期刊论文

Rail Wear Detection Algorithm Based on Line Laser-Machine Vision

中文

信息与电脑

1003-9767

2018

19

77-80

国家自然科学基金项目（项目编号:61272278,61401319）； 湖北省自然科学基金项目（项目编号:2015CFA061,2014CFB270,2018CFB408）； 湖北省重大科技专项（项目编号:2018ABA099）； 湖北省教育厅科学技术研究项目（项目编号:Q20181807）；

本校为第一机构

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...