Large-Scale Airborne LiDAR Point Cloud Building Extraction Based on Improved Voxelized Deep Learning Network
Buildings · 07.04.2026 · 16 (7) · 1450
DOI: 10.3390/buildings16071450 ↗Kurz gefasst
Fachartikel in der Zeitschrift Buildings, online erschienen am 07.04.2026. Verfasst von Bai Xue und 5 weiteren Autor:innen. Die Arbeit nennt 28 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Guizhou Provincial Basic Research Program, High-Level Innovative Talents in Guizhou Province, Bijie Scientist Workstation Project.
Abstract (Originalsprache)
High-precision 3D building data are pivotal for smart city development, urban planning, and disaster management. However, large-scale building extraction from airborne LiDAR point clouds remains challenging due to semantic ambiguity, uneven point density, and complex architectural structures. To address these limitations, we propose a novel framework integrating geometric topology perception with cross-dimensional attention mechanisms within a Sparse Voxel Convolutional Neural Network (SPVCNN). The key contributions include: (1) an enhanced LaserMix++ multi-scale hybrid augmentation strategy featuring cross-scene block replacement, ground normal–constrained rotation, and non-uniform scaling; (2) a dual-branch SPVCNN architecture embedding a collaborative module of Geometric Self-Attention (GSA) and Cross-Space Residual Attention (CSRA) to preserve topological consistency and enable cross-dimensional feature interaction; and (3) a Boundary Enhancement Module (BEM) specifically designed to resolve boundary ambiguity and overlapping predictions. Evaluated on a 177 km2 dataset covering Washington, D.C., our method significantly outperforms the baseline SPVCNN, improving accuracy by 12.04 percentage points (0.8212 to 0.9416) and Intersection over Union (IoU) by 9.96 percentage points (0.866 to 0.9656). Furthermore, it surpasses mainstream networks such as Cylinder3D and MinkResNet by over 50% in absolute accuracy gain. These results demonstrate the effectiveness of synergistically combining geometric perception with adaptive attention for robust building extraction from large-scale LiDAR data.
Förderung
- Guizhou Provincial Basic Research Program
- High-Level Innovative Talents in Guizhou Province
- Bijie Scientist Workstation Project
- Karst Plateau Resources and Environment Remote Sensing Talent Team
- Intelligent Geospatial Information Application Engineering Center
Zitieren
APA
Xue, B., Song, Y., Ai, P., Li, H., Liu, S., & Guo, L. (2026). Large-Scale Airborne LiDAR Point Cloud Building Extraction Based on Improved Voxelized Deep Learning Network. Buildings, 16(7), 1450. https://doi.org/10.3390/buildings16071450
Vancouver
Xue B, Song Y, Ai P, Li H, Liu S, Guo L. Large-Scale Airborne LiDAR Point Cloud Building Extraction Based on Improved Voxelized Deep Learning Network. Buildings. 2026;16(7):1450. doi:10.3390/buildings16071450
BibTeX
@article{Xue2026,
author = {Xue, Bai and Song, Yanru and Ai, Pi and Li, Hongzhou and Liu, Shuhan and Guo, Li},
title = {Large-Scale Airborne LiDAR Point Cloud Building Extraction Based on Improved Voxelized Deep Learning Network},
journal = {Buildings},
year = {2026},
volume = {16},
number = {7},
pages = {1450},
doi = {10.3390/buildings16071450}
}
EUVIMED