Generative Design of Cold-Region Waterfront Ice and Snow Cultural Scenes Based on LiDAR Point Clouds and Generative Adversarial Network (GAN)
Main Article Content
Abstract
Accurate reconstruction and generation of complex cold-region waterfront environments remain challenging due to the coupled influence of geometric structures, environmental conditions, and physical constraints. This study proposes a multimodal conditional latent-space disentanglement and physically constrained adversarial generation framework (MC-PAG) for LiDAR-based 3D scene modeling and synthesis. The framework integrates LiDAR point-cloud measurements, semantic descriptors, and environmental parameters into a structured latent representation through multimodal feature encoding and latent-space disentanglement. A cascaded generative adversarial architecture is subsequently employed to progressively reconstruct dense spatial point-cloud distributions, while differentiable physical-consistency modules are embedded into the generation process to enforce environmental plausibility and geometric continuity. Furthermore, a multi-scale discrimination strategy is introduced to jointly evaluate spatial fidelity, semantic consistency, and physical-rule compliance. Experimental results on a dedicated cold-region waterfront dataset demonstrate that the proposed framework achieves a semantic segmentation IoU of 0.845 for ice–snow structures, maintains an average physical-rule violation rate of only 3.37%, and generates spatial layouts with an average path curvature of 0.12. The generated scenes further exhibit high geometric fidelity, environmental consistency, and controllable semantic representation. By establishing a unified pipeline for LiDAR sensing information processing, spatial environment reconstruction, and physically constrained scene generation, the proposed framework provides an effective methodology for digital environment modeling and propagation-aware spatial representation in complex outdoor environments.
Downloads
Article Details
This work is licensed under a Creative Commons Attribution 4.0 International License.
Authors who publish with this journal agree to the following terms:
- Authors retain copyright and grant the journal right of first publication with the work simultaneously licensed under a Creative Commons Attribution License that allows others to share the work with an acknowledgement of the work's authorship and initial publication in this journal.
- Authors are able to enter into separate, additional contractual arrangements for the non-exclusive distribution of the journal's published version of the work (e.g., post it to an institutional repository or publish it in a book), with an acknowledgement of its initial publication in this journal.
- Authors are permitted and encouraged to post their work online (e.g., in institutional repositories or on their website) prior to and during the submission process, as it can lead to productive exchanges, as well as earlier and greater citation of published work (See The Effect of Open Access).
References
A. A. Fime, S. Mahmud, A. Das, M. S. Islam, and J.-H. Kim, “Automatic Scene Generation: State-of-the-Art Techniques, Models, Datasets, Challenges, and Future Prospects,” IEEE Access, vol. 13, no. 1, pp. 95753-95796, 2025, doi: 10.1109/ACCESS.2025.3574298.
Z. X. Chew, J. Y. Wong, Y. H. Tang, C. C. Yip, and T. Maul, “Generative design in the built environment,” Automation in Construction, vol. 166, no. 1, Art. no. 105638, 2024, doi: 10.1016/j.autcon.2024.105638.
X. Zhuang, Y. Ju, A. Yang, and L. Caldas, “Synthesis and generation for 3D architecture volume with generative modeling,” International Journal of Architectural Computing, vol. 21, no. 2, pp. 297-314, 2023, doi: 10.1177/14780771231168233.
D. Yang, J. Wang, and X. Yang, “3D Point Cloud Shape Generation with Collaborative Learning of Generative Adversarial Network and Auto-Encoder,” Remote Sensing, vol. 16, no. 10, pp. 1772-1777, 2024, doi: 10.3390/rs16101772.
X. Han, C. Liu, Y. Zhou, K. Tan, Z. Dong, and B. Yang, “WHU-Urban3D: An urban scene LiDAR point cloud dataset for semantic instance segmentation,” ISPRS Journal of Photogrammetry and Remote Sensing, vol. 209, no. 1, pp. 500-513, 2024, doi: 10.1016/j.isprsjprs.2024.02.007.
H. Yoo, Y. Kim, J. H. Ryu, S. Lee, and J. H. Lee, “Semi-Automated Building Dataset Creation for 3D Semantic Segmentation of Point Clouds,” Electronics, vol. 14, no. 1, pp. 108-114, 2024, doi: 10.3390/electronics14010108.
K. A. Tychola, E. Vrochidou, and G. A. Papakostas, “Deep learning based computer vision under the prism of 3D point clouds: A systematic review,” The Visual Computer, vol. 40, no. 11, pp. 8287-8329, 2024, doi: 10.1007/s00371-023-03237-7.
Q. C. Xu, T. J. Mu, and Y. L. Yang, “A survey of deep learning-based 3D shape generation,” Computational Visual Media, vol. 9, no. 3, pp. 407-442, 2023, doi: 10.1007/s41095-022-0321-5.
Z. Qiu, J. Liu, Y. Xia, H. Qi, and P. Liu, “Text Semantics to Controllable Design: A Residential Layout Generation Method Based on Stable Diffusion Model,” Developments in the Built Environment, vol. 23, no. 1, Art. no. 100691, 2025, doi: 10.1016/j.dibe.2025.100691.
M. Shi, J. O. Seo, S. H. Cha, B. Xiao, and H. L. Chi, “Generative AI-powered architectural exterior conceptual design based on the design intent,” Journal of Computational Design and Engineering, vol. 11, no. 5, pp. 125-142, 2024, doi: 10.1093/jcde/qwae077.
R. Maskeliūnas, S. Maqsood, M. Vaškevičius, and J. Gelšvartas, “Fusing LiDAR and photogrammetry for accurate 3D data: A hybrid approach,” Remote Sensing, vol. 17, no. 3, pp. 1-27, 2025, doi: 10.3390/rs17030443.
