摘要
为了满足企业在实际生产中焊后微通道换热器缺陷自动化检测需求,应用机器视觉技术,设计了一种换热器翅片虚焊和翅片损伤的缺陷检测方法。首先利用双边滤波对样本图像进行滤波处理,随后采用y方向一阶Sobel边缘检测算子对扁管轮廓进行检测,并使用二值化和形态学开运算对干扰斑点进行过滤;其次针对处理后图像,对其进行直线检测并画出像素点横向累加投影折线图从而提取翅片损伤和翅片虚焊缺陷检测区域;最后根据缺陷特征同样运用直线检测技术对翅片虚焊缺陷区域进行检测,利用斑点轮廓面积检测对翅片损伤缺陷进行检测。实验表明,该检测算法对换热器翅片缺陷检测识别率可达97%,实现了快速、无接触且较为准确的检测。
In order to meet the requirements of automatic defect detection of microchannel heat exchanger after welding in actual production, a defect detection method of fin false welding and fin damage of heat exchanger was designed by using machine vision technology.Firstly, the sample image was filtered by bilateral filtering, then the first-order Sobel edge detection operator in the y direction was used to detect the contour of the flat tube, and the binarization and morphological operation were used to filter the interference spots.Secondly, for the processed image, the straight line was detected and the pixel point transverse cumulative projection broken line graph was drawn, so as to extract the detection area of fin damage and fin false welding defects.Finally, according to the defect characteristics, the line detection technology was also used to detect the fin false welding defect area, and the spot contour area detection was used to detect the fin damage defect.The experimental test shows that the detection algorithm has a recognition rate of 97% for the defect detection of heat exchanger fins, realizing fast, contactless and more accurate detection.
作者
杨晨
徐志鹏
曹松晓
宋涛
蒋庆
YANG Chen;XU Zhi-peng;CAO Song-xiao;SONG Tao;JIANG Qing(College of Metrology and Measurement Engineering,China Jiliang University,Hangzhou 310018,China)
出处
《仪表技术与传感器》
CSCD
北大核心
2022年第12期102-107,共6页
Instrument Technique and Sensor
基金
浙江省自然科学基金(LY18E050009)。
关键词
微通道换热器
机器视觉
缺陷检测
直线检测
microchannel heat exchanger
machine vision
defect detection
line detection
