摘要
为了对生产线上的轮毂进行识别分类,本文开发了一套基于OpenCV和MFC平台的轮毂型号在线识别系统.首先提取轮毂的高度、外直径、中心孔直径、辐条数目、幅窗的周长面积比等特征参数.其中,通过图像预处理、边缘检测、圆拟合、系统标定等方法获取轮毂外直径,来表征各类轮毂的尺寸;通过提取辐条数目、中心孔直径、幅窗的周长面积比等具有旋转不变性的常量来表征各类轮毂的形状.然后为提取到的特征参数生成序列号,作为型号识别的特征参数.最后将生成的特征序列号与模板库中的标准数值进行比对,达到在线实时分类的效果.实验结果表明:该系统的识别准确率为98.7%,能够有效地完成轮毂的在线识别分类,为轮毂缺陷检测的自动化、智能化提供了保障.
In order to identify and classify the hub on the production line,an on-line wheel hub type recognition system based on OpenCV and MFC platform is developed.The parameters of whell hub height,diameter,diameter of central hole,number of spokes and the spoke window’s ratio of perimeter to area are extracted as the characteristic parameters of whell hub recognition.The hub diameter is obtained by image preprocessing,edge detection,circle fitting,system calibration and other methods to represent the size of the hub.The hub shape is characterized by the number of spokes,the diameter of the central hole,the circumference area ratio of the spoke window and other constant with rotation invariance.Finally,the serial number of the extracted hub characteristic parameters is generated and compared with the standard values in the hub template library to achieve the online real-time classification effect.The experimental results show that the recognition accuracy of the system is 98.7%,which can effectively complete the online identification and classification of the hub and provides guarantee for the automation and intelligentization of the wheel defect detection.
作者
郭智杰
王明泉
张俊生
焦腾云
GUO Zhijie;WANG Mingquan;ZHANG Junsheng;JIAO Tengyun(School of Information and Communication Engineering,North University of China,Taiyuan 030051,China)
出处
《测试技术学报》
2019年第3期233-237,共5页
Journal of Test and Measurement Technology
基金
国家重大科学仪器设备开发专项基金资助项目(2013YQ240803)
山西省重点研发计划资助项目(201803D121069)
关键词
铝合金轮毂
机器视觉
相机标定
识别分类
aluminum alloy wheel hub
machine vision
camera calibration
identification classification