基于视觉的集装箱海上并靠吊装位姿识别  

Container Posture Recognition of Skin-to-skin Lifting at Sea Based on Vision

作　　者：卢道华[1] 王超 王佳[1] 王恺 LU Daohua;WANG Chao;WANG Jia;WANG Kai(School of Mechanical Engineering,Jiangsu University of Science and Technology,Jiangsu Zhenjiang 212003,China)

机构地区：[1]江苏科技大学机械工程学院,江苏镇江212003

出　　处：《船舶工程》2020年第11期100-105,共6页Ship Engineering

基　　金：国家重点研发计划资助项目(2018YFC0309100)。

摘　　要：舰船在远海进行并靠补给时受波浪影响两船之间会产生无规律的相对运动,给补给集装箱吊装工作带来了很大挑战,急需一种能够在较高海况下辅助人工的自动化吊装系统。针对这一问题,提出一种基于单目视觉和Ar Uco标志点相结合的集装箱位姿检测方法。首先,介绍了带有视觉系统的并联吊装系统主要组成部分,简要分析了控制流程及方法;然后,根据实际集装箱码放流程,设计了适用的ArUco标志点布放和检测方案,介绍了相机标定、ArUco标志点检测及集装箱位姿估计方法;最后,搭建模拟试验平台测试视觉位姿识别的精度、速度和不同光照条件下的鲁棒性。试验结果表明,在普通PC工控平台上,最低识别速度在15Hz以上,姿态角平均误差为±0.5°,位移量平均误差为±0.4%,可以满足并靠吊装要求。The irregular relative motion between the two ships will occur when the ships are doing skin-to-skin replenishment at the open sea,which brings great challenges to the lifting of supply containers.It is urgently needed to find an automatic lifting system to assist humans under severe sea state.Aiming at the problem,the paper proposes a container pose detection method based on a combination of monocular vision and ArUco landmarks.Firstly,the main components of a parallel movement crane system with vision is introduced,and the control process and method are briefly analyzed.Then,according to the actual container stacking process,an applicable ArUco landmark placement and detection solution is designed.The ArUco landmark detection and container pose estimation method are introduced.Finally,a simulation test platform is built to test the accuracy,speed and robustness under different lighting conditions of visual posture recognition.The test results show that,on a common Industrial computer platform,the minimum recognition speed is above 15 Hz;the average error of the attitude angle is±0.5°;the average error of the displacement is±0.4%;which can meet the requirements of skin-to-skin lifting.

关 键 词：并靠补给 波浪补偿 单目视觉 位姿估计 ArUco 

分 类 号：TP27[自动化与计算机技术—检测技术与自动化装置]