基于机器视觉的车速传感器焊点检测  

Welding spot detection of speed sensor based on machine vision

作　　者：谢俊[1] 李畅 左飞飞 李玉萍 杨启志[2] Xie Jun;Li Chang;Zuo Feifei;Li Yuping;Yang Qizhi(School of Mechanical Engineering,Jiangsu University,Zhenjiang 212013,Jiangsu,China;School of Agricultural Equipment Engineering,Jiangsu University,Zhenjiang 212013,Jiangsu,China)

机构地区：[1]江苏大学机械工程学院,镇江212013 [2]江苏大学农业装备工程学院,镇江212013

出　　处：《现代制造工程》2020年第6期52-56,共5页Modern Manufacturing Engineering

基　　金：江苏自然科学基金项目(BK20161344)。

摘　　要：应用机器视觉检测技术对车速传感器焊点进行图像处理与检测,通过CCD工业相机进行图像的采集,根据图像的特点划定感兴趣区域ROI;分析椒盐噪声干扰的影响,采用中值滤波算法对焊点图像进行去噪;采用OSTU算法推导出最佳阈值并用阈值法进行图像分割;运用开运算算法去除因阈值分割生成的细小区块;调用Open e Vision软件算法函数库计算焊点区域面积;调试检测系统程序并进行实际生产试验,结果表明:检测系统对车速传感器焊点检测精确且高效。Machine vision technology was applied to process image and detect the speed sensor welding spots,CCD industrial camera was used to collect welding spots images of the speed sensor and determine the Region Of Interest(ROI)according to the characteristics of the image.The influence of salt and pepper noise was analyzed,and the median filtering algorithm was used to remove noise in welding spots images.Using OSTU algorithm to deduce the best threshold value and using the threshold value to segment the image.Using open operation algorithm to eliminate the small blocks generated by threshold segmentation.Using Open e Vision software algorithm library to calculate the welding spots area.Debugging the program of the detection system,and carrying out the actual production test.The results show that the testing system is accurate and efficient for the welding spot detection of the speed sensor.

关 键 词：机器视觉 图像处理 车速传感器 焊点检测 

分 类 号：TN606[电子电信—电路与系统] TP391[自动化与计算机技术—计算机应用技术]