基于视觉的船舶水下机械臂自主抓取系统研究

doi:10.19423/j.cnki.31-1561/u.2025.183

船舶 ›› 2026, Vol. 37 ›› Issue (04): 124-133.DOI: 10.19423/j.cnki.31-1561/u.2025.183

基于视觉的船舶水下机械臂自主抓取系统研究

张春文¹, 刘晨², 刘义³, 吴乃龙^2,*, 王旭阳¹

1.上海交通大学船舶海洋与建筑工程学院上海 200240;
2.东华大学信息与智能科学学院上海 201620;
3.中国船舶及海洋工程设计研究院上海 200011

收稿日期:2025-11-06 修回日期:2026-03-15 发布日期:2026-08-28
通讯作者: 吴乃龙（1987—）,男,博士,副教授/博士生导师。研究方向：无人系统技术及智能感测;王旭阳（1977—）,男,博士,高级工程师。研究方向：水下机器人和深海作业装备。
作者简介:张春文（1987—）,男,硕士,助理工程师。研究方向：水下机器人;刘晨（1994—）,男,博士研究生。研究方向：智能船舶;刘义（1988—）,女,博士,高级工程师。研究方向：智能船舶。
基金资助:
中央高校基本科研业务费专项资金资助（2232025D-48）

Vision-Based Autonomous Grasping for Shipborne Underwater Robotic Arm Using MB-Grasp

ZHANG Chunwen¹, LIU Chen², LIU Yi³, WU Nailong^2,*, WANG Xuyang¹

1. School of Naval Architecture, Ocean & Civil Engineering, Shanghai Jiao Tong University, Shanghai 200240, China;
2. School of Information and Intelligent Science, Donghua University, Shanghai 201620, China;
3. Marine Design & Research Institute of China, Shanghai 200011, China

Received:2025-11-06 Revised:2026-03-15 Published:2026-08-28

摘要/Abstract

摘要： 为提升船舶与海洋工程水下作业的自主化水平,该文提出一种基于MB-Grasp模型的水下目标物抓取方法。首先通过搭载于作业平台的双目相机获取目标物RGB图像,再结合半全局立体匹配算法得到视差深度图;随后将RGB图像和深度图输入MB-Grasp网络,估计目标物的位置、种类及抓取角,并通过坐标变换计算其在机械臂基坐标系下的位姿;最终利用RRT*（rapidly-exploring random tree star）算法进行路径规划,通过逆运动学求解各路径点关节角,经多项式插值生成关节空间平滑轨迹,驱动机械臂完成抓取。该研究搭建了模拟船舶水下作业环境的实验平台,成功实现了对模拟缆线和小型部件等目标的自主抓取,为船舶水下自主作业提供了可行方案。

关键词: 水下机器人, 机械臂系统, MB-Grasp, 半全局立体匹配算法, RRT*算法, 船舶水下作业

Abstract: To enhance the autonomy of underwater operations in marine and ship engineering, this paper proposes a method for object grasping based on an MB-Grasp model. First, an RGB image of the target is captured by a binocular camera mounted on the operation platform, and a disparity depth map is generated using a semi-global stereo matching (SGM) algorithm. The RGB and depth images are then fed into the MB-Grasp network to estimate the target's position, category, and grasp angle. Subsequently, the target's pose in the robotic arm base coordinate system is derived through coordinate transformation. Path planning is performed using the rapidly-exploring random tree star (RRT*) algorithm, and inverse kinematics solves the joint angles for each path point. A smooth trajectory in joint space is generated via polynomial interpolation to drive the robotic arm to execute the grasp. An experimental platform simulating the underwater operation environment of a ship was constructed, and autonomous grasping of simulated cables and small components was successfully achieved. This study provides a feasible solution for autonomous underwater operations in ship underwater engineering.

