基于三维激光成像技术的扣件缺损自动检测

Automatic detection of fastener defects based on 3D laser imaging technology

随着我国轨道交通的快速发展,其维护工作日益繁重,传统检测手段及技术难以满足现有的需求,轨道检测的快速化及智能化迫在眉睫。为了弥补这些检测方法的不足,本文基于非接触式激光三角测距原理,自主研发了一套高速线激光测量系统。首先是通过改进的灰度权重模型法提取激光中心线,可以得到更加精确的轨道扣件高度信息;其次使用Phone光照模型对点云数据渲染,可以直观显示采集对象的高低起伏情况;最后利用K均值聚类法进行扣件病害类别识别。检测结果为:扣件缺失与扣件异物检测成功率分别为98%、100%。尤其使用改进的灰度权重模型法,精确的提取激光中心线,得到高精度的三维纹理信息,具有很高的推广应用价值。

With the rapid development of rail transit in China,its maintenance is becoming increasingly heavy.The traditional detection methods and technologies are difficult to meet the existing needs.The rapid and intelligent rail detection is imminent.In order to make up for the shortcomings of these detection methods,a high-speed line laser measurement system is developed based on the principle of non-contact laser triangulation.More accurate track fastener height information of the orbital fastener can be obtained with an improved gray weight model method to extract the laser centerline.Secondly,the phone illumination model is used to render the point cloud data,which can visually display the ups and downs of the collected object.Finally,the K-means clustering method is used to identify the category of fastener diseases.The test results show that the success rates of fastener missing and fastener foreign matter detection were 98%and 100%,respectively.In particular,the improved gray weight model method is used to accurately extract the laser centerline and obtain high-precision 3D texture information,which has high popularization and application value.