Development and research of software for detection and tracking of fast-moving objects
Abstract
The article is devoted to the development and research of software for detecting and tracking fast-moving objects in a video sequence. The paper considers the main tasks of detection and tracking and provides an overview of existing architectures and algorithms used to improve the accuracy and speed of data processing. Special attention is paid to the application of a Siamese neural network for assessing the visual similarity of objects and validating detections during the tracking process.
Within the framework of the study, software was developed that implements modules for detection, validation of detections using a Siamese neural network, and tracking. Experimental testing and a comparative analysis of the proposed approach with traditional methods based on handcrafted features were conducted. The quality of recognition was evaluated using precision and recall metrics, and the performance of multi-object tracking was assessed using the MOTA and IDF1 metrics. The results of the experiments confirm the superiority of the neural network approach: the object detection accuracy reached 0,84 with a recall of 0,87. The obtained results show higher robustness to variations in the appearance of objects compared to classical methods. The use of a Siamese network made it possible to reduce the number of object losses and identity switch errors, which is reflected in the MOTA (0,64) and IDF1 (0,73) values. The developed approach demonstrates high efficiency in solving the problems of detecting and tracking fast-moving objects and confirms the promise of using Siamese neural networks for the further development of intelligent computer vision systems
Full Text:
PDF (Russian)References
Cui Y. et al. Mixformer: End-to-end tracking with iterative mixed attention // Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. – 2022. – P. 13608–13618.
Ye B. et al. Joint feature learning and relation modeling for tracking: A one-stream framework // European conference on computer vision. – Cham : Springer Nature Switzerland, 2022. – P. 341–357.
Meinhardt T. et al. Trackformer: Multi-object tracking with transformers // Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. – 2022. – P. 8844–8854.
Zeng F. et al. Motr: End-to-end multiple-object tracking with transformer // European conference on computer vision. – Cham : Springer Nature Switzerland, 2022. – P. 659–675.
Kadam P., Fang G., Zou J. J. Object tracking using computer vision: A review // Computers. – 2024. – Т. 13. – № 6. – P. 136.
Zaidi S. S. A. et al. A survey of modern deep learning based object detection models // Digital Signal Processing. – 2022. – Т. 126. – P. 103514.
Abouelyazid M. Comparative evaluation of sort, deepsort, and bytetrack for multiple object tracking in highway videos // International Journal of Sustainable Infrastructure for Cities and Societies. – 2023. – Т. 8. – № 11. – P. 42–52.
Adžemović M. Deep Learning-Based Multi-Object Tracking: A Comprehensive Survey from Foundations to State-of-the-Art // arXiv preprint arXiv:2506.13457. – 2025.
Cheng Y. et al. Improved particle filter algorithm for multi-target detection and tracking // Sensors (Basel, Switzerland). – 2024. – Т. 24. – № 14. – P. 4708.
Javed S. et al. Visual object tracking with discriminative filters and siamese networks: a survey and outlook // IEEE transactions on pattern analysis and machine intelligence. – 2022. – Т. 45. – №. 5. – P. 6552–6574.
Lim S. C., Huh J. H., Kim J. C. Siamese Trackers Based on Deep Features for Visual Tracking // Electronics. – 2023. – Т. 12. – № 19. – P. 4140.
Chicco D. Siamese neural networks: An overview // Artificial neural networks. – 2021. – P. 73–94.
Xie J., Guo L. Siamese Network for Similar Data Recognition Filtering // Proceedings of the 2024 7th International Conference on Big Data Technologies. – 2024. – P. 81–85.
Vieira L. Comparing Performance of Preprocessing Techniques for Traffic Sign Recognition Using a HOG-SVM // arXiv preprint arXiv:2504.09424. – 2025.
Siamese neural network for few-shot visual inspection of steel surface defects in predictive maintenance systems / E. Antonyants, A. Aletdinova, V. Zaozernov, A. Leontiev, A. Vasiliev. – DOI 10.1109/APEIE66761.2025.11289355. – Text : electronic // 17 International scientific and technical conference actual problems of electronic instrument engineering (APEIE–2025) : proc., Novosibirsk, 14–16 Nov. 2025. – Novosiblrsk : Publ. NSTU, 2025. – 6 p. – URL: https://ieeexplore.ieee.org/document/11289355 (access date: 21.02.2026).
Sharma D., Singh B. K. Smart CCTV detection using local binary pattern histogram (LBPH) // Int Res J Adv Sci Hub. – 2023. – Т. 5. – № 08. – P. 290–298.
Refbacks
- There are currently no refbacks.
Abava Кибербезопасность Monetec 2026 СНЭ
ISSN: 2307-8162