基于单目视觉语义地图构建的列车里程定位方法

doi:10.3778/j.issn.1002-8331.2501-0216

摘要/Abstract

摘要： 针对铁路检测列车传统里程定位方法均存在的缺陷，将语义信息融合到三维重建技术中，提出了基于先验信息的单目视觉语义地图构建方法。该方法包含目标检测模块、目标跟踪模块、单目SLAM模块和点云统计处理模块，并且对铁路场景中进行三维重建所存在的主要问题分别进行优化：通过定位稳定的语义目标，来缓解尺度漂移问题；通过设定自适应阈值和筛选“三维重建一致性”良好的特征点，来缓解铁路场景模式单一和运行速度过快，所导致的特征点匹配易发生错误的问题；通过筛选本征矩阵奇异值分解的结果，来提升算法运行速度。实验结果表明，该方法是一种可行性高、便于应用推广的高精度铁路里程定位方法。对比仅依靠轮速计的里程定位，提出的方法可将检测系统的定位误差控制在1?m左右，最大误差由11.3?m降低为1.7?m，实现了检测列车精确里程定位。

关键词: 列车里程定位, 三维重建, 同步定位与制图, 语义地图

Abstract: To address the limitations of traditional mileage positioning methods used in railway inspection trains, this paper integrates semantic information into 3D reconstruction technology and proposes a monocular vision-based semantic map construction approach using prior knowledge. The method consists of modules for object detection, object tracking, monocular SLAM, and point cloud statistical processing. It optimizes key challenges in 3D reconstruction of railway scenes by mitigating the scale drift issue through locating stable semantic targets, reducing feature point matching errors caused by the uniformity of railway scene patterns and excessive train speed by setting adaptive thresholds and selecting feature points with good 3D reconstruction consistency, and enhancing the algorithm’s computational efficiency by filtering the results of the intrinsic matrix’s singular value decomposition. Experimental results demonstrate that the proposed method is highly feasible, easy to apply, and provides high-precision railway mileage positioning. Compared to mileage positioning based solely on wheel encoders, the method can limit the positioning error of the detection system to approximately 1 meter, reducing the maximum error from 11.3 meters to 1.7 meters, thereby achieving precise mileage positioning for inspection trains.

Key words: train mileage positioning, 3D construction, simultaneous localization and mapping (SLAM), semantic map

蒋欣兰, 王胜春, 沈彦龙. 基于单目视觉语义地图构建的列车里程定位方法[J]. 计算机工程与应用, 2025, 61(19): 336-347.

JIANG Xinlan, WANG Shengchun, SHEN Yanlong. Train Mileage Positioning Method Based on Monocular Vision Semantic Map Construction[J]. Computer Engineering and Applications, 2025, 61(19): 336-347.

