同時定位與地圖構建在自動駕駛系統的實地部署中扮演關鍵角色。隨著幾何同時定位與地圖構建技術的成熟，無論是基於點雲或影像的語意輔助定位方法也在近年獲得顯著發展。然而，在基於光學雷達的同時定位與地圖構建研究中，現有工作多著重於降低動態環境對定位的干擾，對於所建立語意地圖的進一步利用仍有開發空間。本文提出一套基於光學雷達的物件追蹤架構，特別針對室內封閉場景進行定位優化。我們透過牆面偵測與追蹤，在多時段資料間建立關聯，使每次定位能更有效地依賴既有區域進行未知區域的拓展。此方法提升了同時定位與地圖構建在時序上的一致性與地圖建構的連續性，進而增強了室內環境下的定位穩定性。

SLAM plays a pivotal role in the real-world deployment of autonomous driving systems. With the advancement of geometric SLAM techniques, both point cloud-based and image-based semantic-assisted localization methods have seen significant progress in recent years. However, in LiDAR-based SLAM research, most existing works primarily focus on mitigating the impact of dynamic environments on localization accuracy. The potential for further leveraging the constructed semantic maps remains underexplored.
In this thesis, we propose a LiDAR-based object tracking framework specifically designed to enhance localization performance in enclosed indoor environments. By detecting and tracking wall structures, our approach establishes temporal correspondences across multi-session data, enabling each localization task to better rely on previously mapped regions when extending into unknown areas. This method improves temporal consistency in SLAM and enhances the continuity of map construction, thereby strengthening localization robustness in indoor settings.

[1] C. Choy, J. Gwak, and S. Savarese, “4D Spatio-Temporal ConvNets: Minkowski Convolutional Neural Networks,” in Proc. IEEE/CVF Conf. Comput. Vis. Pattern Recognit. (CVPR), pp. 3075–3084, 2019.
[2] Z. Zheng, X. Wang, D. Yang, et al., “Cylinder3D: An Effective 3D Framework for Driving-Scene LiDAR Semantic Segmentation,” in Proc. IEEE/CVF Conf. Comput. Vis. Pattern Recognit. (CVPR), pp. 9939–9948, 2021.
[3] A. Hornung, K. M. Wurm, M. Bennewitz, C. Stachniss, and W. Burgard, “OctoMap: An Efficient Probabilistic 3D Mapping Framework Based on Octrees,” Autonomous Robots, vol. 34, no. 3, pp. 189–206, 2013.
[4] C. R. Qi, L. Yi, H. Su, and L. J. Guibas, “Panoptic Multi-TSDFs: Object-Aware 3D Mapping,” in Proc. IEEE Int. Conf. Robot. Autom. (ICRA), 2021.
[5] A. Dai, A. Chang, M. Savva, M. Halber, T. Funkhouser, and M. Nießner, “ScanNet: Richly-annotated 3D Reconstructions of Indoor Scenes,” in Proc. IEEE/CVF Conf. Comput. Vis. Pattern Recognit. (CVPR), 2017.
[6] I. Armeni, O. Sener, A. R. Zamir, H. Jiang, I. Brilakis, M. Fischer, and S. Savarese, “3D Semantic Parsing of Large-Scale Indoor Spaces,” in Proc. IEEE/CVF Conf. Comput. Vis. Pattern Recognit. (CVPR), 2016.
[7] T. Shan, B. Englot, and D. Meyers, “LIO-SAM: Tightly-coupled Lidar Inertial Odometry via Smoothing and Mapping,” in Proc. IEEE/RSJ Int. Conf. Intell. Robots Syst. (IROS), pp. 5135–5142, 2020.
[8] Y. Anebarassane, D. K. P., and A. Chandru, “Enhancing ORB-SLAM3 with YOLO-based Semantic Segmentation in Robotic Navigation,” in Proc. IEEE Conf., 2023.
[9] T. Qin, P. Li, and S. Shen, “VINS-Mono: A Robust and Versatile Monocular Visual-Inertial State Estimator,” IEEE Trans. Robot., pp. 1004–1020, 2018.
[10] J. Zhang and S. Singh, “LOAM: Lidar Odometry and Mapping in Real-time,” in Robotics: Science and Systems (RSS), 2018.
[11] T. Qin, S. Cao, and J. Pan, “A General Optimization-based Framework for Global Pose Estimation with Multiple Sensors,” arXiv preprint arXiv:1901.03642, 2019.
[12] T. Shan, B. Englot, and C. Ratti, “LVI-SAM: Tightly-coupled Lidar-Visual-Inertial Odometry via Smoothing and Mapping,” in Proc. IEEE Int. Conf. Robot. Autom. (ICRA), pp. 5192–5198, 2021.
[13] C. Yu, X.-J. Liu, and F. Xie, “DS-SLAM: A Semantic Visual SLAM towards Dynamic Environments,” in Proc. IEEE/RSJ Int. Conf. Intell. Robots Syst. (IROS), pp. 1168–1174, 2018.
[14] B. Bescos, J. D. Tardós, and J. Neira, “DynaSLAM II: Tightly-Coupled Multi-Object Tracking and SLAM,” IEEE Robot. Autom. Lett., pp. 5191–5198, 2021.
[15] S. Yang and S. Scherer, “CubeSLAM: Monocular 3-D Object SLAM,” IEEE Trans. Robot., pp. 925–938, 2019.
[16] Livox, “LiDAR istortion removal,” [Online]. Available: https://www.livoxtech.com. Accessed: Jul. 2025.
[17] J. Redmon, S. Divvala, R. Girshick, and A. Farhadi, “You Only Look Once: Unified, Real-Time Object Detection,” in Proc. IEEE/CVF Conf. Comput. Vis. Pattern Recognit. (CVPR), pp. 779–788, 2016.
[18] K. He, G. Gkioxari, P. Dollár, and R. Girshick, “Mask R-CNN,” in Proc. IEEE Int. Conf. Comput. Vis. (ICCV), pp. 2961–2969, 2017.
[19] A. Rosinol, M. Abate, Y. Chang, and L. Carlone, “Kimera: An open-source library for real-time metric-semantic localization and mapping,” in Proc. IEEE Int. Conf. Robot. Autom. (ICRA), 2020.
[20] J. McCormac, A. Handa, A. J. Davison, and S. Leutenegger, “SemanticFusion: Dense 3D Semantic Mapping with Convolutional Neural Networks,” in Proc. IEEE Int. Conf. Robot. Autom. (ICRA), pp. 4628–4635, 2017.
[21] M. Aygün, “4D Panoptic LiDAR Segmentation,” in Proc. IEEE/CVF Conf. Comput. Vis. Pattern Recognit. (CVPR), 2021.
[22] Z. Zhou, Y. Zhang, and H. Foroosh, “Panoptic-PolarNet: Proposal-free LiDAR Point Cloud Panoptic Segmentation,” in Proc. IEEE/CVF Conf. Comput. Vis. Pattern Recognit. (CVPR), 2021.
[23] B. Cao, C.-N. Ritter, D. Göhring, and R. Rojas, “Accurate Localization of Autonomous Vehicles Based on Pattern Matching and Graph-Based Optimization in Urban Environments,” in Proc. IEEE Intell. Transp. Syst. Conf. (ITSC), 2020.
[24] B. Cao, R. C. Mendoza, A. Philipp, and D. Göhring, “LiDAR-Based Object-Level SLAM for Autonomous Vehicles,” in Proc. IEEE/RSJ Int. Conf. Intell. Robots Syst. (IROS), pp. 4397–4404, 2021.