Share Email Print

Proceedings Paper

Planar segmentation and topological reconstruction for urban buildings with lidar point clouds
Author(s): Yunfan Li; Hongchao Ma; Jianwei Wu
Format Member Price Non-Member Price
PDF $14.40 $18.00
cover GOOD NEWS! Your organization subscribes to the SPIE Digital Library. You may be able to download this paper for free. Check Access

Paper Abstract

This paper presents a framework for segmentation and topological relationship reconstruction of building planar surfaces by using airborne LiDAR point clouds. The analysis of a planar surface structure is fundamental to almost any applications involving LiDAR data, especially building reconstruction. The proposed framework consists of two steps. Firstly, point clouds is segmented using an improved RANSAC (RANdom SAmple Consensus) algorithm with variant consensus set threshold. It is designed to solve under- or no- segmentation problem. It reduces consensus set threshold when the original RANSAC could not find valid planes, hence small planar surfaces would be extracted. Then, the topological relationship planar surface is constructed based on estimating connectedness of connecting point pairs between each pair of adjacent planar surfaces. The types of connectedness of planar surface are divided into three categories and a statistical method is used to estimate the connectedness type. The reconstructed topological relationship is described by an adjacent graph and can be utilized in the building modeling. Experiments show the effectiveness and efficiency of the proposed framework.

Paper Details

Date Published: 24 October 2011
PDF: 7 pages
Proc. SPIE 8286, International Symposium on Lidar and Radar Mapping 2011: Technologies and Applications, 828615 (24 October 2011); doi: 10.1117/12.912838
Show Author Affiliations
Yunfan Li, Wuhan Univ. (China)
Hongchao Ma, Wuhan Univ. (China)
Jianwei Wu, Wuhan Univ. (China)

Published in SPIE Proceedings Vol. 8286:
International Symposium on Lidar and Radar Mapping 2011: Technologies and Applications
Jonathan Li, Editor(s)

© SPIE. Terms of Use
Back to Top