基于机器视觉的大型换热器管材孔系测量

Measurement of Pipe Hole System in Large Heat Exchanger Based on Machine Vision

针对大型换热器管材孔系,提出基于精密运动控制的机器视觉测量方案。运动控制系统携带相机完成工件全局扫描。机器视觉系统采集工件局部位置图像,使用高斯滤波法、最大类间方差法和Canny边缘检测方法自适应地检测孔系边缘,采用Hough梯度法提取单孔轮廓;利用优化的最小二乘拟合法测量单孔的孔心坐标和孔径,计算相邻孔的孔桥距。对测量方法进行实验验证,实验结果表明,孔径偏差低于0.031 mm,孔桥距偏差低于0.032 mm,满足测量精度要求。

A machine vision measurement scheme based on precision motion control is proposed for large heat exchanger tube hole system.The motion control system carries a camera to complete the global scan of the workpiece.The machine vision system collects the local position image of the workpiece,and adaptively detects the edge of the hole system using the Gaussian filter method,the OTSU method and the Canny edge detection method.The Hough gradient method is used to extract the single hole contour.The coordinates of the hole center and the aperture of the single hole are measured by the optimized least square fitting method,and the distance between the adjacent holes is calculated.The experimental results show that the aperture deviation is lower than 0.031 mm,and the hole bridge distance deviation is lower than 0.032 mm,which meet the measurement accuracy requirements.