水下机器人回收过程中目标光源识别与定位研究综述  

Review on Target Light Source Identification and Location in the Process of Underwater Robot Recovery

作　　者：王迪 赵文洁 王昱[1] 黄浩乾 王俊玮 王广才 WANG Di;ZHAO Wenjie;WANG Yu;HUANG Haoqian;WANG Junwei;WANG Guangcai(College of Artificial Intelligence and Automation,Hohai University,Changzhou 213022,China;College of Mechanical and Vehicle Engineering,Hunan University,Changsha 410082,China)

机构地区：[1]河海大学人工智能与自动化学院,常州213022 [2]湖南大学机械与运载工程学院,长沙410082

出　　处：《无人系统技术》2025年第2期45-56,共12页Unmanned Systems Technology

基　　金：江苏省自然科学基金(BK20240349);中央高校基本科研业务费专项资金(B250201090,B240201120)。

摘　　要：对水下机器人回收过程中目标光源识别与定位的相关研究进行了综合评述,并对未来发展趋势进行展望。首先阐述水下机器人回收的背景及目标光源识别与定位的重要性;随后从水下光源设计及图像处理、目标光源识别、水下光源视觉定位三个方面,详细总结现有技术与成果;接着分析该领域面临的难点,包括水下成像环境复杂、光源不确定性以及计算量和精度的平衡问题。综述表明,水下目标光源识别与定位技术虽取得一定进展,但仍需深入研究。未来应聚焦于物理模型优化,通过实验和模拟技术建立更精确的光传播模型;实现实时性和智能化,借助硬件技术提升处理速度,融入人工智能增强系统适应性;开展跨学科研究,整合多学科优势资源攻克难题。A comprehensive review of the research on target light source identification and positioning during the recovery of underwater robots is conducted,and the future development trends are prospected.Firstly,the background of underwater robot recovery and the importance of target light source identification and positioning are expounded.Subsequently,the existing technologies and achievements are summarized in detail from three aspects:underwater light source design and image processing,target light source identification,and underwater light source visual positioning.Then,the challenges faced by this field are analyzed,including complex underwater imaging environments,light source uncertainties,and the balance between computational complexity and accuracy.The review indicates that although some progress has been made in underwater target light source identification and positioning technology,in-depth research is still required.In the future,the focus should be on optimizing physical models,establishing more accurate light propagation models through experimental and simulation techniques;achieving real-time performance and intelligence,enhancing processing speed with the aid of hardware technology and integrating artificial intelligence to improve system adaptability;and conducting interdisciplinary research to integrate the superior resources of multiple disciplines to overcome difficulties.

关 键 词：水下机器人 自主回收 水下视觉 目标光源定位 水下对接 视觉导引 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]