A Low-Cost and Accurate Train Localization Method Based on Vision-Aided IMU Calibration
IEEE Open Journal of Vehicular Technology · 2026 · 7 · 2122-2132
DOI: 10.1109/ojvt.2026.3711628 ↗Kurz gefasst
Fachartikel in der Zeitschrift IEEE Open Journal of Vehicular Technology, 2026. Verfasst von Kensuke Nagai und 4 weiteren Autor:innen. Die Arbeit nennt 43 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Japan Society for the Promotion of Science.
Abstract (Originalsprache)
Accurate train localization is crucial for collision avoidance. However, the high costs of ground-based infrastructure, such as balises and track circuits, burden railway operators. In this paper, we propose a low-cost and accurate train localization method that uses measurements from inexpensive onboard sensors and image recognition. By combining trackside object detection via a monocular camera with a track map, we compensate for the gravitational influence inherent in acceleration measurements. Furthermore, by acquiring the train’s absolute position through landmark detection, we calibrate the inevitable accumulated errors in acceleration and velocity integration. Experimental results from a commercial railway line in Japan demonstrate that our method achieves a localization accuracy of approximately 5 meters in Root Mean Square Error using only low-cost sensors, without relying on GNSS, even in rainy conditions. In summary, our proposed localization approach can be deployed in GNSS-denied environments as well as on low-budget regional lines, thereby contributing to the sustainable operation of railway systems.
Förderung
- Japan Society for the Promotion of Science
Zitieren
APA
Nagai, K., Ohnishi, W., Koseki, T., Miyoshi, S., & Hasegawa, T. (2026). A Low-Cost and Accurate Train Localization Method Based on Vision-Aided IMU Calibration. IEEE Open Journal of Vehicular Technology, 7, 2122-2132. https://doi.org/10.1109/ojvt.2026.3711628
Vancouver
Nagai K, Ohnishi W, Koseki T, Miyoshi S, Hasegawa T. A Low-Cost and Accurate Train Localization Method Based on Vision-Aided IMU Calibration. IEEE Open Journal of Vehicular Technology. 2026;7:2122-2132. doi:10.1109/ojvt.2026.3711628
BibTeX
@article{Nagai2026,
author = {Nagai, Kensuke and Ohnishi, Wataru and Koseki, Takafumi and Miyoshi, Shota and Hasegawa, Tomonori},
title = {A Low-Cost and Accurate Train Localization Method Based on Vision-Aided IMU Calibration},
journal = {IEEE Open Journal of Vehicular Technology},
year = {2026},
volume = {7},
pages = {2122--2132},
doi = {10.1109/ojvt.2026.3711628}
}
EUVIMED