EKNet: Graph Structure Feature Extraction and Registration for Collaborative 3D Reconstruction in Architectural Scenes
Source abstract
Abstract
Collaborative geometric reconstruction of building structures can significantly reduce communication consumption for data sharing, protect privacy, and provide support for large-scale robot application management. In recent years, geometric reconstruction of building structures has been partially studied, but there is a lack of alignment fusion studies for multi-UAV (Unmanned Aerial Vehicle)-reconstructed geometric structure models. The vertices and edges of geometric structure models are sparse, and existing methods face challenges such as low feature extraction efficiency and substantial data requirements when processing sparse graph structures after geometrization. To address these challenges, this paper proposes an efficient deep graph matching registration framework that effectively integrates interpretable feature extraction with network training. Specifically, we first extract multidimensional local properties of nodes by combining geometric features with complex network features. Next, we construct a lightweight graph neural network, named EKNet, to enhance feature representation capabilities, enabling improved performance in low-overlap registration scenarios. Finally, through feature matching and discrimination modules, we effectively eliminate incorrect pairings and enhance accuracy. Experiments demonstrate that the proposed method achieves a 27.28% improvement in registration speed compared to traditional GCN (Graph Convolutional Neural Networks) and an 80.66% increase in registration accuracy over the suboptimal method. The method exhibits strong robustness in registration for scenes with high noise and low overlap rates. Additionally, we construct a standardized geometric point cloud registration dataset.
Source: crossref · Quality: baseline
Research context
OpenAlex topics: 3D Surveying and Cultural Heritage · 3D Shape Modeling and Analysis · Remote Sensing and LiDAR Applications
SciVX discovery tags are terminology-based. OpenAlex topics are source-provided classifications.
Affiliations
College of Intelligence Science and Technology, National University of Defense Technology, Changsha 410073, China
National Key Laboratory of Equipment State Sensing and Smart Support, National University of Defense Technology, Changsha 410073, China
References (55)
Show source-provided references
1. Rigid pairwise 3D point cloud registration: A survey10.1016/j.patcog.2024.110408 ↗
2. A review of point cloud registration algorithms for mobile robotics10.1561/2300000035 ↗
3. Surmann, H., Slomma, D., Grobelny, S., and Grafe, R. (2021, January 25–27). Deployment of Aerial Robots after a major fire of an industrial hall with hazardous substances, a report. Proceedings of the 2021 IEEE International Symposium on Safety, Security, and Rescue Robotics (SSRR), New York, NY, USA.10.1109/SSRR53300.2021.9597677 ↗
4. Paolanti, M., Pierdicca, R., Martini, M., Di Stefano, F., Morbidoni, C., Mancini, A., Malinverni, E.S., Frontoni, E., and Zingaretti, P. (2019, January 9–10). Semantic 3D object maps for everyday robotic retail inspection. Proceedings of the New Trends in Image Analysis and Processing—ICIAP 2019: ICIAP International Workshops, BioFor, PatReCH, e-BADLE, DeepRetail, and Industrial Session, Trento, Italy. Revised Selected Papers 20.10.1007/978-3-030-30754-7_27 ↗
5. Peng, F., Wu, Q., Fan, L., Zhang, J., You, Y., Lu, J., and Yang, J.Y. (2014, January 27–30). Street view cross-sourced point cloud matching and registration. Proceedings of the 2014 IEEE International Conference on Image Processing (ICIP), Paris, France.10.1109/ICIP.2014.7025406 ↗
6. A systematic approach for cross-source point cloud registration by preserving macro and micro structures10.1109/TIP.2017.2695888 ↗
7. Bai, X., Luo, Z., Zhou, L., Fu, H., Quan, L., and Tai, C.L. (2020, January 13–19). D3feat: Joint learning of dense detection and description of 3D local features. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, Seattle, WA, USA.10.1109/CVPR42600.2020.00639 ↗
8. Wang, H., Liu, Y., Dong, Z., Guo, Y., Liu, Y.S., Wang, W., and Yang, B. (2023, January 17–24). Robust multiview point cloud registration with reliable pose graph initialization and history reweighting. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, Vancouver, BC, Canada.10.1109/CVPR52729.2023.00917 ↗
9. Choy, C., Park, J., and Koltun, V. (November, January 27). Fully convolutional geometric features. Proceedings of the IEEE/CVF International Conference on Computer Vision, Seoul, Republic of Korea.
