摘要
提出了一种二维视觉和三维视觉相结合的钢轨表面缺陷检测方法。钢轨踏面分别在LED板光源和激光线光源的照射下,由两台面阵相机单独采集图像。扫描钢轨表面并把图像传送回主机,通过特征提取和图像处理,分别获取表面缺陷的大小和形状,以及缺陷的深度信息。实现了视频和激光检测的融合与互补。实验表明,与单独使用可见光图像识别,或者单独使用线激光检测相比,系统检测效果更好。
A new method for rail surface defects detection is presented, which combines the techniques of ma- chine vision and three-dimensional linear laser detection. Rail surface was scanned by two area scan cameras in the illumination of LED array and linear laser respectively, and then the images were transported to a host computer. The system recognized the rail surface defects in size, shape and depth by feature extraction and image processing. The experimental results show that the performance of the detection system is much better than either of the tradi- tional machine vision system and linear laser inspection system.
出处
《科学技术与工程》
北大核心
2012年第36期9877-9880,9884,共5页
Science Technology and Engineering
基金
四川省科技厅科技支撑计划项目资助
关键词
机器视觉
激光三角法
钢轨表面缺陷
图像处理
machine vision laser triangle method rail surface defects image processing
