Journal of KIBIM (한국BIM학회 논문집)

Korean Institute of Building Information Modeling (한국BIM학회)
권호 표지
DOI QR코드

DOI QR Code

Development of Deep Learning-based Automatic Classification of Architectural Objects in Point Clouds for BIM Application in Renovating Aging Buildings

딥러닝 기반 노후 건축물 리모델링 시 BIM 적용을 위한 포인트 클라우드의 건축 객체 자동 분류 기술 개발

  • Received : 2023.11.02
  • Accepted : 2023.12.19
  • Published : 2023.12.31
Download PDF
⟨ Previous Next ⟩

Abstract

This study focuses on developing a building object recognition technology for efficient use in the remodeling of buildings constructed without drawings. In the era of the 4th industrial revolution, smart technologies are being developed. This research contributes to the architectural field by introducing a deep learning-based method for automatic object classification and recognition, utilizing point cloud data. We use a TD3D network with voxels, optimizing its performance through adjustments in voxel size and number of blocks. This technology enables the classification of building objects such as walls, floors, and roofs from 3D scanning data, labeling them in polygonal forms to minimize boundary ambiguities. However, challenges in object boundary classifications were observed. The model facilitates the automatic classification of non-building objects, thereby reducing manual effort in data matching processes. It also distinguishes between elements to be demolished or retained during remodeling. The study minimized data set loss space by labeling using the extremities of the x, y, and z coordinates. The research aims to enhance the efficiency of building object classification and improve the quality of architectural plans by reducing manpower and time during remodeling. The study aligns with its goal of developing an efficient classification technology. Future work can extend to creating classified objects using parametric tools with polygon-labeled datasets, offering meaningful numerical analysis for remodeling processes. Continued research in this direction is anticipated to significantly advance the efficiency of building remodeling techniques.

Keywords

Acknowledgement

본 연구는 국토교통부/국토교통과학기술진흥원의 지원으로 수행되었음(과제번호 RS-2021-KA163269). 또한 본 과제(결과물)는 교육부와 한국연구재단의 재원으로 지원을 받아 수행된 3단계 산학연협력 선도대학 육성사업(LINC 3.0)의 연구결과임.

References

  1. Kang, K., Wang, X. (2014). Fully Convolutional Neural Networks for Crowd Segmentation, arXiv:1411.4464, https:// arxiv.org/abs/1411.4464 (Nov. 18. 2023).
  2. Kwon, S. C., Kim, M. Y., Kim, T. Y. (2008). The Choice of Measurement Techniques in Survey and Recording of Historic Building in Mordern Ages of Korea, Autumn Annual Conference of AIK 2008, 28(1), pp. 431-434.
  3. Kwon, W., Chun, J. Y. (2006). Constitution of Work Process for the Remodeling Construction Project in Planning Phase, Korean Journal of Construction Engineering and Management, 7(6), pp. 165-174.
  4. Lim, H., Lee, Y., Jee, M., Go, M., Kim, H., Kim, Wo. (2019). Efficient Inference of Image Objects using Semantic Segmentation, Journal of Broadcast Engineering (JBE), 24(1), pp. 67-76.
  5. Lim, H., Lee, Y., Jee, M., Go, M., Kim, H., Kim, Wo. (2019). Efficient Inference of Image Objects Using Semantic Segmentation, Journal of Broadcast Engineering, 24(1), pp. 67-76. https://doi.org/10.5909/JBE.2019.24.1.67
  6. Long, J., Shelhamer, E., Darrell, T. (2015). Fully Convolutional Networks for Semantic Segmentation, CVPR 2015, pp7). 3431-3440.
  7. Long, J., Shelhamer, E., Darrell, T. (2015). Fully Convolutional networks for Semantic Segmentation, Proceedings of the IEEE conference on computer vision and pattern recognition, pp. 3431-3440.
  8. Min, J. S., Park, H. K, (2001). A Study on the Remodeling Process of Building, Autumn Annual Conference of AIK 2001, 21(2), pp. 127-130.
  9. Noh, H., Hong, S., Han, B. (2015) Learning Deconvolution Network for Semantic Segmentation, arXiv:1505.04366, https://arxiv.org/abs/1505.04366 (No. 22. 2023).
  10. Noh, H., Hong, S., Han, B. (2015) Learning Deconvolution Network for Semantic Segmentation, arXiv:1505.04366, https://arxiv.org/abs/1505.04366 (Nov. 22. 2023).
  11. Qi, C. R., Su, H., Mo, K., Guibas, L. J. (2017a). PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation, arXiv:1612.00593v2, https://arxiv.org/abs/1612.00593v2 (Dec. 20. 2023).
  12. Qi, C. R., Su, H., Mo, K., Guibas, L. J. (2017a). Pointnet: Deep Learning on Point Sets for 3D Classification and Segmentation, Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp. 652-660.
  13. Qi, C. R., Yi, L., Su, H., Guibas, L. J, (2017b). PointNet++: Deep Hierarchical Feature Learning on Point Sets in a Metric Space, Advances in Neural Information Processing Systems, 30.
  14. Rukhovich, D., Vorontsova, A., Konushin, A. (2021), FCAF3D: Fully Convolutional Anchor-Free 3D Object Detection, European Conference on Computer Vision, pp. 477-493.
  15. Yang, B., Luo, W., Urtasun, R. (2018). PIXOR: Real-time 3D Object Detection from Point Clouds, Proceedings of the IEEE conference on Computer Vision and Pattern Recognition, pp. 7652-7660.
  16. Zhou, Y., Tuzel, O. (2018). VoxelNet: End-to-End Learning for Point Cloud Based 3D Object Detection, Proceedings of the IEEE conference on computer vision and pattern recognition, pp.4490-4499.