Skip to main navigation menu Skip to main content Skip to site footer

Electrotechnical and Computer Engineering

Vol. 35 No. 11 (2020): Proceedings of the Faculty of Technical Sciences

TRACKING OF MOVING OBJECTS IN VIDEO USING PASSIVE FILTER AND HIGH PERFORMANCE DETECTOR

  • Jelena Mrđen
DOI:
https://doi.org/10.24867/10BE29Mrdjen
Submitted
November 4, 2020
Published
2020-11-04

Abstract

The paper describes the problem of tracking moving objects in video and implements system components based on passive filtering of detections in each frame.  Through numerical experiments based on DETRAC evaluation protocol the best parameters for the configuration of IOU algorithm and YOLOv3 object detector were determined.

References

[1] J. Redmon and A. Fahradi, “YOLOv3: An Incremental Improvement“, 2018.
[2] E. Bochinski, V. Eiselein and T. Sikora, “High-Speed Tracking-by-Detection Without Using Image Information”, 2017.
[3] L. Wena et al., “UA-DETRAC: A New Benchmark and Protocol for Multi-Object Detection and Tracking”, 2020.
[4] "The UA-DETRAC Benchmark Suite", Detrac-db.rit.albany.edu, 2020. [Online]. Dostupno na: http://detrac-db.rit.albany.edu/download.[Pristupljeno: 08- Sep- 2020].