针对装载机动臂孔组同轴度误差传统测量方法存在较大的人为误差且测量效率低等问题,设计基于机器视觉的同轴度误差在线检测系统.该系统使用Labview中的视觉模块设计系统检测界面并联合视觉算法库Hal-con完成对同轴度误差的在线检测;采用全局固定阈值算法进行图像分割,基于形态学的Canny算子边缘检测方法对图像进行形态学处理获得其边缘区域;基于XLD形状选择的最小二乘算法对图像边缘进行拟合圆处理,获得圆搭外圆的边缘特征信息及圆心坐标,进而通过分析计算得到该组孔的同轴度误差.实验表明:该系统检测方案满足工业生产的精度要求,且检测效率优于传统测量方法,适合于大批量的非接触式测量.
Aiming at the problems of large human error and low measurement efficiency in the traditional measure-ment method of coaxiality error of loader arm hole group,an online detection system of coaxiality error based on machine vi-sion is designed.The system uses the vision module in Labview to design the system detection interface and combines the visual algorithm library Halcon to complete the online detection of the coaxiality error.The global fixed threshold algorithm is used for image segmentation,and the morphological-based Canny operator edge detection method is designed to detect the image.Morphological processing is carried out to obtain its edge area,a least squares algorithm based on XLD shape selec-tion is proposed to fit the edge of the image to obtain the edge feature information and center coordinates of the circle and the outer circle,and the coaxiality error of the group of holes is obtained through analysis and calculation.Experiments show that the detection scheme of the system meets the precision requirements of industrial production,and the detection efficien-cy is better than that of traditional measurement methods.It is more suitable for mass non-contact measurement and can be used in actual production.