铁路罐车容积检定点云干扰点的自动去除算法

Automatic Removal Algorithm of Disturbance Points of Point Clouds for Railway Tankers Volume Calibration

运用三维激光扫描法检定铁路罐车容积是铁路罐车容积检定技术发展的趋势,然而目前在扫描形成点云中,由人工去除干扰点的做法严重影响后期数据的处理效率。提出自动去除干扰点的算法,该方法通过切片、投影将点云进行扁平化并进行极坐标系转换,在极坐标系中以确定微分区域内有效点云与干扰点云的特征差异作为判定依据,通过最大类间方差法计算阈值将其做出区分。运用Matlab R2012b平台对算法进行仿真,将G_(60k)型铁路罐车点云作为样本进行算法试验,有效去除了试验数据中的干扰点。

It is a technological development trend for railway tankers volume calibration to employ 3D laser scanning method. However, the disturbance points of point clouds formed during scanning removed manually has serious im- pact on the treatment efficiency of data at the later stage. This paper raises the algorithm that uses braking to remove disturbance points, which flattens the point clouds by slicing and projection and transformed polar coordinate sys- tem. In the polar coordinate system, the characteristic differences of effective point clouds and disturbed point clouds in the divergent region are confirmed, which are taken as the judgment basis and employ maximum between- cluster variance to calculate threshold value so as to distinguish it. Matlab R2012b platform is used to simulate the al- gorithm and the point clouds of G60k railway tankers are taken as the sample to conduct algorithm experiment, which removes the disturbance points in the test data effectively.