TY - GEN
T1 - Building Point Cloud Segmentation via 2D–3D Fusion Based on Colmap
AU - Chuanchuan, Lu
AU - Guanghong, Gong
AU - Ying, Li
AU - Ni, Li
N1 - Publisher Copyright:
© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2026.
PY - 2026
Y1 - 2026
N2 - 3D point cloud segmentation is a key task in urban scene reconstruction, especially for extracting building structures, which are diverse in scale and geometry. Existing segmentation methods mainly rely on supervised deep learning, which suffers from limited generalization across different scenes and requires large amounts of annotated 3D data and computational resources. In contrast, 2D image segmentation has achieved significant progress. This work proposes a generalized 3D building segmentation framework based on 2D–3D fusion. By leveraging state-of-the-art 2D segmentation models such as Mask2Former and SAM, and combining them with 3D point clouds reconstructed by COLMAP, we establish correspondences between 2D masks and 3D points. This approach enables effective segmentation of 3D buildings without 3D supervision, and lays a foundation for downstream tasks such as urban scene reconstruction, measurement, and mapping.
AB - 3D point cloud segmentation is a key task in urban scene reconstruction, especially for extracting building structures, which are diverse in scale and geometry. Existing segmentation methods mainly rely on supervised deep learning, which suffers from limited generalization across different scenes and requires large amounts of annotated 3D data and computational resources. In contrast, 2D image segmentation has achieved significant progress. This work proposes a generalized 3D building segmentation framework based on 2D–3D fusion. By leveraging state-of-the-art 2D segmentation models such as Mask2Former and SAM, and combining them with 3D point clouds reconstructed by COLMAP, we establish correspondences between 2D masks and 3D points. This approach enables effective segmentation of 3D buildings without 3D supervision, and lays a foundation for downstream tasks such as urban scene reconstruction, measurement, and mapping.
KW - 2D-3D fusion
KW - building point cloud
KW - multi-view fusion
KW - point cloud segmentation
UR - https://www.scopus.com/pages/publications/105021937575
U2 - 10.1007/978-981-95-2751-9_19
DO - 10.1007/978-981-95-2751-9_19
M3 - 会议稿件
AN - SCOPUS:105021937575
SN - 9789819527502
T3 - Communications in Computer and Information Science
SP - 282
EP - 295
BT - Intelligent Simulation - 37th China Simulation Conference, CSC 2025, Proceedings
A2 - Liu, Yin
A2 - Li, Ni
A2 - Song, Xiao
A2 - Guo, Yinan
PB - Springer Science and Business Media Deutschland GmbH
T2 - 37th China Simulation Conference, CSC 2025
Y2 - 31 October 2025 through 2 November 2025
ER -