基于机器视觉的堆叠货箱抓取系统设计  

Design of Grasping System of Stacked Container Based on Machine Vision

作　　者：徐志祥[1] 孙文博 高东 刘玮 XU Zhixiang;SUN Wenbo;GAO Dong;LIU Wei(School of Mechanical Engineering,Dalian University of Technology,Dalian 116024,China;Jiuyi Hangyu(Dalian)Technology Co.,Ltd.,Dalian 116085,China)

机构地区：[1]大连理工大学机械工程学院,辽宁大连116024 [2]久亿航宇科技(大连)有限公司,辽宁大连116085

出　　处：《机械制造与自动化》2024年第5期199-202,共4页Machine Building & Automation

摘　　要：针对冷链运输领域自动化水平低的问题,设计一套基于机器视觉的堆叠货箱抓取系统。基于Matlab标定工具完成相机标定,得到坐标系转换数学模型,使用改进Canny边缘检测算法提取图像中堆叠货箱的位置信息,通过PLC控制卸货机器人按照既定程序实现对冷链集装箱中堆叠货箱的抓取作业。实验结果表明:图像处理时间小于2 ms,抓取效率可达5 s/件,抓取位置误差小于5 mm,完全满足工业生产要求。Aimed at the low automation level of cold chain transportation,a set of stacking container grasping system based on machine vision was designed.Matlab calibration tool was used to complete the camera calibration to obtain the mathematical model of coordinate system transformation.Improved Canny edge detection algorithm was applied to extract the position information of stacked cargo boxes in the image.By PLC,the unloading robot was controlled to grasp stacked cargo boxes in cold chain containers according to the established procedures.The experimental results show that the image processing time is less than 2 ms,the grasping efficiency is up to 5 s/piece,and the grasping position error is less than 5 mm,well meeting the requirements of industrial production.

关 键 词：机器视觉 CANNY边缘检测 PLC 卸货机器人 抓取系统 堆叠货箱 

分 类 号：TP242.3[自动化与计算机技术—检测技术与自动化装置]