Vehicle Detection and Tracking Based on YOLOv11

Haoran Gao

2025

Abstract

Since its initial proposal in 2015, the You Only Look Once (YOLO) series of object detection algorithms has rapidly become a popular research direction in real-time object detection due to its efficient single-inference mechanism. YOLO divides the image into grids and simultaneously predicts bounding boxes and class probabilities in a single forward pass, achieving rapid detection. The series has continuously optimized from YOLOv1 to the latest YOLOv11, enhancing feature extraction capabilities, multi-scale perception abilities, and detection accuracy. This paper explores the application of the YOLOv11 algorithm and advanced tracking models (ByteTrack and BoTSORT) in traffic monitoring systems. Ultimately, YOLOv11 achieved a mAP50 of 0.806 and a mAP50-95 of 0.501; precision reached 1.0 under a confidence level of 0.988 and a recall rate of 68.8% when the confidence threshold was 0, with a final frame rate of 63fps. The ByteTrack and BoTSORT tracking algorithms ensured stability and accuracy in tracking through multi-stage data association and trajectory management.

Paper Citation

in Harvard Style

Gao H. (2025). Vehicle Detection and Tracking Based on YOLOv11. In Proceedings of the 2nd International Conference on Data Science and Engineering - Volume 1: ICDSE; ISBN 978-989-758-765-8, SciTePress, pages 481-486. DOI: 10.5220/0013699700004670

in Bibtex Style

@conference{icdse25,
author={Haoran Gao},
title={Vehicle Detection and Tracking Based on YOLOv11},
booktitle={Proceedings of the 2nd International Conference on Data Science and Engineering - Volume 1: ICDSE},
year={2025},
pages={481-486},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0013699700004670},
isbn={978-989-758-765-8},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 2nd International Conference on Data Science and Engineering - Volume 1: ICDSE
TI - Vehicle Detection and Tracking Based on YOLOv11
SN - 978-989-758-765-8
AU - Gao H.
PY - 2025
SP - 481
EP - 486
DO - 10.5220/0013699700004670
PB - SciTePress