基于机器视觉的自冲铆接接头剖面尺寸检测研究

Research on Section Size Detection of Self Piercing Riveting Joint Based on Machine Vision

针对实际工业检测中自冲铆接工艺接头的剖面尺寸人工检测存在精度低、效率低、测量不一致等问题,开发了一个基于机器视觉的自冲铆接工艺接头剖面尺寸检测系统,提出了一种基于边缘提取和角点检测的尺寸缺陷检测算法。对获取的图像进行图像灰度化、图像双边滤波和图像增强,通过Canny边缘检测进行图像分割实现测量区域的边缘提取,采用亚像素级角点检测提取剖面尺寸的测量点,采用最小二乘法拟合直线边缘。通过系统测试对尺寸缺陷检测算法进行了验证和分析,验证结果表明,算法的精确率达到98.67%,测量精度达到0.05 mm,平均检测节拍为0.31 s。该系统检测精度高、稳定性好、检测效率高,可以有效替代传统的人工测量。

Aiming at the problems of low precision,low efficiency and inconsistent measurement in the manual detection of the section size of self-piercing riveting process joints in actual industrial inspection,a machine vision-based self-piercing riveting process joint section size measurement system was developed,and a size defect detection algorithm based on edge extraction and corner detection is proposed.The obtained image is subjected to image grayscale,image bilateral filtering and image enhancement,and the edge of the measurement area is extracted by image segmentation through Canny edge detection,and the sub-pixel level corner detection is used to extract the measurement points of the section size.Least squares fit straight line edges.The dimensional defect detection algorithm is verified and analyzed through the system test.The verification results show that the accuracy of the algorithm reaches 98.67%,the measurement accuracy reaches 0.05 mm,and the average detection time is 0.31 s.The system has high detection accuracy,good stability and high detection efficiency,which can effectively replace the traditional manual measurement.