| 时间: 2021-11-10 | 次数: |
李国清, 汤子梦, 吴喜芳,等.导向滤波和自动白平衡融合的遥感图像除雾算法应用研究[J].河南理工大学学报（自然科学版）,2021,40(6):77-83.
LI G Q, TANG Z M, WU X F, et al.Research on the application of remote sensing image defogging algorithm with guided filtering and automatic white balance fusion[J].Journal of Henan Polytechnic University(Natural Science) ,2021,40(6):77-83.
导向滤波和自动白平衡融合的遥感图像除雾算法应用研究
李国清1, 汤子梦2, 吴喜芳2, 张合兵2, 王春阳2, 李保莲3
1.河南省遥感测绘院，河南 郑州 450000;2.河南理工大学 测绘与国土信息工程学院，河南 焦作 454000;3.河南省国土资源调查规划院，河南 郑州 450016
摘要:针对目前除雾算法实时性差、除雾后影像色彩失真严重等问题，从带雾图像本质出发，对导向滤波和自动白平衡算法进行深入研究，提出一种基于导向滤波和自动白平衡融合的除雾新算法。该算法首先对暗通道后的图像进行导向滤波计算，其次将自动白平衡算法应用于导向滤波处理后的图像中，最后获得除雾后的图像。研究结果表明，该方法能有效解决先前除雾方法引起的色彩暗淡和颜色失真等问题，获取的地物特征更明显，有效提高除雾速度。研究成果解决了遥感图像的全天候使用问题，为遥感图像的除雾算法提供了一种新思路。
关键词:雾天退化模型;导向滤波;自动白平衡;除雾;透射率;信息熵
doi:10.16186/j.cnki.1673-9787.2020020064
基金项目:河南省国土资源科研项目（豫财招标采购-2019-379-7）；国家自然科学基金资助项目（41801318）；河南理工大学博士基金 资助项目（B2021-19）
收稿日期:2021/02/27
修回日期:2021/05/17
出版日期:2021/11/15
Research on the application of remote sensing image defogging algorithm with guided filtering and automatic white balance fusion
LI Guoqing1, TANG Zimeng2, WU Xifang2, ZHANG Hebing2, WANG Chunyang2, LI Baolian3
1.Institute of Remote Sensing Surveying and Mapping of Henan Province， Zhengzhou 450000 ， Henan， China;2.School of Surveying and Land Information Engineering， Henan Polytechnic University， Jiaozuo 454000 ， Henan， China;3.Land and Resources Survey and Planning Institute of Henan Province，Zhengzhou 450016，Henan，China
Abstract:Aiming at the problems of poor real-time performance and serious image color distortion after defog-ging ，a deep research on guided filtering and automatic white balance algorithm was made based on the essence of foggy image and an optimization algorithm combining guided filter and automatic white balance algorithm was proposed. First ， the image after dark channel was calculated by guided filtering ， and then the automatic white balance algorithm was applied to the image after guided filtering to obtain the defogged image. The results showed that the optimization method proposed in this study could effectively solve the problems of color dimming and color distortion caused by previous defogging methods. In addition ， the new defogging method could also obtain more obvious features of ground objects and effectively improve the defogging speed. This study solved the problem of remote sensing image all-weather use and proposed a new idea for remote sensing image defogging algorithm.
Key words:fog degeneration model;guided filtering;auto matic white balance;defogging;transmissivity;information entropy
