AI-Based Train Localization Using Railway Infrastructure Object Detection

Authors

  • Vladimir A. Fedorov *

    1. Department of Electronic Engineering, Ural Federal University, Yekaterinburg 620002, Russia
    2. Department of Research and Development for Integrated Security Systems, NPO SAUT LLC, Yekaterinburg 620027, Russia

DOI:

https://doi.org/10.55121/tdr.v2i2.379

Keywords:

Railway Automation, Train Localization, Railway Infrastructure, Object Detection, Convolutional Neural Networks, YOLOv11, NPU

Abstract

This paper presents an Artificial Intelligence (AI)-based approach to train localization through the detection of
railway infrastructure objects using convolutional neural networks. The proposed system identifies key visual landmarks such as traffic lights, level crossings, tunnels, bridges, and passenger platforms directly from live video streams captured by onboard cameras during train operation. This enables accurate and continuous localization without relying on satellite navigation systems or additional trackside infrastructure. The object detection model is based on the You Only Look Once (YOLOv11) architecture. It is trained using high-performance Graphics Processing Unit (GPU) resources and subsequently converted and optimized for deployment on the energy-efficient RK3588 neural processing unit (NPU). The system achieves a mean average precision of mAP@0.5:0.95 = 0.52 and operates in real time at approximately 35 frames per second, meeting the practical requirements for onboard applications. Compared to traditional Global Navigation Satellite System (GNSS)-based solutions, the proposed method is inherently resilient to signal jamming and spoofing while significantly reducing infrastructure costs. Its low power consumption and high-speed inference make it especially well-suited for integration into modern railway systems operating at higher automation levels. The results confirm the feasibility of this AI-driven approach as a scalable and robust solution for train localization in diverse operational conditions.

References

[1] Barruffo, L., Caiazzo, B., Petrillo, A., & Santini, S. (et al., 2024. A GoA4 control architecture for the autonomous driving of high-speed trains over ETCS: Design and experimental validation. IEEE Transactions on Intelligent Transportation Systems. 25(6), 5096–5111. DOI: https://doi.org/10.1109/TITS.2023.3338295

[2] Zidan, J., Adegoke, E.I., Kampert, E., et al., 2020. GNSS vulnerabilities and existing solutions: A review of the literature. IEEE Access. 9, 153960–153976. DOI: https://doi.org/10.1109/ACCESS.2020.2973759

[3] Li, Z., Liu, F., Yang, W., et al., 2021. A survey of convolutional neural networks: Analysis, applications, and prospects. IEEE Transactions on Neural Networks and Learning Systems. 33(12), 6999–7019. DOI: https://doi.org/10.1109/TNNLS.2021.3084827

[4] Khanam, R., Hussain, M., 2024. Yolov11: An overview of the key architectural enhancements. [Preprint]. DOI: https://doi.org/10.48550/arXiv.2410.17725

[5] Jiang, S., Lu, D., Cai, B., 2019. GNSS NLOS Signal Modeling and Quantification Method in Railway Urban Canyon Environment. Proceedings of The 2019 IEEE Intelligent Vehicles Symposium (IV); June 9–12, 2019; Paris, France. pp. 1268–1273. DOI: https://doi.org/10.1109/IVS.2019.8814194

[6] Dong, D., Wang, M., Chen, W., et al., 2016. Mitigation of multipath effect in GNSS short baseline positioning by the multipath hemispherical map. Journal of Geodesy. 90, 255–262. DOI: https://doi.org/10.1007/s00190-015-0870-9

[7] Seferoglu, K.T., Turk, A.S., 2019. Review of Spoofing and Jamming Attack on the Global Navigation Systems Band and Countermeasure. Proceedings of The 2019 9th International Conference on Recent Advances in Space Technologies (RAST); June 11–14, 2019; Istanbul, Turkey. pp. 513–520. DOI: https://doi.org/10.1109/RAST.2019.8767871

[8] Broumandan, A., Siddakatte, R., Lachapelle, G., 2017. An approach to detect GNSS spoofing. IEEE Aerospace and Electronic Systems Magazine. 32(8), 64–75. DOI: https://doi.org/10.1109/MAES.2017.160190

[9] Zhuang, Q., Dong, D., Cui, K., et al., 2015. Detection and Compensation for Coded Odometer Errors Due to Train Wheel Slips and Slides. Proceedings of The Fifth International Conference on Transportation Engineering (ICTE 2015); September 26–27, 2015; Dalian, China. pp. 1640–1648. DOI: https://doi.org/10.1061/9780784479384.207

[10] Chen, Q., Niu, X., Zuo, L., et al., 2018. A railway track geometry measuring trolley system based on aided INS. Sensors. 18(2), 538. DOI: https://doi.org/10.3390/s18020538

