›› 2021, Vol. 38 ›› Issue (05): 33-39.doi: 10.12052/gdutxb.210036

• • 上一篇    下一篇

协同超像素和视觉显著性的图像质量评价

邓杰航, 袁仲鸣, 林好润, 顾国生   

  1. 广东工业大学 计算机学院，广东 广州 510006
  • 收稿日期:2021-03-04 出版日期:2021-09-10 发布日期:2021-07-13
  • 通信作者: 顾国生(1978–)，男，讲师，主要研究方向为多媒体信息安全和图像处理，E-mail：gsgu@gdut.edu.cn E-mail:gsgu@gdut.edu.cn
  • 作者简介:邓杰航(1979–)，男，副教授，博士研究生，主要研究方向为图像处理和模式识别，E-mail：dengjiehang@163.com
  • 基金资助:
    国家自然科学基金资助项目(61202267)；广东省重点领域研发项目(2019B010139002)；广州市科技计划项目(201902020007，202007010004，201807010058)

Superpixel and Visual Saliency Synergetic Image Quality Assessment

Deng Jie-hang, Yuan Zhong-ming, Lin Hao-run, Gu Guo-sheng   

  1. School of Computers, Guangdong University of Technology, Guangzhou 510006, China
  • Received:2021-03-04 Online:2021-09-10 Published:2021-07-13

HTML

PDF (PC)

2510

摘要： 为实现全参考图像质量客观评估与人类主观评估更高的一致性, 本文提出了协同超像素和视觉显著性双重策略的图像质量评价方法。该方法通过融合4个图像特征相似度得到局部图像质量得分。这4个相似度分别是超像素局部亮度相似度、超像素局部色度相似度、视觉显著性相似度和Scharr梯度相似度。为了解决过去的评价方法中不同的特征相似度仅凭经验确定参数的问题, 提出相似度量参数修正模型对各相似度的参数进行自适应调整。最终的质量得分由视觉显著性构造的加权函数与局部质量得分池化获得。大量的比较实验表明, 本文方法的综合性能表现优异, 与主观评估具有更高的相关性。

关键词: 全参考图像质量评价, 超像素, 视觉显著性, 梯度, 参数自适应模型

Abstract: In order to achieve higher consistency between objective and subjective evaluation of image quality, a full-reference image quality evaluation method based on a synergetic strategy between superpixel and visual saliency is proposed. This method obtains the image quality score by pooling four similarities of the image features. The four similarities are: the superpixel-based local luminance similarity, the superpixel-based local chrominance similarity, the visual saliency similarity, and the Scharr gradient similarity. In particular, a parameter correction model is proposed to adaptively adjust the parameters of each similarity, to address the problem of determining the parameters of each similarity empirically. The visual saliency is introduced to design a weighting function to fuse the four similarities to obtain the global quality score. A large number of comparative experiments show that the proposed method outperforms the baseline full-reference image quality assessment methods, and has a higher correlation with the subjective evaluation.

Key words: full reference image quality assessment, superpixel, visual saliency, gradient, adaptive parameter model

中图分类号: 

