电解电容视觉定位算法研究  

Research on Vision Location Algorithm of Electrolytic Capacitor

作　　者：朱晓凡 严国平 钟飞[1] 黄振[1] ZHU Xiaofan;YAN Guoping;ZHONG Fei;HUANG Zhen(School of Mechanical Engineering,Hubei Univ.of Tech.,Hubei Wuhan 430068 China)

机构地区：[1]湖北工业大学机械工程学院,湖北武汉430068

出　　处：《湖北工业大学学报》2020年第2期18-21,共4页Journal of Hubei University of Technology

基　　金：湖北省技术创新专项(2018AAA026);湖北工业大学博士启动基金(BSQD2016003)。

摘　　要：工业电解电容的生产包装中,需要精确定位电容的位置和针脚朝向,视觉系统获取到电容的位置信息后,将其传递给机械手,抓取电容。根据此需求,提出基于图像二阶矩的图像识别算法,即采用一阶矩获取电容图像的质心,然后采用二阶矩获取电容图像的倾斜角度;利用电容图像的外接矩形获取电容的中心位置,通过外接矩形中心与电容图像质心的差,可以计算出电容针脚的朝向,从而确定电容的位姿。结合实验与仿真的数据,该算法的运行时间为748ms,识别准确度为100%,电容识别和抓取的平均用时为1s,满足工业要求。In the industrial production and packaging of electrolytic capacitors,it is necessary to accurately locate the position and pin orientation of capacitors.The position information of capacitor captured by the vision system is transmitted to the robot hand to grasp the capacitor.According to this requirement,an image recognition algorithm based on second moments is proposed.The first moment is used to get the centroid of the capacitance image,and the second moment is used to get the tilt angle of the capacitance image.The outer rectangle of the capacitance image is used to get the central position of the capacitance.Through the difference between the center of rectangle and the barycenter of capacitance image,the orientation of capacitance pin can be calculated and the position and posture of capacitance can be determined.Combining the experimental and simulation data,the running time of the algorithm is 748 ms and the recognition accuracy is 100%.The average time of capacitance identification and grabbing is 1 s,which meets the industrial requirements.

关 键 词：电容装箱 最小外接矩形 图像二阶矩 位置识别 

分 类 号：TP278[自动化与计算机技术—检测技术与自动化装置]