10. Huang, S., Gojcic, Z., Usvyatsov, M., Wieser, A., and Schindler, K. (2021, January 20–25). Registration of 3D Point Clouds with Low Overlap. Proceedings of the Computer Vision and Pattern Recognition, Nashville, TN, USA.10.1109/CVPR46437.2021.00425 ↗
11. Qin, Z., Yu, H., Wang, C., Peng, Y., and Xu, K. (2023, January 17–24). Deep graph-based spatial consistency for robust non-rigid point cloud registration. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, Vancouver, BC, Canada.10.1109/CVPR52729.2023.00522 ↗
12. Yu, Z., Qin, Z., Zheng, L., and Xu, K. (2024, January 17–21). Learning Instance-Aware Correspondences for Robust Multi-Instance Point Cloud Registration in Cluttered Scenes. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, Seattle, WA, USA.10.1109/CVPR52733.2024.01854 ↗
13. Challenging data sets for point cloud registration algorithms10.1177/0278364912458814 ↗
14. Koide, K., Yokozuka, M., Oishi, S., and Banno, A. (June, January 30). Voxelized GICP for fast and accurate 3D point cloud registration. Proceedings of the 2021 IEEE International Conference on Robotics and Automation (ICRA), Xi’an, China.10.1109/ICRA48506.2021.9560835 ↗
15. Rusu, R.B., Blodow, N., and Beetz, M. (2009, January 12–17). Fast point feature histograms (FPFH) for 3D registration. Proceedings of the 2009 IEEE International Conference on Robotics and Automation, Kobe, Japan.10.1109/ROBOT.2009.5152473 ↗
16. Gojcic, Z., Zhou, C., Wegner, J.D., and Wieser, A. (2019, January 15–20). The perfect match: 3D point cloud matching with smoothed densities. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, Long Beach, CA, USA.10.1109/CVPR.2019.00569 ↗
17. Point cloud registration based on one-point ransac and scale-annealing biweight estimation10.1109/TGRS.2020.3045456 ↗
18. Fast and robust iterative closest point
19. Adiyatov, O., and Varol, H.A. (2017, January 6–9). A novel RRT*-based algorithm for motion planning in dynamic environments. Proceedings of the 2017 IEEE International Conference on Mechatronics and Automation (ICMA), Takamatsu, Japan.10.1109/ICMA.2017.8016024 ↗
20. Shiarlis, K., Messias, J., and Whiteson, S. (June, January 29). Rapidly exploring learning trees. Proceedings of the 2017 IEEE International Conference on Robotics and Automation (ICRA), Singapore.10.1109/ICRA.2017.7989184 ↗
21. Wang, X., Zhu, M., Bo, D., Cui, P., Shi, C., and Pei, J. (2020, January 6–10). Am-gcn: Adaptive multi-channel graph convolutional networks. Proceedings of the 26th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining, Virtual.10.1145/3394486.3403177 ↗
22. Feng, A., You, C., Wang, S., and Tassiulas, L. (March, January 22). Kergnns: Interpretable graph neural networks with graph kernels. Proceedings of the AAAI Conference on Artificial Intelligence, Virtual.
23. Perozzi, B., Al-Rfou, R., and Skiena, S. (2014, January 24–27). Deepwalk: Online learning of social representations. Proceedings of the 20th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, New York, NY, USA.10.1145/2623330.2623732 ↗
24. Grover, A., and Leskovec, J. (2016, January 13–17). node2vec: Scalable feature learning for networks. Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, San Francisco, CA, USA.10.1145/2939672.2939754 ↗
25. Semi-supervised graph regularized deep NMF with bi-orthogonal constraints for data representation10.1109/TNNLS.2019.2939637 ↗
26. Besl, P.J., and McKay, N.D. (1991, January 14–15). Method for registration of 3-D shapes. Proceedings of the Sensor Fusion IV: Control Paradigms and Data Structures, Boston, MA, USA.
27. Greenspan, M., and Yurick, M. (2003, January 6–10). Approximate kd tree search for efficient ICP. Proceedings of the Fourth International Conference on 3-D Digital Imaging and Modeling, 2003 (3DIM 2003), Proceedings, Banff, AB, Canada.
