基于人眼色彩差异化感知的图像质量评价研究

计算机工程与科学 ›› 2023, Vol. 45 ›› Issue (02): 295-303.

基于人眼色彩差异化感知的图像质量评价研究

王杨1,2,隆海燕1,2,贾曦然1,2

(1.河北工业大学电子信息工程学院，天津300401；2.河北工业大学天津市电子材料与器件重点实验室，天津300401)

收稿日期:2021-07-05 修回日期:2021-11-16 接受日期:2023-02-25 出版日期:2023-02-25 发布日期:2023-02-15
基金资助:
河北省教育厅重点项目（ZD2020304）

Research on image quality evaluation based on color differential perception of human eye

WANG Yang1,2,LONG Hai-yan1,2,JIA Xi-ran1,2

(1.College of Electronics and Information Engineering,Hebei University of Technology,Tianjin 300401;
2.Tianjin Key Laboratory of Electronic Materials & Devices,Hebei University of Technology,Tianjin 300401,China)

Received:2021-07-05 Revised:2021-11-16 Accepted:2023-02-25 Online:2023-02-25 Published:2023-02-15

摘要/Abstract

摘要： 针对显著性图像质量评价问题，参考人眼视觉对图像色彩的差异化感知，提出一种基于孪生神经网络对图像色彩对比显著区域进行质量评价的算法。首先，根据图像中的色彩对比和语义信息分别提取原始图像和失真图像中的色彩对比显著区域；然后，将原始图像和失真图像中对应的色彩区域作为子图像，以样本对的形式输入孪生神经网络；最后，计算主客观图像质量评估值的相关性。实验采用残差结构的Inception-ResNet-V2网络作为基础模型，同时增加EMD损失函数优化对图像质量的距离损失，经过Softmax层后输出图像质量评估值，并在TID2013数据集上进行了测试。结果表明，提出的算法在该数据集上性能良好。

关键词: 显著性图像质量评价, 色彩对比度, 孪生神经网络, 语义提取

Abstract: Aiming at the problem of saliency image quality evaluation, referring to the differential perception of image color by human vision, an algorithm based on siamese neural network for quality evaluation of image color contrast salient regions is proposed. Firstly, according to the color contrast and its semantic information of the image, the significant regions of color contrast are extracted respectively. Secondly, the extracted color regions are inputted into the siamese neural network as sub-images in the form of sample pairs. Finally, the correlation between the subjective and objective image quality evaluation values is calculated. In the experiment, the Inception-ResNet-V2 network with residual structure is used as the basic model. The EMD loss function is introduced to optimize the distance loss. After being filtered through the Softmax layer, the image quality evaluation value is obtained. Tests are carried out on the TID2013 dataset and the test results show that the proposed algorithm has good performance on this dataset.

Key words: saliency image quality evaluation, color contrast, siamese neural network, semantic extraction

王杨, 隆海燕, 贾曦然, . 基于人眼色彩差异化感知的图像质量评价研究[J]. 计算机工程与科学, 2023, 45(02): 295-303.

WANG Yang, LONG Hai-yan, JIA Xi-ran, . Research on image quality evaluation based on color differential perception of human eye[J]. Computer Engineering & Science, 2023, 45(02): 295-303.

