Lin Chen
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Lin Chen has a total of 4 co-authors. Here are the top co-authors by number of shared publications:
- Fengyu Zhang
Department of Disaster Mitigation for Structures, Tongji University, Shanghai 200092, China
1 shared publication - Jinyang Liu
Department of Disaster Mitigation for Structures, Tongji University, Shanghai 200092, China
1 shared publication - Peizhen Li
Department of Disaster Mitigation for Structures, Tongji University, Shanghai, People’s Republic of China
1 shared publication - Qingsong Xiong
Department of Civil and Environmental Engineering, The Hong Kong Polytechnic University, Hung Hom, Kowloon, Hong Kong
1 shared publication
Latest Publications
- Open Access
- 9 January 2026
- Published by MDPI AG in Electronics
- Vol. 15 (2) , 293
Automated Dimension Recognition and BIM Modeling of Frame Structures Based on 3D Point Clouds
Building information models (BIMs) serve as a foundational tool for digital management of existing structures. Traditional methods suffer from low automation and heavy reliance on manual intervention. This paper proposes an automated method for structural component dimension recognition and BIM modeling based on 3D point cloud data. The proposed methodology follows a three-step workflow. First, the raw point cloud is semantically segmented using the PointNet++ deep learning network, and individual structural components are effectively isolated using the Fast Euclidean Clustering (FEC) algorithm. Second, the principal axis of each component is determined through Principal Component Analysis, and the Random Sample Consensus (RANSAC) algorithm is applied to fit the boundary lines of the projected cross-sections, enabling the automated extraction of geometric dimensions. Finally, an automated script maps the extracted geometric parameters to standard IFC entities to generate the BIM model. The experimental results demonstrate that the average dimensional error for beams and columns is within 3 mm, with the exception of specific occluded components. This study realizes the efficient transformation from point cloud data to BIM models through an automated workflow, providing reliable technical support for the digital reconstruction of existing buildings.
- 9 January 2026
- Published by MDPI AG in Electronics
- Vol. 15 (2) , 293
Open AccessAutomated Dimension Recognition and BIM Modeling of Frame Structures Based on 3D Point Clouds
Building information models (BIMs) serve as a foundational tool for digital management of existing structures. Traditional methods suffer from low automation and heavy reliance on manual intervention. This paper proposes an automated method for structural component dimension recognition and BIM modeling based on 3D point cloud data. The proposed methodology follows a three-step workflow. First, the raw point cloud is semantically segmented using the PointNet++ deep learning network, and individual structural components are effectively isolated using the Fast Euclidean Clustering (FEC) algorithm. Second, the principal axis of each component is determined through Principal Component Analysis, and the Random Sample Consensus (RANSAC) algorithm is applied to fit the boundary lines of the projected cross-sections, enabling the automated extraction of geometric dimensions. Finally, an automated script maps the extracted geometric parameters to standard IFC entities to generate the BIM model. The experimental results demonstrate that the average dimensional error for beams and columns is within 3 mm, with the exception of specific occluded components. This study realizes the efficient transformation from point cloud data to BIM models through an automated workflow, providing reliable technical support for the digital reconstruction of existing buildings.