利用车载激光点云的分车带识别及单木分割方法

提出了一种基于车载激光点云数据的城区分车带识别及单木点云分割方法,首先通过布料模拟算法进行点云滤波去除地面点,然后利用基于八叉树连通性分析对非地面点进行聚类并构建聚类单元的最小包围矩形,基于先验知识和高差约束进行分车带识别,最后根据单木的空间几何特征,引入基于局部最高点的区域生长算法实现分车带内点云单木分割。选取北京市某道路的车载激光点云数据进行实验,结果表明:该方法能够从车载激光点云中快速识别出分车带点云并完成单木分割,能达到较好的识别和分割效果,具有抗噪性强和提取精度高的特点。

Single tree segmentation method of urban distributing belt based on vehicle-borne laser point cloud data

In this paper,a method based on the vehicle-borne laser point cloud is proposed for the identification of urban distributing belts and the segmentation of single tree inside distributing belts.Firstly,the point cloud is filtered by cloth simulation algorithm to remove the ground points,then the non-ground points are clustered by octree connectivity analysis,and the distributing belts are identified by using the minimum bounding rectangle and height difference constraints.Then,according to the spatial geometric characteristics of single tree,the region growing algorithm of the highest point is used to segment the point cloud in the distributing belt.The vehicle-borne laser point cloud data of a road in Beijing is selected for experiments.Results show that the proposed method can quickly identify the distributing belt point cloud from the vehicle laser point cloud and complete the single tree segmentation,which can achieve better recognition effect,with the characteristics of strong noise resistance and high extraction accuracy.