[11] He, Y., Li, J., Liu, J., 2023. Research on GNSS INS & GNSS/INS integrated navigation method for autonomous vehicles: A survey. IEEE Access. 11, 79033–79055. DOI: https://doi.org/10.1109/ACCESS.2023.3299290

[12] Malakar, B., Roy, B.K., 2014. Survey of RFID Applications in Railway Industry. Proceedings of The 2014 First International Conference on Automation, Control, Energy and Systems (ACES); February 1–2, 2014; Adisaptagram, India. pp. 1–6. DOI: https://doi.org/10.1109/ACES.2014.6807999

[13] Dhahbi, S., Abbas-Turki, A., Hayat, S., et al., 2011. Study of the High-Speed Trains Positioning System: European Signaling System ERTMS/ETCS. Proceedings of The 2011 4th International Conference on Logistics; May 31–June 3, 2011; Hammamet, Tunisia. pp. 468–473. DOI: https://doi.org/10.1109/LOGISTIQUA.2011.5939444

[14] Fedorov, V.A., 2024. Recognizing Railway Infrastructure Using CNN and Stereoscopic Vision. Proceedings of The 2024 International Russian Smart Industry Conference (SmartIndustryCon); March 25–29, 2024; Sochi, Russia. pp. 13–18. DOI: https://doi.org/10.1109/SmartIndustryCon61328.2024.10516208

[15] Fedorov, V.A., 2024. Railway infrastructure detection based on YOLOv8 with npu acceleration. Doklady Mathematics. 110(Suppl 1), S42–S48. DOI: https://doi.org/10.1134/S1064562424601951

[16] Jegham, N., Koh, C.Y., Abdelatti, M., et al., 2024. Evaluating the evolution of yolo (you only look once) models: A comprehensive benchmark study of yolo11 and its predecessors. [Preprint]. DOI: https://doi.org/10.48550/arXiv.2411.00201

[17] Dewi, C., Juli Christanto, H., 2022. Combination of deep cross-stage partial network and spatial pyramid pooling for automatic hand detection. Big Data and Cognitive Computing. 6(3), 85. DOI: https://doi.org/10.3390/bdcc6030085

[18] Chen, X., Jiang, N., Yu, Z., et al., 2025. Citrus Leaf Disease Detection Based on Improved YOLO11 with C3K2. Proceedings of The International Conference on Computer Graphics, Artificial Intelligence, and Data Processing (ICCAID 2024); November 8–10, 2025; Guangzhou, China. pp. 746–751. DOI: https://doi.org/10.1109/CEI63587.2024.10871723

[19] He, K., Zhang, X., Ren, S., et al., 2015. Spatial pyramid pooling in deep convolutional networks for visual recognition. IEEE Transactions on Pattern Analysis and Machine Intelligence. 37(9), 1904–1916. https://doi.org/10.1109/TPAMI.2015.2389824

[20] Hidayatullah, P., Syakrani, N., Sholahuddin, M.R., et al., 2025. YOLOv8 to YOLO11: A comprehensive architecture in-depth comparative review. [Preprint]. DOI: https://doi.org/10.48550/arXiv.2501.13400

[21] Imambi, S., Prakash, K.B., Kanagachidambaresan, G.R., 2021. PyTorch. In: Prakash, K.B., Kanagachidambaresan, G.R. (eds.). Programming with TensorFlow: Solution for Edge Computing Applications. Springer, Cham, Switzerland. pp. 87–104. DOI: https://doi.org/10.1007/978-3-030-57077-4_10

[22] Jin, T., Bercea, G.T., Le, T.D., et al., 2020. Compiling ONNX neural network models using MLIR. [Preprint]. DOI: https://doi.org/10.48550/arXiv.2008.08272

[23] Ismagilov, R., 2023. Performance Evaluation of the Rockchip Systems-on-Chip Through YOLOv4 Object Detection Model. Proceedings of The 2023 IEEE Ural-Siberian Conference on Biomedical Engineering, Radioelectronics and Information Technology (USBEREIT); May 15–17, 2023; Yekaterinburg, Russia. pp. 241–243. DOI: https://doi.org/10.1109/USBEREIT58508.2023.10158842

[24] Uijlings, J.R., Van De Sande, K.E., Gevers, T., et al., 2013. Selective search for object recognition. International Journal of Computer Vision. 104, 154–171. DOI: https://doi.org/10.1007/s11263-013-0620-5

[25] Lin, T.Y., Maire, M., Belongie, S., et al., 2014. Microsoft Coco: Common Objects in Context. Proceedings of The Computer Vision–ECCV 2014: 13th European Conference; September 6–12, 2014; Zurich, Switzerland. pp. 740–755. DOI: https://doi.org/10.1007/978-3-319-10602-1_48

Downloads

Issue

Vol. 2 No. 2 (November 2024): In progress

Section

Article