基于线激光-机器视觉的钢轨磨耗检测算法

Rail Wear Detection Algorithm Based on Line Laser-Machine Vision

提出一种基于机器视觉的钢轨磨耗检测算法,给出激光迹线的细化与多段合成,摄像机标定与边缘配准的处理算法。实验结果表明,该方法能在40ms内完成一次钢轨轮廓检测,平均误差为0.06mm,能够较好的达到实际检测所要求的精度与实时性要求。

A detection algorithm for rail wear detection based on machine vision is proposed. The algorithm of laser trace refinement and multi-segment synthesis, camera calibration and edge registration was presented. The experimental results show that the method can complete the detection of a rail profile within 40ms with the average error is 0.06mm, which can meet the requirements of precision and real-time performance required by actual testing.