基于机器视觉的圆形部件中心点定位系统研究

Research on the Location System for Circular Component Center Point Based on Machine Vision

为实现通过机器视觉对气门导管、闷盖孔、护套孔等零部件的定位,特设计一种适用于圆形部件中心点的检测方法。首先,使用阈值分割、膨胀、腐蚀等算法对图像进行预处理,筛选出圆形部件所在的区域。其次,找到圆形部件所在区域的中心,将此中心作为圆形卡尺工具的测量中心,设置卡尺的检测半径和动态测量范围,使测量范围覆盖圆形部件的外轮廓。利用卡尺测量算子找到圆形轮廓与卡尺栅格的交点集。最后,通过最小二乘圆弧逼近的方法得到圆的中心点,实现对圆形部件的定位。研究表明,该方法的鲁棒性较强,精确度较高,提高了装配的准确率。

A detection method suitable for circular component center point is designed to achieve the positioning of valve guides,blind cover holes,protective sleeve holes and other components through Machine Vision.Firstly,it uses threshold segmentation,dilation,corrosion and other algorithms to preprocess the image and filter out the region where circular components are located.Secondly,it finds the center of the area where the circular component is located,uses this center as the measurement center for the circular caliper tool,sets the detection radius and dynamic measurement range of the caliper,so that the measurement range covers the outer contour of the circular component.It uses the caliper measurement operator to find the intersection point set between the circular contour and the caliper grid.Finally,the center point of the circle is obtained using the method of least squares arc approximation to locate the circular component.Research has shown that this method has strong robustness,high accuracy,and improves the accuracy of assembly.