Author(s): Long Huu Nguyen and Ha Le Thi Thu

Abstract: High-resolution three-dimensional (3D) reconstruction of urban buildings plays a crucial role in infrastructure management, urban planning, cultural heritage preservation, and the development of smart cities. Unmanned Aerial Vehicle (UAV)-based photogrammetry and Terrestrial Laser Scanning (TLS) are two of the most representative geospatial technologies, both of which have made significant contributions to geospatial data acquisition. However, each data acquisition technique has inherent limitations in terms of observation coverage and the ability to capture information in occluded areas, resulting in 3D models generated from a single data source that often fail to represent the complete geometric characteristics of a building. This study focuses on the integration of UAV- and TLS-derived point clouds to develop an effective approach for high-resolution 3D reconstruction of urban buildings. The proposed approach is validated through a case study conducted on an urban building located in Quang Ninh Province, Vietnam. The proposed workflow consists of the following steps: (i). data acquisition using an Unmanned Aerial Vehicle (UAV) and a Terrestrial Laser Scanning (TLS) system; (ii). Point cloud preprocessing, including noise filtering, point cloud decimation, and selection of overlapping regions; (iii). Registration of the two point cloud datasets using the Iterative Closest Point (ICP) algorithm implemented in Cloud Compare; (iv). Point cloud fusion; and (v). High-resolution 3D reconstruction of the building. Experimental results demonstrate that the integration of UAV and TLS point clouds effectively compensates for missing data over roof surfaces, building façades, and occluded areas. Compared with models generated from a single data source, the integrated approach significantly improves both the geometric completeness and the level of detail of the reconstructed 3D model. The experimental results demonstrate that the geometric accuracy of the integrated UAV-TLS point cloud is location-dependent. Specifically, millimeter-level accuracy is achieved for objects located at the base of the building, whereas objects in the roof area are reconstructed with centimeter-level accuracy. The experimental results demonstrate the feasibility and effectiveness of integrating UAV and TLS point cloud data for high-resolution 3D reconstruction of urban buildings. Furthermore, the proposed approach provides a scientific and technical foundation for the implementation of urban spatial digitalization applications in Vietnam.

DOI: https://www.doi.org/10.22271/2707840X.2026.v7.i2a.66

Pages: 01-10 | Views: 2 | Downloads: 2

Download Full Article: Click Here