28. Min, Z., Wang, J., and Meng, M.Q.H. (2018, January 21–25). Robust generalized point cloud registration using hybrid mixture model. Proceedings of the 2018 IEEE International Conference on Robotics and Automation (ICRA), Brisbane, QLD, Australia.10.1109/ICRA.2018.8460825 ↗
29. Local feature extraction and matching on range images: 2.5 D SIFT10.1016/j.cviu.2009.06.005 ↗
30. Maximum likelihood from incomplete data via the EM algorithm10.1111/j.2517-6161.1977.tb01600.x ↗
31. Harris 3D: A robust extension of the Harris operator for interest point detection on 3D meshes10.1007/s00371-011-0610-y ↗
32. Zhang, J., and Singh, S. (2015, January 26–30). Visual-lidar odometry and mapping: Low-drift, robust, and fast. Proceedings of the 2015 IEEE International Conference on Robotics and Automation (ICRA), Seattle, WA, USA.10.1109/ICRA.2015.7139486 ↗
33. Prakhya, S.M., Liu, B., and Lin, W. (October, January 28). B-SHOT: A binary feature descriptor for fast and efficient keypoint matching on 3D point clouds. Proceedings of the 2015 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), Hamburg, Germany.10.1109/IROS.2015.7353630 ↗
34. Rusu, R.B., Blodow, N., Marton, Z.C., and Beetz, M. (2008, January 22–26). Aligning point cloud views using persistent feature histograms. Proceedings of the 2008 IEEE/RSJ International Conference on Intelligent Robots and Systems, Nice, France.10.1109/IROS.2008.4650967 ↗
35. Chen, C., Li, G., Xu, R., Chen, T., Wang, M., and Lin, L. (2019, January 15–20). Clusternet: Deep hierarchical cluster network with rigorously rotation-invariant representation for point cloud analysis. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, Long Beach, CA, USA.10.1109/CVPR.2019.00513 ↗
36. Aoki, Y., Goforth, H., Srivatsan, R.A., and Lucey, S. (2019, January 15–20). Pointnetlk: Robust & efficient point cloud registration using pointnet. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, Long Beach, CA, USA.10.1109/CVPR.2019.00733 ↗
37. Qi, C.R., Su, H., Mo, K., and Guibas, L.J. (2017, January 21–26). Pointnet: Deep learning on point sets for 3D classification and segmentation. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, Honolulu, HI, USA.
38. Wang, Y., and Solomon, J.M. (November, January 27). Deep closest point: Learning representations for point cloud registration. Proceedings of the IEEE/CVF International Conference on Computer Vision, Seoul, Republic of Korea.
39. Attention is all you need
40. Li, J., Zhang, C., Xu, Z., Zhou, H., and Zhang, C. (2020, January 23–28). Iterative distance-aware similarity matrix convolution with mutual-supervised point elimination for efficient point cloud registration. Proceedings of the Computer Vision—ECCV 2020: 16th European Conference, Glasgow, UK. Proceedings, Part XXIV 16.10.1007/978-3-030-58586-0_23 ↗
41. Prnet: Self-supervised learning for partial-to-partial registration
42. Yuan, W., Eckart, B., Kim, K., Jampani, V., Fox, D., and Kautz, J. (2020, January 23–28). Deepgmr: Learning latent gaussian mixture models for registration. Proceedings of the Computer Vision—ECCV 2020: 16th European Conference, Glasgow, UK. Proceedings, Part V 16.10.1007/978-3-030-58558-7_43 ↗
43. Yew, Z.J., and Lee, G.H. (2020, January 13–19). Rpm-net: Robust point matching using learned features. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, Seattle, WA, USA.10.1109/CVPR42600.2020.01184 ↗
44. Graph attention networks
45. Xu, B., Shen, H., Cao, Q., Qiu, Y., and Cheng, X. (2019). Graph wavelet neural network. arXiv.
46. Wijesinghe, A., and Wang, Q. (2022, January 25–29). A new perspective on “How graph neural networks go beyond Weisfeiler-Lehman?”. Proceedings of the International Conference on Learning Representations, Virtual.
47. Feature selection and learning for graphlet kernel10.1016/j.patrec.2020.05.023 ↗
48. Dgcnn: A convolutional neural network over large-scale labeled graphs10.1016/j.neunet.2018.09.001 ↗
49. Clevert, D.A., Unterthiner, T., and Hochreiter, S. (2015). Fast and accurate deep network learning by exponential linear units (elus). arXiv.
50. Zhang, F., and Zhang, X. (2024). GraphKAN: Enhancing Feature Extraction with Graph Kolmogorov Arnold Networks. arXiv.
51. Liu, Z., Wang, Y., Vaidya, S., Ruehle, F., Halverson, J., Soljačić, M., Hou, T.Y., and Tegmark, M. (2024). Kan: Kolmogorov-arnold networks. arXiv.
52. Zhou, Y., Huo, H., Hou, Z., Bu, L., Mao, J., Wang, Y., Lv, X., and Bu, F. (2023). Co-embedding of edges and nodes with deep graph convolutional neural networks. Sci. Rep., 13.10.1038/s41598-023-44224-1 ↗
53. Diffusion improves graph learning
54. Wu, F., Souza, A., Zhang, T., Fifty, C., Yu, T., and Weinberger, K. (2019, January 10–15). Simplifying graph convolutional networks. Proceedings of the International Conference on Machine Learning, PMLR, Long Beach, CA, USA.
55. Huang, K., Tang, J., Liu, J., Yang, R., and Xiao, X. (May, January 30). Node-wise diffusion for scalable graph learning. Proceedings of the ACM Web Conference 2023, Austin, TX, USA.10.1145/3543507.3583408 ↗