摘要
针对目前常规岔心测量方法作业效率低下、测量人员上线作业存在较大安全隐患等问题,提出一种基于三维激光扫描技术的既有铁路道岔岔心自动提取方法:①对原始点云数据进行预处理,通过轨面高程滤波剔除枕木等无用数据(只保留铁轨轨面数据);②利用Hough变换算法检测直线,并采用最小二乘法对检测出的直线进行拟合,以提高直线检测精度;③为每条直线设置1个邻域,搜索邻域内的点作为该条直线新的直线点集,采用最小二乘法对新筛选出的直线点集进行第二次拟合;④将直线参数相近的直线合并为一条直线,得到最终的轨道直线,根据提取出的直线参数计算岔心坐标。选取某车站既有道岔的点云数据进行实例验证,提取出的岔心坐标与传统人工测量方式得到的岔心坐标差值均在3 cm以内,证明了该方法的可行性。
Aiming at the problem that the current conventional surveying method of the railway turnout center is inefficient and too dangerous for surveying on the railway, this paper presented an automatic extraction method of railway turnout center based on 3D laser scanning technology:①Firstly this paper preprocessed the original point cloud data, and eliminated the useless data such as sleeper by filtering based on rail surface elevation (only rail surface data remained);②Then Hough transforming algorithm was used to detect the straight lines, and the least square fitting method was used to fit the detected straight line to improve the accuracy of line detection;③Then a neighborhood was set for every straight line, and the points in the neighborhood were regarded as the new points set of the straight lines, and fitted again by the least square fitting method;④The straight lines with similar line parameters were integrated into one line to get the final rail line, and the coordinates of railway turnout center were calculated according to the extracted line parameters. The point cloud data of railway turnouts from an existing railway station were selected to test this method. The difference between the coordinates of railway turnout center extracted by the new method and the ones got by the traditional manual surveying method was all within 3 cm, which proved the feasibility of this method.
作者
张云鹏
Zhang Yunpeng(China Railway Liuyuan Group Co. , Ltd. , Tianjin 300308, China)
出处
《铁道勘察》
2019年第5期31-36,共6页
Railway Investigation and Surveying
