摘要
针对轨道交通门系统门扇拆装过程中人工托举存在劳动强度大、安全隐患等问题,为实现门扇自动化抓取、搬运等任务,本研究提出一种基于视觉的门扇定位抓取方法。采用2D视觉相机获取门扇玻璃窗边缘图像,通过对图像进行中值滤波、灰度化等处理后进行边缘线的特征识别及Canny检测,随后采用Hough变换算法获取玻璃窗边缘线信息,最后以两侧、下侧边缘线交点作为定位基准,实时获得门扇相对于基准点的位置偏移量,从而引导搬运机构完成门扇抓取、搬运等任务。实验验证了所提方法的有效性,达到了减轻工人劳动强度、提高作业安全性的目标。与传统方法相比,本文所提定位方法实现方式简单,所采用的视觉识别算法识别效率高、准确度高,便于现场工程技术人员实施。
Aiming at the issues of the high labor intensity and safety during the manual installation process of the door leaves in the rail transit systems,this study proposes the vision—based positioning method for the door leaves to achieve the automated grasping task.A 2D vision camera is used to capture the images of the edges of the door leaf window.The images are processed with the methods of the median filter and grayscale conversion,and the feature recognition of edge lines is conducted with the Canny detection.Subsequently the Hough transform algorithm is employed to obtain the edge lines information of the door leaf window.Finally,the intersection points of the left,right,and bottom edge lines are obtained as the positioning references to determine the real-time position offset of the door leaf relative to the reference point;and then the handling mechanism is guided to complete the tasks of grasping and transporting the door.The effectiveness of the proposed method is validated with the experimental results,achieving the goals of reducing workers'labor intensity and improving operational safety.Compared with the traditional grasping method,the proposed method has the advantages of the hardware and software implementation,the higher efficiency and accuracy,making it easy to be conducted for the field technicians.
作者
张飞
陈婷
刘东海
李波
ZHANG Fei;CHEN Ting;LIU Donghai;LI Bo(School of Intelligent Manufacturing,Guangxi Vocational&Technical College,Nanning 530226,China;College of Mechanical&Electrical Engineering,Nanjing University of Aeronautics and Astronautics,Nanjing 210016,China)
出处
《智能制造》
2026年第1期43-48,共6页
Intelligent Manufacturing
基金
2022年江苏省卓越博士后计划(2022ZB814)
2022年紫金山英才先锋计划(宁开委科字[2022]115号)
2023年江苏省双创人才计划(JSSCRC2023374)
广西职业技术学院校级课题基金项目(桂职院[2025]131号251105y)。
