[1]张一鸣,杨曦晨.基于特征融合的雾化图像质量评价方法[J].计算机技术与发展,2022,32(11):72-80.[doi:10. 3969 / j. issn. 1673-629X. 2022. 11. 011]
　ZHANG Yi-ming,YANG Xi-chen.Hazy Image Quality Assessment Based on Multi-feature Fusion[J].,2022,32(11):72-80.[doi:10. 3969 / j. issn. 1673-629X. 2022. 11. 011]
点击复制

基于特征融合的雾化图像质量评价方法()
分享到：

《计算机技术与发展》[ISSN:1006-6977/CN:61-1281/TN]

卷:
32
期数:
2022年11期
页码:
72-80
栏目:
媒体计算
出版日期:
2022-11-10

文章信息/Info

Title:
Hazy Image Quality Assessment Based on Multi-feature Fusion
文章编号:
1673-629X(2022)11-0072-09
作者:
张一鸣; 杨曦晨
南京师范大学 计算机与电子信息 / 人工智能学院,江苏 南京 210046
Author(s):
ZHANG Yi-ming; YANG Xi-chen
School of Computer and Electronic Information / School of Artificial Intelligence,Nanjing Normal University,Nanjing 210046,China
关键词:
数字图像处理; 图像质量评价; 雾化失真; 特征融合; 支持向量回归
Keywords:
digital image process; image quality assessment; hazy distortion; feature fusion; SVR
分类号:
TP751. 1
DOI:
10. 3969 / j. issn. 1673-629X. 2022. 11. 011
摘要:
针对雾、霾等天气导致监控图像质量降低的问题,提出一种基于特征融合的雾化图像质量评价方法。 该方法首先提取雾化图像的梯度方向特征、梯度强度特征和亮度特征,计算雾化图像的局部梯度强度标准差和局部亮度标准差以分析图像降质所导致局部结构信息变化。 其次,计算参考图像和失真图像不同类型特征之间的差异得到差异特征图。 最后,融合不同类型的差异特征实现雾化图像质量的定量评价。 为验证该方法的性能,分别在雾化图像数据库 exBeDDE、公开自然场景图像数据库与公开截屏图像数据库上,将该方法与雾化图像数据库自带的雾化图像质量评价方法以及几种主流的全参考图像质量评价方法进行了性能比较。 实验结果表明,该方法能够准确评价雾化图像的质量,适用于不同失真类型,与人眼主观评价结果有较高的一致性。
Abstract:
Aiming at the problem of the quality degradation caused by the bad weather such as fog and haze,an hazy image assessmentbased on multi - feature difference is proposed. Firstly, the gradient direction, gradient intensity and brightness are extracted, and thestandard deviation of local gradient intensity and local brightness of the hazy image are employed to represent? ? ? the variations of localstructure information caused by image degradation. Secondly,the distance between different types of features of the reference image andthe distorted image is calculated to obtain the difference feature map. Finally,the difference features are fused to train image quality assessment model to measure the quality of hazy image. To verify the performance of the proposed method,it is compared with the hazyimage quality evaluation method of hazy image database and several mainstream full-reference image quality evaluation methods on hazyimage database exBeDDE,open natural scene image database and open screenshot image database,respectively. The experimental resultsdemonstrate that the proposed method can evaluate the quality of the hazy image accurately,and is also suitable for different distortiontypes with a high consistency with the subjective quality assessment results.

相似文献/References:

[1]刘萌萌.基于无标度摄像机的车流跟踪与速度估计算法[J].计算机技术与发展,2008,(06):111.
　LIU Meng-meng.Algorithm on Vehicle Tracking and Speed Estimating Based on Non- Calibrated Camera[J].,2008,(11):111.
[2]刘学练 张航 熊富强.TMS320在交通流视频检测中的应用[J].计算机技术与发展,2008,(12):200.
　LIU Xue-lian,ZHANG Hang,XIONG Fu-qiang.Traffic Flow Video Detection Achievement Based on TMS320DSP[J].,2008,(11):200.
[3]朱大龙 明军.基于结构失真的图像质量评价方法的研究[J].计算机技术与发展,2006,(03):56.
　ZHU Da-long,MING Jun.Research of Image Quality Assessment Based on Structural Distortion[J].,2006,(11):56.
[4]马飞 吕海莲 杨帅 程荣花.基于图像处理的客观题自动阅卷系统研究开发[J].计算机技术与发展,2012,(07):242.
　MA Fei,LU Hai-lian,YANG Shuai,et al.Research of Objective Examination Paper System Automatically Marking Based on Image Processing[J].,2012,(11):242.
[5]王宇庆.基于梯度复数矩阵的图像质量客观评价方法[J].计算机技术与发展,2013,(01):63.
　WANG Yu-qing.Objective Image Quality Assessment Based on Gradient Complex Matrix[J].,2013,(11):63.
[6]肖鹏,闫建国,赵元伟.基于图像处理的堆积物计数方法研究[J].计算机技术与发展,2013,(09):182.
　XIAO Peng,YAN Jian-guo,ZHAO Yuan-wei.Research on Counting Methods of Deposits Based on DSP[J].,2013,(11):182.
[7]安军,周宁宁. 一种基于视觉注意模型的SSIM改进方法[J].计算机技术与发展,2015,25(01):226.
　AN Jun,ZHOU Ning-ning. An Improved Method of SSIM Based on Visual Attention Model[J].,2015,25(11):226.
[8]李江龙,鲍义东,陈 果.基于视觉显著性的农作物图像评价方法研究[J].计算机技术与发展,2020,30(04):211.[doi:10. 3969 / j. issn. 1673-629X. 2020. 04. 040]
　LI Jiang-long,BAO Yi-dong,CHEN Guo.Study on Crop Image Evaluation Methods Based on Visual Saliency[J].,2020,30(11):211.[doi:10. 3969 / j. issn. 1673-629X. 2020. 04. 040]
[9]孙鹏崴,王 俊,王树军,等.基于 MATLAB GUI 的图像处理系统的设计[J].计算机技术与发展,2022,32(04):215.[doi:10. 3969 / j. issn. 1673-629X. 2022. 04. 037]
　SUN Peng-wei,WANG Jun,WANG Shu-jun,et al.Design of Image Processing System Based on MATLAB GUI[J].,2022,32(11):215.[doi:10. 3969 / j. issn. 1673-629X. 2022. 04. 037]

更新日期/Last Update: 2022-11-10