参考文献

[1] 夏博光. 基于RFID技术的高速综合检测列车时空校准系统研究[D]. 北京: 中国铁道科学研究院, 2011.
XIA B G. The research of high-speed comprehensive inspection train space-time calibration system based on RFID technology[D]. Beijing: China Academy of Railway Sciences, 2011.
[2] 龚利, 赵延杰, 朱明辉. 一种基于北斗和5G技术融合的复杂环境下机车定位方法[J]. 北京交通大学学报, 2021, 45(2): 44-51.
GONG L, ZHAO Y J, ZHU M H. A fusion method based on Beidou and 5G technology for locomotive positioning in complex environment[J]. Journal of Beijing Jiaotong University, 2021, 45(2): 44-51.
[3] MUR-ARTAL R, MONTIEL J M M, TARDóS J D. ORB-SLAM: a versatile and accurate monocular SLAM system[J]. IEEE Transactions on Robotics, 2015, 31(5): 1147-1163.
[4] KHALIQ S, ANJUM M L, HUSSAIN W, et al. Why ORB-SLAM is missing commonly occurring loop closures?[J]. Autonomous Robots, 2023, 47(8): 1519-1535.
[5] KOESTLER L, YANG N, ZELLER N, et al. TANDEM: tracking and dense mapping in real-time using deep multi-view stereo[C]//Proceedings of the Conference on Robot Learning, 2022: 34-45.
[6] MILDENHALL B, SRINIVASAN P P, TANCIK M, et al. NeRF: representing scenes as neural radiance fields for view synthesis[J]. Communications of the ACM, 2021, 65(1): 99-106.
[7] 朱沛尧, 周海波, 张浩宇, 等. 移动机器人视觉同步定位与建图方法[J/OL]. 天津理工大学学报: 1-10[2025-01-05]. https://link.cnki.net/urlid/12.1374.N.20240514.1049.004.
ZHU P Y, ZHOU H B, ZHANG ＨY, et al. Visual simultaneous localization and mapping method for mobile robot[J/OL]. Journal of Tianjin University of Technology: 1-10[2025-01-05]. https://link.cnki.net/urlid/12.1374.N.20240514.1049.004.
[8] 张干, 周非, 张阔, 等. 基于语义和几何一致性的视觉SLAM回环检测算法[J]. 计算机工程与应用, 2024, 60(20): 180-188.
ZHANG G, ZHOU F, ZHANG K, et al. Visual SLAM loop closure algorithm based on semantic and geometric consistency[J]. Computer Engineering and Applications, 2024, 60(20): 180-188.
[9] 沈斯杰, 田昕, 魏国亮, 等. 基于2D激光雷达的SLAM算法研究综述[J]. 计算机技术与发展, 2022, 32(1): 13-18.
SHEN S J, TIAN X, WEI G L, et al. Review of SLAM algorithm based on 2D lidar[J]. Computer Technology and Development, 2022, 32(1): 13-18.
[10] PARKINSON B, SPILKER JR J J, AXELRAD P, et al. Global positioning system: theory and applications, volume I[M]. [S. l.]: American Institute of Aeronautics and Astronautics, 1996.
[11] PEREZ-RUIZ M, SLAUGHTER D C, GLIEVER C, et al. Tractor-based real-time kinematic-global positioning system (RTK-GPS) guidance system for geospatial mapping of row crop transplant[J]. Biosystems Engineering, 2012, 111(1): 64-71.
[12] WANG Y S, SONG W W, LOU Y D, et al. Rail vehicle locali-
zation and mapping with LiDAR-vision-inertial-GNSS fusion[J]. IEEE Robotics and Automation Letters, 2022, 7(4): 9818-9825.
[13] CHEN C, ZHU H, WANG L, et al. A stereo visual-inertial SLAM approach for indoor mobile robots in unknown environments without occlusions[J]. IEEE Access, 2019, 7: 185408-185421.
[14] TSCHOPP F, SCHNEIDER T, PALMER A W, et al. Experimental comparison of visual-aided odometry methods for rail vehicles[J]. IEEE Robotics and Automation Letters, 2019, 4(2): 1815-1822.
[15] OTEGUI J, BAHILLO A, LOPETEGI I, et al. A survey of train positioning solutions[J]. IEEE Sensors Journal, 2017, 17(20): 6788-6797.
[16] DENG Z X, SONG H F, HUANG H, et al. Multi-sensor based train localization and data fusion in autonomous train control system[C]//Proceedings of the 2020 Chinese Automation Congress. Piscataway: IEEE, 2020: 5702-5707.
[17] SONG H F, SUN Z Y, WANG H W, et al. Enhancing train position perception through AI-driven multi-source information fusion[J]. Control Theory and Technology, 2023, 21(3): 425-436.
[18] 杜少聪, 张红钢, 王小敏. 基于改进YOLOv5的钢轨表面缺陷检测[J]. 北京交通大学学报, 2023, 47(2): 129-136.
DU S C, ZHANG H G, WANG X M. Rail surface defect detection based on improved YOLOv5[J]. Journal of Beijing Jiaotong University, 2023, 47(2): 129-136.
[19] 刘海斌, 张友兵, 周奎, 等. 改进YOLOv5-S的交通标志检测算法[J]. 计算机工程与应用, 2024, 60(5): 200-209.
LIU H B, ZHANG Y B, ZHOU K, et al. Traffic sign detection algorithm based on improved YOLOv5-S[J]. Computer Engineering and Applications, 2024, 60(5): 200-209.
[20] JIANG P Y, ERGU D J, LIU F Y, et al. A review of YOLO algorithm developments[J]. Procedia Computer Science, 2022, 199: 1066-1073.
[21] CAO J K, PANG J M, WENG X S, et al. Observation-centric SORT: rethinking SORT for robust multi-object tracking[C]//Proceedings of the 2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE, 2023: 9686-9696.
[22] FORTUN D, BOUTHEMY P, KERVRANN C. Optical flow modeling and computation: a survey[J]. Computer Vision and Image Understanding, 2015, 134: 1-21.
[23] FEI C J, ZHANG Q L, CAI Z J, et al. Edge assisted fast optical flow matching SLAM in underground rescue environments[C]//Proceedings of the Chinese Conference on Pattern Recognition and Computer Vision. Singapore: Springer, 2025: 3-17.