A 3D modeling method for buildings based on LiDAR point cloud and DLG

Existing 3d modeling methods for urban buildings suffer from the problem that data acquisition cost and model accuracy are difficult to balance. In this paper, we propose a method for 3D building modeling that is driven by knowledge rules using multi-source data. We summarize building structure know...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Computers, environment and urban systems environment and urban systems, 2023-06, Vol.102, p.101974, Article 101974
Hauptverfasser: Zhao, Qian, Zhou, Liangchen, Lv, Guonian
Format: Artikel
Sprache:eng
Schlagworte:
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:Existing 3d modeling methods for urban buildings suffer from the problem that data acquisition cost and model accuracy are difficult to balance. In this paper, we propose a method for 3D building modeling that is driven by knowledge rules using multi-source data. We summarize building structure knowledge rules by analyzing the building design standards, and these knowledge rules are applied to assist the process of extracting modeling information from topographic map DLG and airborne laser point cloud data and reconstructing single building 3D models. Our method achieves an efficient extraction of building models with a level of detail that can meet the needs of urban management. Based on an experiment using real data from the Muyan area of Nanjing, it is demonstrated that this method can balance the data acquisition cost and model accuracy within a certain range. This provides a low-cost and high-efficiency technical route for large-scale acquisition of 3D models of urban buildings. •3D modeling of buildings is a key issue for real-world 3D construction and city information modeling (CIM).•LoD-2.2 building models can meet the practical urban management needs.•We propose a method for 3D building modeling that is driven by knowledge rules using multi-source data.
ISSN:0198-9715
1873-7587
DOI:10.1016/j.compenvurbsys.2023.101974