Volume 18 • Issue 2 • PP: 447-464 • 2026
Thermal Vehicle Detection and Tracking for Intelligent Transportation Systems: A Modular IoT Architecture and Staged Deployment Roadmap
Open Access & Copyright
© 2026 The Author(s). Published by ASPG. This article is licensed under the Creative Commons Attribution 4.0 International License (CC BY 4.0).
View PDF Abstract
Automated vehicle monitoring in intelligent transportation systems must operate reliably around the clock, including under conditions that routinely cripple conventional visible-light cameras: night, glare, shadows, and adverse weather. This paper proposes a modular Internet of Things (IoT) architecture for thermal-based vehicle detection, classification, and trajectory analysis, together with a four-phase deployment roadmap that connects public-dataset evaluation to live-traffic field validation. The system integrates longwave infrared (LWIR) imaging (8–14 𝜇m) with YOLO-family deep learning detectors (YOLOv8/v11/v12) and multi-object tracking algorithms (ByteTrack, BoTSORT, StrongSORT), deployed across NVIDIA Jetson edge devices and cloud infrastructure through JSON/MQTT formalized data contracts. The primary novel contribution is a system-level integration framework that bridges the gap between component-level algorithmic research and operational deployment. Concretely, this work: (i) defines five functionally independent modules with explicit interface specifications and latency budgets not previously formalized in the thermal-ITS literature; (ii) introduces quantified decision gates linking progression criteria directly to published benchmark values; (iii) provides region-specific operational availability estimates derived from empirical weather-degradation data; and (iv) integrates domain adaptation, GDPR compliance, edge hardware budgets, and regulatory WIM frameworks within a single coherent system blueprint. Domain adaptation strategies reported in peer-reviewed literature recover 20–50% of cross-dataset mAP degradation (typically 10–30%) caused by sensor and scene variability; these figures are literature benchmarks, not results obtained in this work. An optional weight-estimation module (Module 4) based on recent vision-based and bridge WIM validation studies is treated as an exploratory extension requiring site-specific validation.
Keywords
References
[1] M. Adresi, M. Lacidogna, and G. Grasso, “A review of different types of weigh-in-motion sensors: State ofthe- art,” Measurement, vol. 225, article 113976, 2024. DOI: 10.1016/j.measurement.2023.113976.
[2] J. M. R. Velázquez, L. Khoudour, G. Saint Pierre, P. Duthon, S. Liandrat, F. Bernardin, S. Fiss, I. Ivanov, and R. Peleg, “Analysis of thermal imaging performance under extreme foggy conditions: Applications to autonomous driving,” J. Imaging, vol. 8, no. 11, article 306, 2022. DOI: 10.3390/jimaging8110306.
[3] M. Vollmer and K.-P. Möllmann, Infrared Thermal Imaging: Fundamentals, Research and Applications, 2nd ed. Weinheim, Germany: Wiley-VCH, 2018, ISBN: 978-3-527-41351-5. DOI: 10.1002/9783527693306.
[4] NVIDIA Corporation, “Jetson AGX Xavier, Xavier NX and Nano developer kits,” [Online]. Available: https: //developer.nvidia.com/embedded/jetson, accessed Apr. 2025.
[5] Ultralytics, “Ultralytics YOLO: State-of-the-art real-time object detection,” [Online]. Available: https: //github.com/ultralytics/ultralytics, accessed Apr. 2025.
[6] M. Ding, Z. Han, and X. Li, “TIR-YOLO-ADAS: A thermal infrared object detection framework for advanced driver assistance systems,” IET Intell. Transp. Syst., vol. 18, no. 3, pp. 512–524, 2024. DOI: 10.1049/itr2.12471.
[7] Y. Zhang, P. Sun, Y. Jiang, D. Yu, F. Weng, Z. Yuan, P. Luo, W. Liu, and X. Wang, “ByteTrack: Multiobject tracking by associating every detection box,” in Proc. Eur. Conf. Comput. Vis. (ECCV), Lecture Notes Comput. Sci., vol. 13682, Springer, Cham, 2022, pp. 1–21. DOI: 10.1007/978-3-031-20047-2_1.
[8] N. Aharon, R. Orfaig, and B.-Z. Bobrovsky, “BoT-SORT: Robust associations multi-pedestrian tracking,” arXiv preprint arXiv:2206.14651, 2022. Available: https://arxiv.org/abs/2206.14651.
[9] N. Wojke, A. Bewley, and D. Paulus, “Simple online and realtime tracking with a deep association metric,” in Proc. IEEE Int. Conf. Image Process. (ICIP), 2017, pp. 3645–3649. DOI: 10.1109/ICIP.2017.8296962.
[10] Y. Du, Z. Zhao, Y. Song, Y. Zhao, F. Su, T. Gong, and H. Meng, “StrongSORT: Make DeepSORT great again,” IEEE Trans. Multimedia, vol. 25, pp. 8725–8737, 2023. DOI: 10.1109/TMM.2023.3240881.
[11] B. Jacob, E. J. O’Brien, and S. Jehaes, Weigh-in-Motion of Road Vehicles: Final Report of the COST 323 Action. LCPC, Paris, France, 1999.
[12] S. Hwang, J. Park, N. Kim, Y. Choi, and I. S. Kweon, “Multispectral pedestrian detection: Benchmark dataset and baseline,” in Proc. IEEE CVPR, 2015, pp. 1037–1045. DOI: 10.1109/CVPR.2015.7298706.
