基于机器视觉的散热器钎焊缺陷检测方法研究

Research on Detection Method of Radiator Brazing Based on Machine Vision

摘要: 为了实现钎焊工业现场复杂环境下散热器缺陷的在线检测，文章提出了基于机器视觉的散热器钎焊缺陷检测方法。采用灰度化和高斯滤波对图像进行预处理，使用Canny边缘检测算子获取边缘图像；利用八邻域连通域分析方法查找所有边缘连通域，使用边界清除算法剔除边界处边缘轮廓的干扰，进而实现散热器缺陷检测。实验结果表明散热器缺陷检测正确率为95%，可以满足散热器工业现场对准确性和检测速度的要求。

Abstract: In order to realize the online detection of radiator defects in the complex environment of the brazing industry, this paper proposes a machine vision-based radiator brazing defect detection method. Grayscale and Gaussian filtering are used to preprocess the image, and the Canny edge detection operator is used to obtain the edge image; use the eight neighbors connected domain analysis method to find all the edge connected domains, and use the boundary clearing algorithm to remove the interference of the edge contours at the boundary, thereby achieving accurate iden-tification of defects. Experimental results show that the correct rate of radiator defect detection is 95%, which can meet the accuracy and speed requirements of the radiator industry site.