基于机器视觉的玻璃马赛克缺陷在线检测系统

On-Line Glass Mosaic Defect Detection Based on Machine Vision

针对玻璃马赛克存在斑点、崩角等常见工艺缺陷,基于机器视觉检测原理,设计了一种采用LED同轴光照明和CCD相机为图像获取工具的在线检测系统。该系统通过相机实时拍摄,获取传送平台上目标玻璃马赛克图像,使用特定的视觉软件对获取到的图像进行中值滤波、锐化、二值化、边缘轮廓提取等图像处理,提取图像中玻璃马赛克轮廓形状与灰度值信息,通过形状匹配与灰度值匹配,实现对玻璃马赛克工艺缺陷的在线检测。实验证明,该在线检测系统具有准确率高、实时性强和速度快等优点。

Based on the principle of machine vision detection, an on-line detection system using LED coaxial light illumination and CCD camera as image acquisition tools is designed for glass mosaic with common defects such as spots and collapse angles. The system obtains the target glass mosaic image on the transmission platform by real-time camera, and uses the special vision software to process the images such as median filtering, sharpening, binarization and edge contour extraction, and extracting the image mosaic Contour shape and gray value information, through the shape matching and gray value matching, to achieve the glass mosaic process defects on-line detection. Experiments show that the on-line detection system has the advantages of high accuracy, high real-time performance and fast speed.