J. Xu, S. Zhang, H. Jing, C. Hancock, P. Qiao, N. Shen, et al., “Improving real-scene 3D model quality of unmanned aerial vehicle oblique-photogrammetry with a ground camera,” Remote Sensing, vol. 16, no. 21, pp. 3933, 2024, doi: 10.3390/rs16213933.
Y. Shen, H. Xu, Y. Bao, Z. Xu, Y. Gu, J. Lu, et al., “Style-Based Tree GAN for Point Cloud Generator,” IEEE Access, vol. 12, no. 1, pp. 27017-27028, 2024, doi: 10.1109/ACCESS.2024.3365519.
W. Li, Y. Chen, Q. Fan, M. Yang, B. Guo, and Z. Yu, “I-PAttnGAN: An Image-Assisted Point Cloud Generation Method Based on Attention Generative Adversarial Network,” Remote Sensing, vol. 17, no. 1, pp. 153-157, 2025, doi: 10.3390/rs17010153.
W. Zhong and H. Meidani, “Physics-informed geometry-aware neural operator,” Computer Methods in Applied Mechanics and Engineering, vol. 434, no. 1, Art. no. 117540, 2025, doi: 10.1016/j.cma.2024.117540.
X. Hu, Q. Ma, P. Zhao, and X. Wang, “Physics-aware neural operator for high-fidelity fluid dynamics modeling with geometric and spectral priors,” Physics of Fluids, vol. 37, no. 11, Art. no. 115111, 2025, doi: 10.1063/5.0299765.
Y. H. Shin, K. W. Son, and D. C. Lee, “Semantic segmentation and building extraction from airborne LiDAR data with multiple return using PointNet++,” Applied Sciences, vol. 12, no. 4, pp. 1975-1982, 2022, doi: 10.3390/app12041975.
X. Nong, W. Bai, and G. Liu, “Airborne LiDAR point cloud classification using PointNet++ network with full neighborhood features,” Plos One, vol. 18, no. 2, Art. no. e0280346, 2023, doi: 10.1371/journal.pone.0280346.
S. Molnar and L. Tamas, “Variational autoencoders for 3D data processing,” Artificial Intelligence Review, vol. 57, no. 2, pp. 42-48, 2024, doi: 10.1007/s10462-023-10687-x.
J. Dai, Q. Guo, G. Wang, X. Liu, and Z. Zheng, “An optimized method for variational autoencoders based on Gaussian cloud model,” Information Sciences, vol. 645, no. 1, Art. no. 119358, 2023, doi: 10.1016/j.ins.2023.119358.
N. Balakrishnan, F. Buono, C. Calì, and M. Longobardi, “Dispersion indices based on Kerridge inaccuracy measure and Kullback-Leibler divergence,” Communications in Statistics-Theory and Methods, vol. 53, no. 15, pp. 5574-5592, 2024, doi: 10.1080/03610926.2023.2222926.
Z. Lu, B. Cao, and Q. Hu, “LiDAR-Camera Continuous Fusion in Voxelized Grid for Semantic Scene Completion,” IEEE Transactions on Circuits and Systems for Video Technology, vol. 34, no. 12, pp. 12330-12344, 2024, doi: 10.1109/TCSVT.2024.3435045.
S. Ye, X. Di, M. Liao, and X. Li, “Enhancing realism in LiDAR scene generation with CSPA-DFN and linear crossattention via Diffusion Transformer model,” Neural Networks, vol. 189, no. 1, Art. no. 107503, 2025, doi: 10.1016/j.neunet.2025.107503.
Q. Xu, X. Guan, J. Cao, Y. Ma, and H. Wu, “MPR-GAN: A novel neural rendering framework for MLS point cloud with deep generative learning,” IEEE Transactions on Geoscience and Remote Sensing, vol. 60, no. 1, pp. 1-16, 2022, doi: 10.1109/TGRS.2022.3212389.
Y. Jia, W. Yu, and L. Zhao, “Generative adversarial networks with texture recovery and physical constraints for remote sensing image dehazing,” Scientific Reports, vol. 14, no. 1, Art. no. 31426, 2024, doi: 10.1038/s41598-024-83088-x.
H. Lee, S. G. Lee, J. Park, and J. Shim, “Improving complex scene generation by enhancing multi-scale representations of gan discriminators,” IEEE Access, vol. 11, no. 1, pp. 43067-43079, 2023, doi: 10.1109/ACCESS.2023.3270561.
X. Zeng, A. Vahdat, F. Williams, Z. Gojcic, O. Litany, S. Fidler, et al., “Lion: Latent point diffusion models for 3d shape generation,” in Koyejo S, Mohamed S, Agarwal A, Belgrave D, Cho K, Oh A, editors. Proceedings of the 36th International Conference on Neural Information Processing Systems, 28 November - 9 December 2022, New Orleans, LA, USA. Red Hook, NY, USA: Curran Associates Inc., 2022, pp. 10021-10039, doi: 10.48550/arXiv.2210.06978.
Y. Li and G. Baciu, “SG-GAN: Adversarial self-attention GCN for point cloud topological parts generation,” IEEE Transactions on Visualization and Computer Graphics, vol. 28, no. 10, pp. 3499-3512, 2021, doi: 10.1109/TVCG.2021.3069195.
R. Li, X. Li, K. H. Hui, and C. W. Fu, “SP-GAN: Sphere-guided 3D shape generation and manipulation,” ACM Transactions on Graphics (TOG), vol. 40, no. 4, pp. 1-12, 2021, doi: 10.1145/3450626.3459766.
M. M. Yapıcı and N. Topaloğlu, “Performance comparison of deep learning frameworks,” Computers and Informatics, vol. 1, no. 1, pp. 1-11, 2021, [Online]. Available: https://izlik.org/JA35GC87FU.