半自动机载LiDAR点云建筑物三维重建方法

针对全自动建筑物3D重建存在需要后续人工检验,且发现重建错误需要花费额外时间修改的问题,提出了一种半自动的面向对象的机载LiDAR点云建筑物3D重建方法。基于建筑物类别点云的联通分析和平面生长分割结果,提出了自动的建筑物栋数检测、单栋建筑物外轮廓提取、单栋建筑物内部结构线提取方法;同时,在计算机无法完成部分工作时,人工辅助计算机完成高程阶越线提取、识别建筑物屋顶附属物点云等工作。实验证明,该方法可以适用于高密度机载LiDAR点云数据中城区大部分建筑物的3D模型重建。

3D reconstruction method of buildings of semi-automatic airborne LiDAR point clouds

To overcome the shortcomings ofautomatic 3D reconstruction ofbuildings,a semi-automatically object-based method for 3D reconstruction ofbuildings from airborne LiDAR point cloud is proposed.First,connected components analysis is employed to separate the building point clouds into individual buildings,and then plane growing method is performed to cluster each building point clouds.At last,the outlines and inner structural lines are extracted for each building based on the segmentation result.At the same time,manual intervention is involved when elevation jump lines are extracted or the identification of the attached objects on the building roofs.Thus,the above processing forms a semi-automatic method for 3D reconstruction of buildings.Experiments suggest that the proposed method is feasible for 3D reconstruction of most of the buildings from the high density point clouds.