摘要
针对控制器生产过程中异形元器件插装存在劳动强度大、自动化程度低等问题,提出一种高精度的视觉定位与检测算法。通过图像形状特征及提取要求,引入高曝光图像引脚粗精定位融合算法,实现引脚的高精度定位;通过对元器件针脚与电路板针孔位置进行拟合与插装判断,解决插装准确率低及元器件插装检测问题;通过实验室机器人进行重复定位精度与4000次重复性测试统计,定位精度对视觉像素坐标与实际坐标进行转换得来,插件定位精度<±0.1 mm,准确率达到99.7%以上。
Aiming at the problems of high labor intensity and low automation in the assembly of special-shaped components in controller production,a high-precision visual positioning and detection algorithm was proposed.Based on the image shape features and extraction requirements,the coarse and fine positioning fusion algorithm of pins in high exposure images was introduced to achieve high precision positioning of pins.By fitting and judging the positions of the components’pins and the circuit board’s pinholes,the problems of low insertion accuracy and component insertion detection were solved.Repeated positioning accuracy and 4000 repeatability test statistics were carried out by the laboratory robot.The positioning accuracy was obtained by converting the visual pixel coordinates to the actual coordinates.The plug-in positioning accuracy is<±0.1 mm,and the accuracy reaches more than 99.7%.
作者
刘志昌
LIU Zhi-chang(School of Computer Science and Technology,Beijing Institute of Technology,Beijing 100081,China;Equipment Power Technology Research Institute,Zhuhai Gree Electric Appliances Limited Company,Zhuhai 519070,China)
出处
《计算机工程与设计》
北大核心
2023年第9期2619-2625,共7页
Computer Engineering and Design
关键词
异形插件
视觉定位
缺陷检测
灰度聚类
像素点拟合
坐标转换
圆度拟合
special-shaped plug-in
visual orientation
defect detection
gray scale clustering
pixel fitting
coordinate transformation
roundness fitting
