摘要
为实现钢轨表面缺陷的实时全方位检测,提出基于多摄像机线结构光的三维检测方法。首先,通过标定线结构光和摄像机平面之间的位置参数,获取该摄像机视角下的钢轨表面线结构光的三维形状变化。其次,标定多个摄像机之间旋转平移参数,将多个视角下的三维点云进行拼接形成完整钢轨三维模型。同时,设计了钢轨三维检测的硬件装置,通过硬件编码器获取传送带运动的距离,形成钢轨在长度方向上的全貌三维形状。针对存在缺陷的钢轨样品进行了三维轮廓测量,实验结果表明,该方法的在钢轨深度方向上的精度能达较高精度,符合钢轨表面缺陷检测的要求。
In order to realize the real-time detection of rail surface defects, a 3D detection method based on multiple camera line structured light is proposed in this paper. Firstly, the three-dimensional shape change of the rail surface line structure light is obtained by calibrating the position parameters between the structured light and the plane of the camera. Secondly, the rotation and translation parameters of multiple cameras are calibrated, and the 3D shape of the line structured light is formed into a complete rail model. At the same time, the hardware device of rail 3D detection was designed to obtain the distance of the conveyor belt movement through hardware encoder, and form the three-dimensional shape of the rail in the length direction. For 3D profile measurement of defective rail samples, experimental results show that the accuracy of the method in the depth direction of the rail can meets the requirements of the rail surface defect detection.
出处
《机械设计与制造》
北大核心
2017年第3期170-172,176,共4页
Machinery Design & Manufacture
基金
四川省重点实验室开放基金(13zxtk05)
西南科技大学校创新团队建设基金(14tdtk01)
关键词
钢轨表面缺陷
多摄像机
线结构光
相机标定
三维重建
Rail Surface Defect
Multi-Camera
Line Structured Light
Camera Calibration
3D Reconstruction