Abstract

This paper presents a model, an integrated system for real-time vehicle tracking and overspeeding violation detection in traffic surveillance video. The system comprises two synergistic mod ules: a deep feature–based Re-Identification (ReID) tracker and a YOLO-powered speed estimation pipeline. The ReID tracker ex tracts 512-dimensional embeddings from pairs of images captured 30 seconds apart, computes cosine similarities to associate vehicle identities across time, and exports match results for further anal ysis. The speed estimation module processes video frames sam pled every 2s, applies YOLOv8x to detect vehicles within a de f ined region of interest, and employs centroid-based distance mea surement calibrated at 20 pixels/m to compute per-vehicle speeds. Vehicles exceeding the 50 km/h limit are flagged as violations and annotated with red bounding boxes, while compliant vehicles are marked in green. The model outputs both a detailed Excel log of identity matches and a fully annotated video illustrating tracked ve hicles with overlaid speed labels. Experimental evaluation demon strates robust identity association under varying viewpoints and accurate speed violation reporting in standard surveillance scenar ios. Future extensions will focus on automated camera calibration, cross-camera tracking, and edge deployment for low-latency, scal able traffic monitoring applications.

References

• Shailendra Singh Kathait, Ashish Kumar, Ram Patidar, Khushi Agrawal, Samay Sawal (2024). Computer Vision and Deep Learning based Approach for Traffic Violations due to Overspeeding and Wrong Direction Detection. International Journal of Computer Applications, paper-id: 6e503f15-f6c9- 4ee2- 9212-4db588484729, DOI: 10.5120/ijca2025924477.
• Shailendra Singh Kathait, Ashish Kumar, Ram Patidar, Khushi Agrawal, Samay Sawal (2024). Computer Vision and Deep Learning based Approach for Violations due to Illegal Parking Detection. International Journal of Computer Applications, DOI: 10.5120/ijca2025924506.
• Shailendra Singh Kathait, Ashish Kumar, Ram Patidar, Khushi Agrawal, Samay Sawal (2024). Deep Learning-based Approach for Detecting Traffic Violations Involving No Helmet Use and Wrong Cycle Lane Usage. International Journal of Computer Applications, DOI: 10.5120/ijca2025924714.
• Shailendra Singh Kathait, Ashish Kumar, Ram Patidar, Khushi Agrawal, Samay Sawal (2024). Deep Learning-Based Person Tracking: A Smart Approach to Security and Civic Monitoring. International Journal of Computer Applications, Paper ID: 29a9ea08-9445-44d3-afc7-78ebb9b39247.
• H. Liu, Y. Tian, Y. Wang, L. Pang, and T. Huang, “PROVID: Progressive and Multi-Granularity Vehicle ReID for Urban Surveillance,” in Proc. IEEE CVPR, 2018.
• D. Zapletal and A. Herout, “Vehicle Re-Identification for Automatic Video Traffic Surveillance,” in CVPR Workshops, 2016.
• M. Amiri, E. Pakzad, and A. Mohammadi, “A Survey on Ve hicle Re-Identification: Datasets, Methods, and Challenges,” arXiv preprint arXiv:2401.10643, 2024.
• M. Jamil, M. Rahman, and M. Hasan, “A Real-Time Vehicle Counting, Speed Estimation, and Classification System Based on Virtual Detection Zone and YOLO,” Sensors, vol. 21, no. 20, pp. 6843, 2021.
• Y. Wang et al., “YOLOv8: Ultralytics Real-Time Object De tection,” arXiv preprint arXiv:2302.05781, 2023.
• A. Kosti´c and M. Subasi´c, “SpeedNet: Real-Time Monocu lar Vehicle Speed Estimation Using Deep Learning,” arXiv preprint arXiv:2505.01203, 2025.
• L. Wang, C. Xu, and J. Liu, “A Multi-Object Vehicle Detec tion and Tracking SystemBasedonYOLOv8andByteTrack,” Electronics, vol. 13, no. 15, pp. 3033, 2024.
• Ultralytics, “YOLOv8 Tracking Documentation,” [Online]. Available: https://docs.ultralytics.com/modes/ track/, Accessed: May 2025.