参考文献［25］

［1］	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.
［2］	Itti L,Koch C,Niebur E.A model of saliency-based visual attention for rapid scene analysis［J］.IEEE Transactions on Pattern Analysis and Machine Intelligence,1998,20(11):1254-1259.
［3］	Liu Di,Li Ying-chun.Quality assessment of remote sensing images based on deep learning and human visual system［J］.Laser & Optoelectronics Progress,2019,56(6):161-169.(in Chinese)
［4］	Jia Hui-zhen,Wang Tong-han,Fu Peng.Multi-feature fusion based image quality assessment method［J］.Pattern Recognition and Artificial Intelligence,2019,32(7):669-675.(in Chinese)
［5］	Shen Fei-peng,Zhu Tong,Zhang He-nan,et al.Non-reference blur image quality evaluation based on saliency object classification［J］.Laser & Optoelectronics Progress,2021,58(22):237-247.(in Chinese)
［6］	García-Lamont F,Cervantes J,López-Chau A,et al.Color image segmentation using saturated RGB colors and decoupling the intensity from the hue［J］.Multimedia Tools and Applications, 2020, 79(1):1555-1584.
［7］	Xu Li-na, Xiao Qi,He Lu-xiao.Fused image quality assessment based on human visual characteristics［J］.Geomatics and Information Science of Wuhan University,2019,44(4):546-554.(in Chinese)
［8］	Hou Chun-ping,Li Hao,Yue Guang-hui.Quality assessment of tone-mapped images using local and global features［J］.Journal of Hunan University (Natural Sciences),2019,46(8):132-140.(in Chinese)
［9］	Zhao Yue,Wang Lai-hua,Qi Su-min,et al.No-reference quality assessment of night-time images based on global and local features［J］.Computer Engineering & Science,2021,43(7):1283-1290.(in Chinese)
［10］	Chopra S,Hadsell R,Lecun Y.Learning a similarity metric discriminatively，with application to face verification［C］∥Proc of 2005 IEEE Conference on Computer Vision and Pattern Recognition,2005:539-546.
［11］	Yang Guang-yi,Ye Ling-xuan,Long Min-yi,et al.Full re- ference color image quality assessment algorithm based on siamese network［J］.Journal of Wuhan University(Natural Science Edition),2022,66(6):589-596.(in Chinese)
［12］	Gregory R L. Eye and brain［M］.Qu Jin-chun, Zhang Fen-fen, translation. Taibei:WU-Nan Book Incorporation,2006.(in Chinese)
［13］	Yuan Ya-hui.Research on saliency detection methods based on the visual sensitivities of RGB colors［D］. Jinan:Shandong Univerity,2016.(in Chinese)
［14］	Raman C V.Light,colour and vision［J］.Current Science,1959,28(11):525.
［15］	Piao Da-zhi,Li Zai-qing. Advancement on research of human vision characteristics［J］.Modern Measurement and Test,1996,4(1):2-6.(in Chinese)
［16］	Nadenau M. Integration of human color vision models into high quality image compression［D］.Lausanne,Switzerland:école Polytechnique Fédérale de Lausanne,2000.
［17］	Yao J.Measurements of human vision contrast sensitivity to opposite colors using a cathode ray tube display［J］.Chinese Science Bulletin,2011,56(23):2425-2432.
［18］	Guo Ying-chun,Li Zhuo.Video saliency detection based on edge features and adaptive fusion［J］.Journal of Hebei University of Technology,2019,48(1):1-7.(in Chinese)
［19］	Wang Peng-yu, You You-peng,Yang Xue-feng.Color image segmentation based on color quantization and density peak clustering［J］.Computer Engineering and Applications,2020,56(2):211-215.(in Chinese)
［20］	Avcibasi I,Sanbu B,Satood K.Statistical evaluation of image quality measures［J］.Journal of Electronic Imaging,2002,11(2):206-213.
［21］	Zhang L,Zhang L,Mou X Q,et al.FSIM:A feature similarity index for image quality assessment［J］.IEEE Transactions on Image Processing,2011,20(8):2378-2386.
［22］	Chandler D M, Hemami S S. VSNR:A wavelet-based visual signal-to-noise ratio for natural images［J］.IEEE Transactions on Image Processing,2007,16(9):2284-2298.
［23］	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):1-21.
［24］	Nafchi H Z,Shahkolaei A,Hedjam R,et al.Mean deviation similarity index:Efficient and reliable full-reference image quality evaluator［J］.IEEE Access,2016,4:5579-5590.
［25］	Yao Jun-cai, Shen Jing.Objective assessment of image quality based on image content contrast perception［J］.Acta Physica Sinica,2020,69(14):287-302.(in Chinese)
	附中文参考文献：
［3］	刘迪,李迎春.基于深度学习和人眼视觉系统的遥感图像质量评价［J］.激光与光电子学进展,2019,56(6):161-169.
［4］	贾惠珍,王同罕,傅鹏.多特征融合的图像质量评价方法［J］.模式识别与人工智能,2019,32(7):669-675.
［5］	沈飞鹏,朱彤,张赫男,等.基于显著性目标分类的无参考模糊图像质量评价［J］.激光与光电子学进展,2021,58(22):237-247.
［7］	许丽娜,肖奇,何鲁晓.考虑人类视觉特征的融合图像评价方法［J］.武汉大学学报(信息科学版),2019,44(4):546-554.
［8］	侯春萍,李浩,岳广辉.局部和全局特征融合的色调映射图像质量评价［J］.湖南大学学报(自然科学版),2019,46(8):132-140.
［9］	赵月,王来花,齐苏敏,等.基于全局和局部特征的无参考夜间图像质量评价［J］.计算机工程与科学,2021,43(7):1283-1290.
［11］	杨光义,叶凌轩,龙敏义,等.基于孪生网络的全参考彩色图像质量评价算法［J］.武汉大学学报（理学版）,2022，66(6):589-596.
［12］	Gregory R L. 视觉心理学［M］.瞿锦春，张芬芬，译.台北：五南图书出版有限公司，2006.
［13］	苑亚会.基于RGB视觉敏感度的显著性检测方法研究［D］.济南：山东大学,2016.
［15］	朴大植,李在清.人眼视觉特性研究的进展［J］.现代计量测试,1996,4(1):2-6.
［18］	郭迎春,李卓.基于边缘特征和自适应融合的视频显著性检测［J］.河北工业大学学报,2019,48(1):1-7.
［19］	王鹏宇,游有鹏,杨雪峰.基于颜色量化和密度峰聚类的彩色图像分割［J］.计算机工程与应用,2020,56(2):211-215.
［25］	姚军财,申静.基于图像内容对比感知的图像质量客观评价［J］.物理学报,2020,69(14):287-302.