摘要
针对钢球表面缺陷的在线检测要求,提出了一种基于机器视觉的钢球表面缺陷轨道自动检测方法。该方法采用6个CCD摄像头,保证摄取到整个钢球的表面;采用自行设计的碗状红色LED圆顶漫射光源,克服了钢球表面易反光的难题,得到了真实清晰的钢球表面图像;并通过图像缺陷识别、球形表面重构等技术,准确地识别出了钢球表面所存在的缺陷。
In view of the requirements for online ball surface defect detection,provide an automatic detection method of surface defects based on orbit.By using six CCD cameras,it can be sure to have image of the entire ball surface;and using self-designed red LED diffuse light source with bowl shape to overcome the problem of surface reflection,the real clear picture of the ball＇s surface is got.With the image defect identification,spherical surface reconstruction technique,the existence of the ball＇s surface defects is accurately identified.
出处
《轴承》
北大核心
2010年第5期45-48,共4页
Bearing
关键词
滚动轴承
钢球
缺陷检测
图像识别
rolling bearing
steel ball
defect detection image recognition
