Premium
Robust and fast reconstruction of complex roofs with active sampling from 3D point clouds
Author(s) -
Dehbi Youness,
Henn André,
Gröger Gerhard,
Stroh Viktor,
Plümer Lutz
Publication year - 2021
Publication title -
transactions in gis
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.721
H-Index - 63
eISSN - 1467-9671
pISSN - 1361-1682
DOI - 10.1111/tgis.12659
Subject(s) - ransac , computer science , point cloud , sampling (signal processing) , preprocessor , consistency (knowledge bases) , lidar , artificial intelligence , cascade , ranking (information retrieval) , azimuth , nonparametric statistics , data mining , pattern recognition (psychology) , algorithm , computer vision , mathematics , filter (signal processing) , remote sensing , geography , statistics , engineering , geometry , chemical engineering , image (mathematics)
This article proposes a novel method for the 3D reconstruction of LoD2 buildings from LiDAR data. We propose an active sampling strategy which applies a cascade of filters focusing on promising samples at an early stage, thus avoiding the pitfalls of RANSAC‐based approaches. Filters are based on prior knowledge represented by (nonparametric) density distributions. In our approach samples are pairs of surflets—3D points together with normal vectors derived from a plane approximation of their neighborhood. Surflet pairs provide parameters for model candidates such as azimuth, inclination and ridge height, as well as parameters estimating internal precision and consistency. This provides a ranking of roof model candidates and leads to a small number of promising hypotheses. Building footprints are derived in a preprocessing step using machine learning methods, in particular support vector machines.