EUVIMEDEUVIMED entdecken
Crossref

Enhanced vehicle detection via YOLOv7-tiny: Advancements in feature extraction and bounding box regression

Xiaole Ge, Zhaofeng Chen, Feng Zhou, Binghe Sun, Shuaiting Chen, Gan Gao, Rugang Wang

Proceedings of the Institution of Mechanical Engineers, Part D: Journal of Automobile Engineering · 2026-05 · 240 (6) · 3485-3499

DOI: 10.1177/09544070251348698 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Proceedings of the Institution of Mechanical Engineers, Part D: Journal of Automobile Engineering, online erschienen am 23.07.2025. Verfasst von Xiaole Ge und 6 weiteren Autor:innen. Die Arbeit wurde bisher 3-mal zitiert (Crossref) und nennt 44 Literaturangaben. Gefördert von: National Natural Science Foundation of China, Graduate Research and Innovation Projects of Jiangsu Province, Joint Project of Industry-University-Research of Jiangsu Province.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

With the proliferation of intelligent transportation systems, vehicle detection plays a pivotal role in autonomous driving and traffic management. However, real-time vehicle detection faces challenges, particularly in dense traffic scenarios with overlapping and occluded vehicles, small targets at long distances, and variable shapes of large vehicles. To address these issues, this paper presents improvements to the YOLOv7-tiny algorithm. Firstly, we introduce the Sim_DFC feature extraction module, which combines multi-scale feature extraction with an attention mechanism to enhance recognition accuracy for large and special vehicles. Secondly, we optimize object box regression by integrating Shape IoU with Inner IoU, improving the detection of overlapping vehicles and small targets. Finally, we adopt an improved BiFPN structure to enhance multi-scale feature fusion without increasing computational complexity. Experimental results on the UA-DETRAC and Visdrone2019 datasets demonstrate that our proposed method significantly improves mAP50 accuracy by 9.8% and 1.8%, respectively. Meanwhile, the model maintains real-time detection performance with competitive FPS and achieves higher precision and recall compared to the baseline model. The detection effect graphs further validate the reliability and feasibility of our approach across various traffic scenarios.

Förderung

  • National Natural Science Foundation of China
  • Graduate Research and Innovation Projects of Jiangsu Province
  • Joint Project of Industry-University-Research of Jiangsu Province
  • Jiangsu Province Higher Education Teaching Reform Research Project Funded Project
  • Research on the Teaching Reform of High-quality Public Courses in Jiangsu Universities’ funded project
  • Yancheng institute of technology high-level talent research initiation project
  • Top-notch Academic Programs Project of Jiangsu Higher Education Institutions

Zitieren

APA

Ge, X., Chen, Z., Zhou, F., Sun, B., Chen, S., Gao, G., & Wang, R. (2026). Enhanced vehicle detection via YOLOv7-tiny: Advancements in feature extraction and bounding box regression. Proceedings of the Institution of Mechanical Engineers, Part D: Journal of Automobile Engineering, 240(6), 3485-3499. https://doi.org/10.1177/09544070251348698

Vancouver

Ge X, Chen Z, Zhou F, Sun B, Chen S, Gao G, et al. Enhanced vehicle detection via YOLOv7-tiny: Advancements in feature extraction and bounding box regression. Proceedings of the Institution of Mechanical Engineers, Part D: Journal of Automobile Engineering. 2026;240(6):3485-3499. doi:10.1177/09544070251348698

BibTeX
@article{Ge2026,
  author = {Ge, Xiaole and Chen, Zhaofeng and Zhou, Feng and Sun, Binghe and Chen, Shuaiting and Gao, Gan and Wang, Rugang},
  title = {Enhanced vehicle detection via YOLOv7-tiny: Advancements in feature extraction and bounding box regression},
  journal = {Proceedings of the Institution of Mechanical Engineers, Part D: Journal of Automobile Engineering},
  year = {2026},
  volume = {240},
  number = {6},
  pages = {3485--3499},
  doi = {10.1177/09544070251348698}
}
Kostenlos starten →