测绘科学技术学报2025，Vol.41Issue(6)：612-618,7.DOI:10.3969/j.issn.1673-6338.2025.06.009
联合虚拟格网和LPD-DBSCAN算法的建筑物立面点云提取方法
An Extraction Method of Building Façade Point Cloud by Combining Pseudo-grid and LPD-DBSCAN Algorithm
熊峰 1李宗春 1付永健 1汪文琪 1赵昭明 2何华1
作者信息
- 1. 信息工程大学,河南 郑州 450001
- 2. 信息工程大学,河南 郑州 450001||93920部队||陕西 西安 710001
- 折叠
摘要
Abstract
To address the problems that the building façade extraction algorithms in the light detection and ranging(LiDAR)point cloud data processing are difficult to set parameters reasonably and extract accurately,a method that combines pseudo-grid and density-based spatial clustering of applications with noise(DBSCAN)clustering al-gorithm considering the local point density for extracting building façade point cloud is proposed.Firstly,the ground point cloud is separated adaptively by using the elevation distribution characteristics based on pseudo-grid,and the low ground object point cloud is filtered out according to the altitude difference of the grid.Secondly,based on calculating the optimal neighborhood,the DBSCAN clustering algorithm is utilized according to the local point density of each point.Finally,the building façade point cloud is extracted precisely by combining the propor-tion and size of the linear and planar points in the clustering block.Experimental verification is performed on two datasets in different scenes,the evaluation results show that the proposed method can obtain better results of build-ing façade extraction,with higher accuracy compared to MRG,RealWorks and EPSB methods as well as F1 scores are higher than 97.69％.Reliable building façade information for the applications of building reconstruction and ur-ban fine-graining can be provided by this method.关键词
激光雷达/建筑物立面提取/虚拟格网/最优邻域/密度聚类Key words
LiDAR/building façade extraction/pseudo-grid/optimal neighborhood/density clustering分类
天文与地球科学引用本文复制引用
熊峰,李宗春,付永健,汪文琪,赵昭明,何华..联合虚拟格网和LPD-DBSCAN算法的建筑物立面点云提取方法[J].测绘科学技术学报,2025,41(6):612-618,7.
