A staged tunnel registration method integrating dynamic pooling-based adaptive registration and neighborhood feature-assisted DTW

doi:10.11947/j.AGCS.2026.20250550

Abstract

Abstract:

To address cross-epoch misalignment of multi-epoch tunnel mobile laser scanning data caused by accumulated mileage errors and variations in acquisition poses, a staged registration method integrating dynamic pooling-based adaptive registration and neighborhood feature-assisted dynamic time warping (DTW) is proposed. First, two-dimensional tunnel images are generated from point-cloud intensity information, and local features are extracted using SuperPoint. Building upon the rotary positional encoding and attention-based matching framework of LightGlue, a dynamic pooling strategy tailored to the feature distribution of tunnel images is introduced to achieve coarse mileage alignment over large offsets between multi-epoch datasets. Subsequently, tunnel cross-sectional point clouds are transformed into radial sequences in polar coordinates, and dynamic time warping is employed to establish cross-epoch correspondences between tunnel cross-sections. Adaptive-neighborhood intrinsic shape signature (ISS) keypoint extraction and least-squares optimization are then applied to achieve refined cross-sectional registration. Experiments were conducted on multi-epoch mobile laser scanning data acquired from shield and mined tunnels of Chongqing Metro. The image-registration and point-cloud-registration errors were 1.3 and 1.2 cm, respectively, for the shield tunnel, and 3.7 and 3.6 cm, respectively, for the mined tunnel. After refined registration, the Chamfer distances converged to 0.1~0.3 cm, with an average of approximately 0.2 cm, representing a reduction of approximately 90% compared with those before registration. The results demonstrate that the proposed method can effectively accommodate cross-epoch alignment of tunnels with different structural characteristics and provide a reliable spatial reference for tunnel deformation monitoring, defect evolution analysis, and multi-epoch data fusion.

Key words: tunnel, mobile laser scanning, multi-epoch registration, LightGlue, dynamic time warping, ISS

CLC Number:

P232

Haili Sun, Xinwu Yan, Ping Yang, Sheng Hai, Ruofei Zhong, Mingquan Wang. A staged tunnel registration method integrating dynamic pooling-based adaptive registration and neighborhood feature-assisted DTW[J]. Acta Geodaetica et Cartographica Sinica, 2026, 55(9): 1555-1566.

