基于PCA法向量估计与RANSAC算法的隧道中轴线提取

Tunnel Axis Extraction Based on PCA Normal Vector Estimation with RANSAC Algorithm

为了减少隧道点云噪声对隧道中轴线提取精度的影响,提出主成分分析(PCA,Principal Components Analysis)算法点云法向量估计与随机抽样一致性(RANSAC,Random Sample Consensus)算法相结合的隧道中轴线提取方法。首先对隧道点云数据进行ROI(Region of Interest)区域提取及统计滤波预处理,其次通过PCA算法估计隧道点云法向量并进行点云法向量方向一致化处理,通过设定阈值分别提取满足要求的隧道拱顶及拱腰处法向量,最后用RANSAC算法代替最小二乘算法对筛选出来的法向量提取出精度较高的隧道中轴线。实验结果表明,本文方法所提取出来的隧道中轴线受噪声影响较小,斜率最大偏差0.007 1,最小偏差0.000 3,算法稳定性好,结果精度较高。

In order to reduce the influence of tunnel point cloud noise on the accuracy of tunnel central axis extraction,a tunnel central axis extraction method combined with Principal Components Analysis point cloud vector estimation and Random Sample Consensus algorithm is proposed.Firstly,the tunnel point cloud data is extracted by ROI region and statistical filter pretreatment,followed by the PCA algorithm to estimate the tunnel point cloud vector and perform point cloud vector direction consistency processing,and then the tunnel vault and arch normal vectors that meet the requirements are extracted by setting a threshold,and finally the RANSAC algorithm is used instead of the least squares algorithm to extract the tunnel axis with high accuracy from the filtered normal vectors.Experimental results show that the axis of the tunnel extracted by the proposed method is less affected by noise,the maximum deviation of slope is 0.007 1,and the minimum deviation is 0.000 3,the algorithm stability is good,and the accuracy of the results is high.