基于机器视觉的玻璃缺陷检测技术

Research on glass defect detection technology based on machine vision

缺陷检测是玻璃生产技术的重要环节,影响着产品质量和生产效率。设计了一种玻璃缺陷自动检测系统:采集玻璃图像信息并传输至计算机,对其进行了图像滤波,图像分割,轮廓提取,缺陷分类等运算,得到缺陷位置,大小等信息,实现玻璃缺陷的检测。针对图像处理中手动确定阈值的方法的不足,设计了自动阈值算法。实验表明,这种算法对不同成像质量照片的适应性较强,提高了玻璃缺陷检测的准确率。

Defect detection is an important link in glass production technology,which affects product quality and production efficiency. An automatic detection system for glass defects is designed. The image information of glass is collected and transmitted to the computer. The image filtering,image segmentation,contour extraction,defect classification and other operations are carried out to obtain the information of defect location,size and so on,thus realizing the detection of glass defects. Aiming at the shortage of manual threshold determination method in image processing,an automatic threshold algorithm is designed.Experiments show that the algorithm has strong adaptability to photos with different imaging quality and improves the accuracy of glass defect detection.