[13] Y. Tian, D. Ye, and L. Zhang, “YOLOv12: Attention-centric real-time object detectors,” arXiv preprint
arXiv:2502.12524, Feb. 2025. https://arxiv.org/abs/2502.12524
[14] R. Ahmad and M. S. Khan, “Benchmarking YOLO variants for thermal image object detection,” MSRA Journal Review, vol. 2, no. 2, pp. 45–58, 2025. Zenodo record: https://zenodo.org/records/17310153. Note: This outlet is not indexed in Web of Science or Scopus; the YOLOv11n/v12n mAP values cited herein require independent replication on FLIR ADAS V2 before use as primary ISI benchmarks.
[15] U. Uzun, C. Sissons, and F. Murat, “Self-supervised domain adaptation for thermal object detection,” in Proc. IEEE/CVF Winter Conf. Appl. Comput. Vis. (WACV), 2021, pp. 2750–2759. DOI: 10.1109/WACV48630.2021.00280.
[16] U. Ustun, M. Ozay, and F. Murat, “Spectral transfer guided active domain adaptation for thermal imagery,” in Proc. IEEE/CVF CVPR Workshops (CVPRW), 2023, pp. 1–10. DOI: 10.1109/CVPRW59228.2023.00449.
[17] Y. Wang, C. Li, Z. Sun, and J. Liu, “Unsupervised thermal-to-visible domain adaptation for pedestrian detection,” Pattern Recognit. Lett., vol. 153, pp. 222–228, 2022. DOI: 10.1016/j.patrec.2021.11.024.
[18] K. Karthikeyan, P. Rajasekaran, S. Selvi, and M. Kalaivani, “A secured IoT-based intelligent transport system framework for smart cities,” Cluster Comput., vol. 26, pp. 2703–2719, 2023. DOI: 10.1007/s10586-022-03694- 6.
[19] M. Urblik, M. Kopp, and M. Dolanský, “A modular framework for data processing at the edge,” Sensors, vol. 23, no. 18, article 7774, 2023. DOI: 10.3390/s23187774.
[20] Y. Gong, X. Li, Z. Chen, T. Liu, and W. Wang, “A multispectral vision-based machine learning framework for non-contact vehicle weigh-in-motion,” Measurement, vol. 228, article 114162, 2024. DOI: 10.1016/j.measurement.2024.114162.
[21] A. Tempelhahn, H. Budzier, V. Krause, and G. Gerlach, “Shutter-less calibration of uncooled infrared cameras,” J. Sensors Sens. Syst., vol. 5, no. 1, pp. 9–16, Jan. 2016. DOI: 10.5194/jsss-5-9-2016.
[22] V. S. Vibashan, D. Poster, S. You, S. Hu, and V. M. Patel, “Meta-UDA: Unsupervised domain adaptive thermal object detection using meta-learning,” in Proc. IEEE/CVFWinter Conf. Appl. Comput. Vis. (WACV),Waikoloa, HI, USA, Jan. 2022, pp. 1412–1423. https://openaccess.thecvf.com/content/WACV2022/html/VS_ Meta-UDA_Unsupervised_Domain_Adaptive_Thermal_Object_Detection_Using_Meta-Learning_ WACV_2022_paper.html
[23] H. Zhao, X. Mao, C. Zhou, H. Li, and X. Liu, “Non-contact vehicle weigh-in-motion using computer vision,” Measurement, vol. 153, article 107415, 2020. DOI: 10.1016/j.measurement.2019.107415.
[24] D. Paul and K. Roy, “Application of bridge weigh-in-motion system in bridge health monitoring: A state ofthe-art review,” Structural Health Monitoring, 2023. DOI: 10.1177/14759217231154431.
[25] European Parliament and Council of the European Union, “Regulation (EU) 2016/679 on the Protection of Natural Persons with Regard to the Processing of Personal Data (General Data Protection Regulation),” Official Journal of the European Union, vol. L 119, pp. 1–88, May 2016. https://eur-lex.europa.eu/ legal-content/EN/TXT/?uri=CELEX:32016R0679.
[26] Teledyne FLIR, “FLIR ADAS Dataset V2,” [Online]. Available: https://oem.flir.com/solutions/ automotive/adas-dataset/, accessed Jan. 2024.
[27] F. McKenna, M. H. Scott, G. L. Fenves, and B. Jeremić, “Open System for Earthquake Engineering Simulation (OpenSees),” [Online]. Available: https://opensees.berkeley.edu/, accessed 2024.
[28] W. G. Cochran, Sampling Techniques, 3rd ed. New York, NY, USA: John Wiley & Sons, 1977.
[29] X. Jia, C. Zhu, M. Li, W. Tang, and W. Zhou, “LLVIP: A visible-infrared paired dataset for low-light vision,” in Proc. IEEE/CVF Int. Conf. Comput. Vis. Workshops (ICCVW), Montreal, QC, Canada, Oct. 2021, pp. 3496–3504. DOI: 10.1109/ICCVW54120.2021.00389.
Cite This Article
Choose your preferred format
Publisher's Note
The statements, opinions, and data presented in this article are solely those of the author(s) and do not necessarily represent those of ASPG, the journal, or its editors. ASPG and the editors disclaim responsibility for any harm arising from the use of any ideas, methods, instructions, or products described in this article, to the fullest extent permitted by applicable law.