BIM-to-Labelled Point Cloud: Automated Point Cloud Annotation from BIM Models using Bounding Boxes and Solid Geometries
Abstract
This paper presents an automated framework for generating semantically labelled building point clouds from their corresponding BIM models. The proposed methodology aims to facilitate the creation of training datasets for deep learning-based indoor semantic segmentation. Two complementary labelling strategies are introduced. The first relies on bounding boxes (BBX) extracted from BIM elements to efficiently assign labels to points based on volumetric inclusion. The second approach uses solid geometry and a nearest-neighbour principle (SG-NN) to compute distances between BIM object meshes and the point cloud, enabling a more precise spatial correspondence. In addition, a room-based geometric grouping strategy is proposed to structure the annotated point clouds into spatial units compatible with common indoor segmentation datasets. The methods are evaluated through a qualitative analysis on several real building datasets of different typologies and acquisition conditions, as well as through a quantitative evaluation based on a manually segmented reference point cloud. Results show that the SG-NN approach achieves higher performance, with an average Recall of 92% and IoU of 88%, compared to 87% of Recall and %78 of IoU for the BBX approach. While the BBX approach provides faster processing, the SG-NN strategy achieves higher labelling accuracy, particularly for geometrically complex elements. The proposed workflow enables scalable dataset generation from Scan-to-BIM projects while significantly reducing manual annotation effort.
Related
- Point-cloud generative models for fast calorimeter simulation across particles and geometries
- The warp factor of supersymmetric D=11 near-horizon geometries: single-point rigidity, the Spin(7) perfect square, and global constraints
- Bounding the entanglement of a state from its spectrum
- A conjectured lower-bounding sequence for the Schur numbers, its reduction to S(6) and S(7), and the question of an upper-bounding sequence
- Universal Relations Meet Causality: Bounding the Dynamical Love Number of Neutron Stars
BibTeX
@article{boudarbala2026bimtolabelled,
title = {{BIM-to-Labelled Point Cloud: Automated Point Cloud Annotation from BIM Models using Bounding Boxes and Solid Geometries}},
author = {Saad Boudarbala and Tania Landes and Hélène Macher and Thibault Bavoux},
journal = {ISPRS International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences},
year = {2026},
month = jul,
doi = {10.5194/isprs-archives-xlix-b2-2026-117-2026},
url = {https://hal.science/hal-05746577v1},
note = {CC BY 4.0},
}
RIS
TY - JOUR
TI - BIM-to-Labelled Point Cloud: Automated Point Cloud Annotation from BIM Models using Bounding Boxes and Solid Geometries
AU - Saad Boudarbala
AU - Tania Landes
AU - Hélène Macher
AU - Thibault Bavoux
PY - 2026
DA - 2026/07/23
JO - ISPRS International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences
DO - 10.5194/isprs-archives-xlix-b2-2026-117-2026
UR - https://hal.science/hal-05746577v1
AB - This paper presents an automated framework for generating semantically labelled building point clouds from their corresponding BIM models. The proposed methodology aims to facilitate the creation of training datasets for deep learning-based indoor semantic segmentation. Two complementary labelling strategies are introduced. The first relies on bounding boxes (BBX) extracted from BIM elements to efficiently assign labels to points based on volumetric inclusion. The second approach uses solid geometry and a nearest-neighbour principle (SG-NN) to compute distances between BIM object meshes and the point cloud, enabling a more precise spatial correspondence. In addition, a room-based geometric grouping strategy is proposed to structure the annotated point clouds into spatial units compatible with common indoor segmentation datasets. The methods are evaluated through a qualitative analysis on several real building datasets of different typologies and acquisition conditions, as well as through a quantitative evaluation based on a manually segmented reference point cloud. Results show that the SG-NN approach achieves higher performance, with an average Recall of 92% and IoU of 88%, compared to 87% of Recall and %78 of IoU for the BBX approach. While the BBX approach provides faster processing, the SG-NN strategy achieves higher labelling accuracy, particularly for geometrically complex elements. The proposed workflow enables scalable dataset generation from Scan-to-BIM projects while significantly reducing manual annotation effort.
SN - 1682-1750
SN - 2194-9034
ER -