  • TP391.41
[1] WANG Z, BOVIK A C. Modern image quality assessment[M]. New York: Morgan and Claypool Publishing Company, 2006.
[2] 褚江, 陈强, 杨曦晨. 全参考图像质量评价综述[J]. 计算机应用研究, 2014, 31(1): 13-22.
CHU J, CHEN Q, YANG X C. Review on full reference image quality assessment algorithms [J]. Application Research of Computers, 2014, 31(1): 13-22.
[3] WANG Z, BOVIK A C. Mean squared error: love it or leave it? [J]. IEEE Signal Processing Magazine, 2009, 26(1): 98-17.
[4] 高方远, 何立火. 基于深度网络和视觉特性的无参考图像质量评价方法[J]. 南京师大学报(自然科学版), 2019, 42(3): 20-26.
GAO F Y, HE L H. No-reference image quality assessment based on deep network and visual characteristics [J]. Journal of Nanjing Normal University(Natural Science Edition), 2019, 42(3): 20-26.
[5] 马华林, 张立燕. 基于人眼感知的无参考色调映射图像质量评价[J]. 计算机应用与软件, 2020, 37(7): 158-164.
MA H L, ZHANG L Y. No-reference Tone-mapper image quality assessment based on human eye perception [J]. Computer Applications and Software, 2020, 37(7): 158-164.
[6] 崔力. 基于多特征相似性融合的半参考图像质量评价[J]. 陕西科技大学学报, 2019, 37(6): 160-164.
CUI L. Reduced-reference image quality assessment base on multi-feature similarity fusion [J]. Journal of Shanxi University of Science & Technology, 2019, 37(6): 160-164.
[7] 吴世海, 鲍义东, 陈果, 等. 基于随机Gabor特征的半参考农作物图像质量评价方法[J]. 计算机与现代化, 2020(5): 70-74.
WU S H, BAO Y D, CHEN G, et al. Reduced-reference crop image quality assessment based on random Gabor feature [J]. Computer and Modernization, 2020(5): 70-74.
[8] WANG Z, BOVIK A C, SHEIKH H R, et al. Image quality assessment: from error visibility to structural similarity [J]. IEEE Transactions on Image Processing, 2004, 13(4): 600-612.
[9] WANG Z, SIMONCELLI E P, BOVIK A C. Multi-scale structural similarity for image quality assessment[C]//The Thirty-Seventh Asilomar Conference on Signals, Systems and Computers. Pacific Grove, CA: IEEE Signal Processing Society, 2003: 1398-1402.
[10] WANG Z, LI Q. Information content weighting for perceptual image quality assessment [J]. IEEE Transaction on Image Processing, 2011, 20(5): 1185-1198.
[11] ZHANG L, ZHANG L, MOU X Q, et al. FSIM: a feature similarity index for image quality assessment [J]. IEEE Transaction on Image Processing, 2011, 20(8): 2378-2386.
[12] LARSON E C, CHANDLER D M. Most apparent distortion: full-reference image quality assessment and the role of strategy [J]. Journal of Electronic Imaging, 2010, 19(1): 011006.
[13] XUE W F, ZHANG L, MOU X Q, et al. Gradient magnitude similarity deviation: a highly efficient perceptual image quality index [J]. IEEE Transaction on Image Processing, 2014, 23(2): 684-695.
[14] SUN W, LIAO Q M, XUE J H, et al. SPSIM: a superpixel-based similarity index for full-reference image Quality Assessment [J]. IEEE Transaction on Image Processing, 2018, 27(9): 4232-4244.
[15] ZHANG L, SHEN Y, LI H Y. VSI: a visual saliency-induced index for perceptual image quality assessment [J]. IEEE Transaction on Image Processing, 2014, 23(10): 4270-4281.
[16] 高俊艳, 刘文印, 杨振国. 结合注意力与特征融合的目标跟踪[J]. 广东工业大学学报, 2019, 36(4): 18-23.
GAO J Y, LIU W Y, YAN Z G. Object tracking combined with attention and feature fusion [J]. Journal of Guangdong University of Technology, 2019, 36(4): 18-23.
[17] HAREL J, KOCH C, PERONA P. Graph-based visual saliency[C]//Proceedings of the 2006 Conference. Advances in Neural Information Processing Systems 19. Vancouver: MIT Press, 2007: 545-552.
[18] ACHANTA R, SHAJI A, SMITH K et al. SLIC superpixels compared to state-of-the-art superpixel methods [J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2012, 34(11): 2274-2281.
[19] ZHANG X D, FENG X C, WANG W W et al. Edge strength similarity for image quality assessment [J]. IEEE Signal Processing Letters, 2013, 20(4): 319-322.
[20] 高敏娟, 党宏社, 魏立力, 等. 结合全局与局部变化的图像质量评价[J]. 自动化学报, 2020, 46(12): 1-10;.
GAO M J, DANG H S, WEI L L et al. Combining global and local variation for image quality assessment [J]. Acta Automatica Sinica, 2020, 46(12): 1-10;.
[21] 高敏娟, 党宏社, 魏立力, 等. 视觉显著性引导的局部图像质量度量[J]. 中国科学: 技术科学, 2019, 49(11): 1350-1360.
GAO M J, DANG H S, WEI L L, et al. Visual saliency induced local image quality metrics [J]. Scientia Sinica (Technological), 2019, 49(11): 1350-1360.
[22] 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[J]. 物理学报, 2020, 69(22): 401-41.
SHI C Y, LIN Y D. Objective image quality assessment based on image color appearance and gradient features [J]. Acta Physica Sinica, 2020, 69(22): 401-41.
[1] 赖峻, 刘震宇, 刘圣海. 基于全局数据混洗的小样本数据预测方法[J]. 广东工业大学学报, 2021, 38(03): 17-21.
[2] 张舒, 莫赞, 柳建华, 杨培琛, 刘洪伟. 基于NWD集成算法的多粒度微博用户兴趣画像构建[J]. 广东工业大学学报, 2020, 37(04): 42-50.
[3] 朱鉴, 张浩晨, 陈炳丰, 蔡瑞初. 基于涡粒子的真实感烟雾快速模拟[J]. 广东工业大学学报, 2019, 36(03): 25-31.
[4] 伍冠健, 宋立锋. HEVC快速帧内模式和深度决策算法[J]. 广东工业大学学报, 2015, 32(04): 132-137.
[5] 姚蕾. 一种粒子群-Mamdani模糊神经网络的参数优化新算法[J]. 广东工业大学学报, 2014, 31(1): 36-39.
[6] 吴建宏, 李伟彤, 廖捷. 一种基于小波-Contourlet的改进图像融合算法[J]. 广东工业大学学报, 2011, 28(1): 12-15.
[7] 刘立程；. 改进的基于形态学梯度法的车辆图像边缘检测方法[J]. 广东工业大学学报, 2007, 24(2): 84-86.
[8] 刘立程； . 图像聚类分析法用于汽车图像边缘的检测[J]. 广东工业大学学报, 2006, 23(2): 69-73.
[9] 郑光昭； . 剪切电子散斑干涉术[J]. 广东工业大学学报, 2003, 20(1): 23-27.
[10] 戴青云； 余英林； 张大鹏； . 掌纹身份识别系统中的定位分割技术[J]. 广东工业大学学报, 2002, 19(1): 1-6.
[11] 赵锋； 余业球； 黎沃光； . BFe30-1-1白铜杆的热型连铸[J]. 广东工业大学学报, 2002, 19(1): 75-80.
[12] 曾文曲； 文有为； 孙海卫； . 图像恢复中的一种新预处理算子[J]. 广东工业大学学报, 2000, 17(2): 76-81.
Viewed
Full text


Abstract

Cited
  Shared   
  Discussed   
No Suggested Reading articles found!
版权所有 © 2017 《广东工业大学学报》编辑部
地址：广州市东风东路729号广东工业大学 邮编：510090
电话：020-37626653,020-37626139,020-37626396
邮箱：xbzrb@gdut.edu.cn
本系统由北京玛格泰克科技发展有限公司设计开发