Add to citation manager EndNote|Reference Manager|ProCite|BibTeX|RefWorks

URL: http://xb.chinasmp.com/EN/10.11947/j.AGCS.2026.20250550

http://xb.chinasmp.com/EN/Y2026/V55/I9/1555

Figures/Tables 9

Fig. 1

Fig. 2

Fig. 3

Tab. 1

Fig. 4

Tab. 2

Fig. 5

Fig. 6

Fig. 7

References 22

[1]	韩宝明, 余怡然, 习喆, 等. 城市轨道交通2023年度统计和分析报告[R]. 北京: 中国城市轨道交通协会, 2024.
	Han Baoming, Yu Yiran, Xi Zhe, et al. Urban rail transit 2023 annual statistics and analysis report[R]. Beijing: China Association of Metros, 2024.
[2]	陈军, 刘万增, 武昊, 等. 智能化测绘的基本问题与发展方向[J]. 测绘学报, 2021, 50(8): 995-1005. DOI: . doi: 10.11947/j.AGCS.2021.20210235
	Chen Jun, Liu Wanzeng, Wu Hao, et al. Smart surveying and mapping: fundamental issues and research agenda[J]. Acta Geodaetica et Cartographica Sinica, 2021, 50(8): 995-1005. DOI: . doi: 10.11947/j.AGCS.2021.20210235
[3]	纪长琦, 郭肇捷, 孙海丽, 等. 基于移动激光扫描的地铁隧道渗漏水定位及快速检测方法[J]. 测绘学报, 2024, 53(6): 1236-1250. DOI: . doi: 10.11947/j.AGCS.2024.20230438
	Ji Changqi, Guo Zhaojie, Sun Haili, et al. Location and rapid detection method of water leakage in subway tunnels based on mobile laser scanning[J]. Acta Geodaetica et Cartographica Sinica, 2024, 53(6): 1236-1250. DOI: . doi: 10.11947/j.AGCS.2024.20230438
[4]	林康, 石波, 杨密, 等. 基于激光点云环缝环号识别的盾构隧道病害定位方法[J]. 铁道科学与工程学报, 2023, 20(12): 4835-4847.
	Lin Kang, Shi Bo, Yang Mi, et al. Disease location method of shield tunnel based on laser point cloud ring seam ring number identification[J]. Journal of Railway Science and Engineering, 2023, 20(12): 4835-4847.
[5]	Lowe D G. Distinctive image features from scale-invariant keypoints[J]. International Journal of Computer Vision, 2004, 60(2): 91-110.
[6]	Bay H, Tuytelaars T, Van Gool L. SURF: speeded up robust features[C]//Proceedings of European Conference on Computer Vision. Berlin: Springer, 2006: 404-417.
[7]	He Kaiming, Zhang Xiangyu, Ren Shaoqing, et al. Deep residual learning for image recognition[C]//Proceedings of 2016 IEEE Conference on Computer Vision and Pattern Recognition. Las Vegas: IEEE, 2016: 770-778.
[8]	Touvron H, Cord M, Sablayrolles A, et al. Going deeper with image transformers[C]//Proceedings of the IEEE/CVF International Conference on Computer Vision. MontrealL: IEEE, 2021: 32-42.
[9]	Chen Hanting, Wang Yunhe, Guo Tianyu, et al. Pre-trained image processing transformer[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. Nashville: IEEE, 2021: 12299-12310.
[10]	De Tone D, Malisiewicz T, Rabinovich A. SuperPoint: self-supervised interest point detection and description[PP/OL].[2025-10-30]. https://doi.org/10.48550/arXiv.1712.07629.
[11]	Lindenberger P, Sarlin P E, Pollefeys M. LightGlue: local feature matching at light speed[C]//Proceedings of the 2023 IEEE/CVF International Conference on Computer Vision. Paris: IEEE, 2023.
[12]	Besl P J, McKay N D. A method for registration of 3D shapes[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 1992, 14(2): 239-256.
[13]	Qin Zheng, Yu Hao, Wang Changjian, et al. GeoTransformer: fast and robust point cloud registration with geometric transformer[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2023, 45(8): 9806-9821.
[14]	Fu Kexue, Luo Jiazheng, Luo Xiaoyuan, et al. Robust point cloud registration framework based on deep graph matching[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2023, 45(5): 6183-6195.
[15]	Zai Dawei, Li Jonathan, Guo Yulan, et al. Pairwise registration of TLS point clouds using covariance descriptors and a non-cooperative game[J]. ISPRS Journal of Photogrammetry and Remote Sensing, 2017, 134: 15-29.
[16]	赵文更, 张旭, 邵晴晴. 融合反射强度图像的地铁隧道点云自动配准[J]. 测绘通报, 2020(9): 38-41.
	Zhao Wengeng, Zhang Xu, Shao Qingqing. Automatic point cloud registration method of subway tunnels based on reflection intensity images[J]. Bulletin of Surveying and Mapping, 2020(9): 38-41.
[17]	Ji Changqi, Sun Haili, Zhong Ruofei, et al. Precise positioning method of moving laser point cloud in shield tunnel based on bolt hole extraction[J]. Remote Sensing, 2022, 14(19): 4791.
[18]	张星, 黄展鹏, 李清泉, 等. 结合配准补偿和空间约束的隧道三维建图方法[J]. 测绘学报, 2025, 54(2): 297-307. DOI: . doi: 10.11947/j.AGCS.2025.20240018
	Zhang Xing, Huang Zhanpeng, Li Qingquan, et al. 3D tunnel mapping method combining registration compensation and spatial constraint[J]. Acta Geodaetica et Cartographica Sinica, 2025, 54(2): 297-307. DOI: . doi: 10.11947/j.AGCS.2025.20240018
[19]	Fischler M A, Bolles R C. Random sample consensus: a paradigm for model fitting with applications to image analysis and automated cartography[J]. Communications of the ACM, 1981, 24(6): 381-395.
[20]	Rakthanmanon T, Campana B, Mueen A, et al. Searching and mining trillions of time series subsequences under dynamic time warping[C]//Proceedings of the 18th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. Beijing: ACM Press, 2012: 262-270.
[21]	Zhong Yu. Intrinsic shape signatures: a shape descriptor for 3D object recognition[C]//Proceedings of 12th IEEE International Conference on Computer Vision Workshops. Kyoto: IEEE, 2009: 689-696.
[22]	Fan H Q, Su H, Guibas L. A point set generation network for 3D object reconstruction from a single image[C]//Proceedings of 2017 IEEE Conference on Computer Vision and Pattern Recognition. Honolulu: IEEE, 2017.

隧道类型	里程范围/m	第一期采集时间	第二期采集时间	文件数	数据总量
盾构法	47 779.0~48 229.0	2022-09	2023-05	26	11.6×10⁹
矿山法	0.0~200.0	2022-08	2023-08	16	6.74×10⁹

隧道数据	影像匹配方法	第一阶段影像配准误差	第二阶段点云配准误差	点云配准优化误差	点云配准前/后倒角距离
盾构法	LightGlue	1.0	0.9	0.1	2.4/0.2
盾构法	动态池化自适应配准	1.3	1.2	0.1	1.8/0.3
矿山法	LightGlue	3.6	3.0	0.6	2.3/0.1
矿山法	动态池化自适应配准	3.7	3.6	0.1	0.3/0.2