Key words: underwater vehicle, manipulator systems, MB-Grasp, semi-global stereo matching (SGM) algorithm, rapidly-exploring random tree star (RRT*) algorithm, ship underwater operations

中图分类号:

张春文, 刘晨, 刘义, 吴乃龙, 王旭阳. 基于视觉的船舶水下机械臂自主抓取系统研究[J]. 船舶, 2026, 37(04): 124-133.

ZHANG Chunwen, LIU Chen, LIU Yi, WU Nailong, WANG Xuyang. Vision-Based Autonomous Grasping for Shipborne Underwater Robotic Arm Using MB-Grasp[J]. Ship & Boat, 2026, 37(04): 124-133.

参考文献

[1] 严新平,刘佳仑,范爱龙,等.智能船舶技术发展与趋势简述[J]. 船舶工程,2020,42(3):15-20.
[2] 徐明浩,司马灿,朱忠,等.水下机器人自主操作——机遇、进展与挑战[J].船舶,2024,35(6):60-80.
[3] 杨学文,王楠.视觉驱动机械臂自主作业技术综述[J]. 南京信息工程大学学报(自然科学版),2021,13(3):281-290.
[4] Jiao H,Wang G,Wang Y,et al.Dynamic robust path-following control of UVMS subject to manipulator disturbance[C]// Journal of Physics:Conference Series. IOP Publishing,2024,2718:012056.
[5] 高剑,何耀祯,陈依民,等. 水下航行器视觉控制技术综述[J]. 水下无人系统学报,2024,32(2):282-294.
[6] 卢有旺,夏英凯,徐国华,等.面向UUV 对接的视觉引导三维轨迹跟踪控制研究[J]. 中国舰船研究,2024,19(1):290-304.
[7] 徐凤强,董鹏,王辉兵,等.基于水下机器人的海产品智能检测与自主抓取系统[J]. 北京航空航天大学学报,2019,45(12):2393-2402.
[8] 孙长迪,潘志松,张艳艳. 基于低开销可变形卷积的 MobileNet 再轻量化方法[J]. 计算机科学,2022,49(12):312-318.
[9] 夏腾辉,王曰英,吴乃龙,等.基于MobileNetV2的船载充电部件检测与岸基机械臂对接方法[J].水下无人系统学报,2025,33(2):359-366.
[10] Gupta S, Girshick R, Arbeláez P, et al.Learning rich features from RGB-D images for object detection and segmentation[C]//Proceedings of the 13th European Conference on Computer Vision,2014:345-360.
[11] Kumra S,Kanan C.Robotic grasp detection using deep convolutional neural networks[C]//2017 IEEE/RSJ International Conference on Intelligent Robots and Systems,2017:769-776.
[12] Sandler M,Howard A,Zhu M,et al.MobileNetV2:Inverted residuals and linear bottlenecks[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition,2018:4510-4520.
[13] Ganesan S,Ramalingam B,Mohan R E.A hybrid sampling-based RRT* path planning algorithm for autonomous mobile robot navigation[J]. Expert Systems with Applications,2024,258:125206-125217.
[14] Zhou Q,Feng H Z,Liu Y Y.Multigene and improved anti-collision RRT* algorithms for unmanned aerial vehicle task allocation and route planning in an urban air mobility scenario[J]. Biomimetics,2024,9(3):125-142.
[15] 殷浩宇. 基于双目视觉的深海机械臂抓取检测方法研究[D]. 合肥:安徽建筑大学,2023.

编辑推荐

Metrics

联系电话：（021）63161688-105006
传真：（021）63151255
E-mail：chuanbo@maric.com.cn

基于视觉的船舶水下机械臂自主抓取系统研究

Vision-Based Autonomous Grasping for Shipborne Underwater Robotic Arm Using MB-Grasp

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 2

编辑推荐

Metrics

[1]	徐明浩, 司马灿, 朱忠, 倪天, 吴海波, 郭佳, 吴雨. 水下机器人自主操作——机遇、进展与挑战[J]. 船舶, 2024, 35(06): 60-80.
[2]	张玲,谢殿伟. 水下机器人惯性类水动力计算研究[J]. 船舶, 2004, 15(03): 8-10.