摘要
针对出厂钢轨表面缺陷检测的难题,提出了一种融合二维图像检测和结构光三维检测的钢轨表面缺陷检测方法;一台线扫描相机和一台面阵相机分别在LED线光源和结构光照明下采集钢轨表面图像,分别获取图像特征,并计算缺陷的参数,然后系统综合利用缺陷在可见光和结构光照明条件下的信息进行缺陷识别;实验结果表明,该方法能有效避开伪缺陷的干扰,并能提高缺陷几何形状的检测精度。
To solve the problem of detecting defects on rail surface during manufacturing process, a method fusing 2D image detection and 3D detection based on structured light was presented. Rails surfaces were scanned by an area scan camera and a line scan camera in the illumination of a LED light source and a structured light respectively, and then the features of images and the parameters of defects were extracted. Information of defects in visible light and structured light was used to recognize the defects. The experimental results indicated that the interference of pseudo--flaws could be avoided and the geometry precision of defects could be improved by using the method.
出处
《计算机测量与控制》
北大核心
2013年第5期1181-1182,1185,共3页
Computer Measurement &Control
基金
四川省2012年科技支撑计划
关键词
结构光
二维图像
融合
钢轨
表面缺陷
structured--light
image
fusion
rail
surface defects
