基于机器视觉的散热器钎焊缺陷检测系统研发
Research and Development of Radiator Brazing Defect detection System Based on Machine Vision

为解决散热器钎焊缺陷在工业检测过程中效率低、差错率高的问题，本文设计了一种基于机器视觉的缺陷检测系统。针对钎焊环节产生的焊缝和阻塞这两种缺陷，通过设计的内外双光源照射模块分别对其打光得到各自的原始图像；其次采用灰度处理、滤波除燥等算法进行预处理；最后采用区域生长算子和设计的双阈值筛选算法分别得到了图像的焊缝和阻塞缺陷。实验表明，本系统检测效率比传统人工检测效率提高了6倍，准确率在97%以上。
In order to solve the problem of low efficiency and high error rate in the industrial detection of radiator brazing defects, a defect detection system based on machine vision is designed in this paper. Aiming at the two defects of welding seam and blocking produced in the brazing link, the original images are obtained by lighting them respectively through the designed internal and external double light source irradiation module. Secondly, the algorithm of gray processing, filtering and removing dryness is used for preprocessing. Finally, the regional growth operator and the designed double threshold screening algorithm were used to obtain the welding seam and blocking defects of the image respectively. The experiment shows that the system greatly improves the detection efficiency of brazing defects, and the accuracy rate is more than 97%.