EUVIMEDEUVIMED entdecken
Crossref

An Efficient YOLOv8–DeepSORT Framework for Real-Time Multi-Object Video Surveillance

N. Devakirubai, G. Kannan, G. Archana, S. Bhuvaneshwari

International Journal of Engineering & Extended Technologies Research · 28.03.2026 · 08 (02)

DOI: 10.15662/ijeetr.2026.0802224 ↗

Kurz gefasst

Fachartikel in der Zeitschrift International Journal of Engineering & Extended Technologies Research, online erschienen am 28.03.2026. Verfasst von N. Devakirubai und 3 weiteren Autor:innen.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Video surveillance systems play a vital role in maintaining safety and security across diverse environments such as educational institutions, transportation hubs, commercial infrastructures, and smart city ecosystems. The exponential increase in the deployment of closed-circuit television (CCTV) cameras has led to the generation of large-scale video data, making continuous manual monitoring inefficient, labor-intensive, and prone to human errors. In addition, existing automated surveillance systems predominantly focus on object detection without maintaining object identities across consecutive frames, resulting in unstable tracking, identity switching, and reduced situational awareness in dynamic environments.To overcome these challenges, this paper presents a robust and scalable AI-based real-time video surveillance framework that integrates the YOLOv8 object detection model with the DeepSORT multi-object tracking algorithm. The proposed approach leverages the high-speed and accurate detection capabilities of YOLOv8 to identify objects of interest, while DeepSORT enhances tracking performance by preserving object identities using motion estimation through Kalman filtering and appearance-based feature matching. This combined framework enables reliable detection and continuous tracking of multiple objects across video frames, even in complex scenarios involving occlusions, crowded scenes, and varying illumination conditions.The system is implemented using Python and OpenCV, ensuring flexibility, ease of deployment, and cost-effectiveness for real-world applications. Extensive experiments are conducted on benchmark datasets and real-time video sequences to evaluate the performance of the proposed model. The results demonstrate that the system achieves an accuracy of 97.62%, precision of 97.53%, recall of 99.94%, and an F1-score of 98.72%, while maintaining real-time processing speed. Furthermore, the integration of DeepSORT significantly reduces identity switching and enhances tracking stability compared to conventional detection-only approaches Overall, the proposed framework provides an efficient, reliable, and scalable solution for intelligent video surveillance, making it suitable for practical deployment in security-critical applications such as traffic monitoring, public safety, and smart surveillance systems

Zitieren

APA

Devakirubai, N., Kannan, G., Archana, G., & Bhuvaneshwari, S. (2026). An Efficient YOLOv8–DeepSORT Framework for Real-Time Multi-Object Video Surveillance. International Journal of Engineering & Extended Technologies Research, 08(02). https://doi.org/10.15662/ijeetr.2026.0802224

Vancouver

Devakirubai N, Kannan G, Archana G, Bhuvaneshwari S. An Efficient YOLOv8–DeepSORT Framework for Real-Time Multi-Object Video Surveillance. International Journal of Engineering & Extended Technologies Research. 2026;08(02). doi:10.15662/ijeetr.2026.0802224

BibTeX
@article{Devakirubai2026,
  author = {Devakirubai, N. and Kannan, G. and Archana, G. and Bhuvaneshwari, S.},
  title = {An Efficient YOLOv8–DeepSORT Framework for Real-Time Multi-Object Video Surveillance},
  journal = {International Journal of Engineering & Extended Technologies Research},
  year = {2026},
  volume = {08},
  number = {02},
  doi = {10.15662/ijeetr.2026.0802224}
}
Kostenlos starten →