水下图像处理技术研究综述  

Survey of Underwater Image Processing Technology

作　　者：林明星[1,2,3] 代成刚[1,2,3] 董雪 张东 管志光[6] LIN Ming-xing;DAI Cheng-gang;DONG Xue;ZHANG Dong;GUAN Zhi-guang(School of Mechanical Engineering,Shandong University,Jinan 250061,China;Key laboratory of High Efciency and Clean Mechanical Manufacture of Ministry of Education,Jinan 250061,China;National Demonstration Center for Experimental Mechanical Engineering Education(Shandong University),Jinan 250061,China;Techanical and Flectrical Engineering Department,Shandong Transport Vocational College,Weifang 261000,China;Institute of Automation Shandong Academy of Sciences,Jinan 250013,China;School of Constrauction Machinery,Shandong Jiaotong University,Jinan 250023,China)

机构地区：[1]山东大学机械工程学院,山东济南250061 [2]高效洁净机械制造教育部重点实验室,山东济南250061 [3]机械工程国家级实验教学示范中心(山东大学),山东济南250061 [4]山东交通职业学院机电工程系,山东潍坊261000 [5]山东省科学院自动化研究所,山东济南250013 [6]山东交通学院工程机械学院,山东济南250023

出　　处：《测控技术》2020年第8期7-20,共14页Measurement & Control Technology

基　　金：山东省重点研发计划(重大科技创新工程)(2019JZZY020703)。

摘　　要：随着自主式水下机器人的发展,水下探测技术成为新的研究热点。然而,吸收效应和散射效应导致水下获取的图像存在雾化和色彩偏差等缺陷。降质的水下图像在一定程度上降低了水下目标识别的准确性。为了改善水下图像质量,国内外学者对水下图像处理方法进行了深入研究。因水下图像处理方法对提升水下目标识别准确性具有良好的促进作用,故其具有重要的研究与分析价值。介绍了水下成像模型,分析了水下图像视觉质量下降的原理;根据水下物理成像模型将水下图像处理方法分为水下图像增强与水下图像复原,并分别对两类方法的研究现状进行分析与归纳;最后,总结与讨论了各类方法的优缺点,并展望了未来的发展方向。Underwater detection technology has become a new research focus with the development of autonomous underwater vehicles.However,underwater images present hazy sense and color deviation due to the scattering and absorption effects of underwater light.The degraded underwater images diminish the accuracy of underwater target recognition to some extent.In order to improve the quality of underwater images,many scholars have made profound researches on underwater image processing methods.The underwater image processing method improves the accuracy of underwater target recognition,which is worth to study and analyze.Accordingly,the underwater imaging model is introduced,and the mechanism of degradation of underwater image visual quality is analyzed.Furthermore,underwater image processing methods are classified to underwater image enhancement and underwater image restoration approaches according to the underwater imaging model,and the research status of these methods are analyzed and summarized respectively.Finally,the advantages and disadvantages of the various methods are summarized and discussed,and the future development direction is prospected.

关 键 词：水下光视觉成像 水下光视觉信号处理 水下图像增强 水下图像复原 

分 类 号：TP37[自动化与计算机技术—计算机系统结构]