水下视觉环境感知方法与技术  

Methods and Technologies for Visual Perception of Underwater Environment

作　　者：喻俊志[1,2] 孔诗涵 孟岩 YU Junzhi;KONG Shihan;MENG Yan(Department of Advanced Manufacturing and Robotics,College of Engineering,Peking University,Beijing 100871,China;State Key Laboratory of Management and Control for Complex Systems,Institute of Automation,Chinese Academy of Sciences,Beijing 100190,China)

机构地区：[1]北京大学工学院先进制造与机器人系,北京100871 [2]中国科学院自动化研究所复杂系统管理与控制国家重点实验室,北京100190

出　　处：《机器人》2022年第2期224-235,共12页Robot

基　　金：国家自然科学基金(61725305,U1909206,61633004);博士后创新人才支持计划(BX2021010).

摘　　要：有效感知水下环境是水下机器人自主运动与作业的关键。本文首先对水下视觉环境感知的共性问题进行分析,总结了视觉退化和折射畸变两大共性问题对水下视觉成像产生的影响。其次,围绕如何应对上述挑战,分别从水下视觉恢复方法、水下3维视觉感知以及水下视野增强方法3个方面对国内外主流的水下视觉环境感知方法的研究现状展开综述。最后,对水下视觉环境感知的未来发展趋势进行展望,重点包括集群化视觉感知、多模态感知信息融合、仿生视觉感知等热点研究方向。Effective perception of underwater environment is the key to the autonomous movements and operations of underwater robots.Firstly,the common problems of underwater visual environment perception are analyzed,and for two of them,i.e.visual degradation and refractive distortion,the influences on underwater visual imaging are summarized.Secondly,the research status of mainstream underwater visual environment perception methods at home and abroad is reviewed from 3 aspects of underwater vision restoration methods,underwater 3D visual perception methods,and underwater vision enhancement methods,to tackle the above-mentioned challenges.Finally,the future development tendencies of underwater visual environment perception are prospected,focusing on the hot research directions such as clustered visual perception,information fusion of multi-modal perception,as well as bionic visual perception.

关 键 词：水下视觉感知 水下机器人 水下图像恢复 水下3维视觉 水下视野增强 

分 类 号：TP242.6[自动化与计算机技术—检测技术与自动化装置]