A platform for research: civil engineering, architecture and urbanism
A platform for research: civil engineering, architecture and urbanism
BIM-driven data augmentation method for semantic segmentation in superpoint-based deep learning network
Abstract This paper describes a universal workflow to synthesize point clouds containing both geometry and color information by utilizing the IFC model or its 3D geometry model to automatically generate annotated point clouds for semantic segmentation in deep learning. In our experiments, we selected 44 scenes from the S3DIS dataset to rebuild BIM models and synthesize point clouds, replaced training datasets in different proportions with synthetic point clouds, and fed the mixed datasets to a superpoint-based network. The training results show that increasing the appropriate proportion of synthetic point clouds in the training dataset can significantly improve the semantic segmentation performance. Furthermore, we extend our experiments to two other state-of-the-art networks and another similar dataset, ScanNet, and the experimental results illustrate that the synthetic point clouds can be applied to augmenting the limited and low-quality annotated point clouds for deep learning and boost the automatic remodeling work in the architectural field.
Highlights A methodology to generate high-quality synthetic point clouds from BIM models. Automatic annotation for synthetic point clouds using Industry Foundation Class. A data augmentation strategy for deep learning on scanning point clouds. Color-based synthetic point clouds are effective for semantic segmentation.
BIM-driven data augmentation method for semantic segmentation in superpoint-based deep learning network
Abstract This paper describes a universal workflow to synthesize point clouds containing both geometry and color information by utilizing the IFC model or its 3D geometry model to automatically generate annotated point clouds for semantic segmentation in deep learning. In our experiments, we selected 44 scenes from the S3DIS dataset to rebuild BIM models and synthesize point clouds, replaced training datasets in different proportions with synthetic point clouds, and fed the mixed datasets to a superpoint-based network. The training results show that increasing the appropriate proportion of synthetic point clouds in the training dataset can significantly improve the semantic segmentation performance. Furthermore, we extend our experiments to two other state-of-the-art networks and another similar dataset, ScanNet, and the experimental results illustrate that the synthetic point clouds can be applied to augmenting the limited and low-quality annotated point clouds for deep learning and boost the automatic remodeling work in the architectural field.
Highlights A methodology to generate high-quality synthetic point clouds from BIM models. Automatic annotation for synthetic point clouds using Industry Foundation Class. A data augmentation strategy for deep learning on scanning point clouds. Color-based synthetic point clouds are effective for semantic segmentation.
BIM-driven data augmentation method for semantic segmentation in superpoint-based deep learning network
Zhai, Ruoming (author) / Zou, Jingui (author) / He, Yifeng (author) / Meng, Liyuan (author)
2022-05-19
Article (Journal)
Electronic Resource
English