МЕТОДИ АВТОМАТИЧНОГО ВИЯВЛЕННЯ СТОРОННІХ ОБ’ЄКТІВ У ВІДЕОСЦЕНІ З ДИНАМІЧНИМИ ПЕРЕШКОДАМИ

P. SABELNIKOV

doi:10.15407/scine21.05.062

Authors

P. SABELNIKOV V.M. Glushkov Institute of Cybernetics of the National Academy of the Sciences of Ukraine https://orcid.org/0000-0003-4709-8693

DOI:

https://doi.org/10.15407/scine21.05.062

Keywords:

Keywords: video systems, video surveillance systems, video fragment, video sequence.

Abstract

Introduction. The automatic detection of foreign and abandoned objects plays a critical role in modern video
surveillance systems, particularly in areas with high pedestrian traffic, where such capabilities are essential for the prevention and investigation of security threats and terrorist activities.
Problem Statement. Video surveillance systems generate vast volumes of data, making it impractical for human operators to monitor all video streams continuously. To mitigate this challenge, automated video analytics
systems have been developed to support or replace manual observation. Since human operators may miss subtle
changes in complex video scenes, there is a clear need for robust algorithms capable of real-time scene analysis.
Purpose. This study aims to develop methods for the automatic detection of foreign objects in dynamic video
environments, enabling early threat identification and supporting security operations at public institutions and
critical infrastructure facilities.
Materials and Methods. Experimental validation has been performed using the PETS 2006 benchmark dataset, which includes seven scenarios of increasing complexity and is widely used to evaluate algorithms for detecting abandoned and removed objects in public environments.
Results. A novel method has been developed that is resilient to changes in lighting conditions and minor camera displacements. Experimental studies have demonstrated that the proposed algorithm reliably detects foreign
objects in dynamic video scenes under varying illumination and scene perturbations. The method has shown stable
performance with minimal feature extraction, achieving consistent results on PETS 2006 video sequences.
Conclusions. The proposed approach enables real-time operation of video surveillance systems, significantly
reducing the cognitive load on operators by automatically detecting potentially dangerous abandoned objects and
issuing timely alerts, thereby enhancing situational awareness and public safety.

Downloads

Download data is not yet available.

References

Patryk, S., Vun, Yu. V., Komly, R. (2011). Performance in dicators for videotracking. IETE Technical review, 28(6), 493. https://doi.org/10.4103/0256-4602.90759 . URL: https://www.researchgate.net/publication/272646487_Performance_ Evaluation_Metrics_for_Video_Tracking (Last accessed: 20.06.2023).

Makfarlein, N. Dzh. B., Shofyld, K. P. (1995). Segmentation and tracking of pigletsinimages. Machine vision and applications, 8, 187—193. URL: https://link.springer.com/article/10.1007/BF01215814 (Last accessed: 20.06.2023).

Kukkyara, R., Hrana, K., Praty, A., Vetstsany, R (2005). Probabilistic pose classification for analyzing human behavior, transactions in systems. IEEE Transactionson Systems, Man, and Cybernetics, 35(1), 42—54. https://doi.org/10.1109/TSMCA.2004.838501

Huwer, S., Niemann, H. (2000). Adaptive change detection for real-time surveillanceapplications. Proceedings Third IEEE International Workshop on Visual Surveillance, 37—46. URL: https://ieeexplore.ieee.org/document/856856 (Last accessed: 20.06.2023). https://doi.org/10.1109/VS.2000.856856

Khalevy, H., Vainsholl, D. (1999). Disturbance motion: detecting and tracking non-rigid multi-body motion. Machine Vision and Applications, 11, 122—137. URL: https://link.springer.com/article/10.1007/s001380050096 (Last accessed: 20.06.2023).

Mokhamad, A. A. Kh., Osman, M. (2013). Adaptive median filter background subtraction technique using fuzzy logic. International Conference on Computer, Electrical and Electronic Engineering (ICCEEE) (Khartoum, Sudan, 2013), 115—120. https://doi.org/10.1109/ICCEEE.2013.6633917

Haritaoglu, I., Harwood, D., Davis, L. S. (2000). W4: Real-time monitoring of people and their actions, pattern analysis and machine intelligence. IEEE Trans. on Pattern Analysis and Machine Intelligence, 8(22), 809—830. URL: https://3dvar. com/Haritaoglu2000W4.pdf (Last accessed: 20.06.2023). https://doi.org/10.1109/34.868683

Boult, T., Micheals, R. J., Gao, X., Eckmann, M. (2001). Into the woods: Visual surveillance of non-cooperative and camo ufl aged targets in compex outdoor settings. Proceedings of the IEEE, 89(10), 1382—1402. https://doi.org/10.1109/5.959337

Abbottand, R., Williams, L. (2009). Multiple target tracking with lazy background subtraction and connected components analysis. Machine Vision and Applications, 20, 93—101. https://doi.org/10.1007/s00138-007-0109-8

Harville, M. (2002). A framework for high-level feedback to adaptive, per-pixel, mixture-of-Gaussian background models. In: Heyden, A., Sparr, G., Nielsen, M., Johansen, P. (Eds.) Computer Vision — ECCV 2002. Lecture Notes in Computer Science, 2352, 28—31. Springer, Berlin, Heidelberg. https://doi.org/10.1007/3-540-47977-5_36

Wang, J. X., Bebis, G. N., Miller, R. (2006). Robust video-based surveillance by integrating target detection with tra cking. IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops (CVPRW) (New York, NY, USA, 2006), 137—137. https://doi.org/10.1109/CVPRW.2006.180

Venetianer, P., Zhang, Z., Yin, W., Lipton, A. (2007). Stationary target detection using the object video surveillance system. Advanced Video and Signal-based Surveillance, 2007. https://doi.org/10.1109/AVSS.2007.4425317

Yaoand, J., Odobez, J.-M. (2007). Multi-Layer Background Subtraction Based on Color and Texture. IEEE Conference on Computer Vision and Pattern Recognition (Minneapolis, MN, USA, 2007), 1—8. https://doi.org/10.1109/CVPR. 2007.383497

Taycher, L., Fisher, J. W., Darrell, T. (2005). Incorporating Object Tracking Feedback into Background Maintenance Framework, (2005 Seventh IEEE Workshops on Applications of Computer Vision (WACV/MOTION’05), Volume 1 (Breckenridge, CO, USA, 2005), 120—125. https://doi.org/10.1109/ACVMOT.2005.63

Hlumov, N. Y., Miasnykov, V. V., Serheev V. V. (2010). Detection and recognition of objects in images: electron. Textbook al lowance. Samara. https://repo.ssau.ru/handle/Uchebnye-posobiya/Obnaruzhenie-i-raspoznavanie-obektov-naizobrazheniyah-Elektronnyi-resurs-elektron-ucheb-posobie-54634 (Last accessed: 20.06.2023).

Sabelnikov, P., Sabelnikov, Yu. (2023). Search for identical points in the inter-pixel space of videoimages. Artificial Intelligence, 28(1), 155—162. https://doi.org/10.15407/jai2023.01.155

Sabelnikov, P. Iu., Sabelnikov, Yu. A., Syniakov, M. M. (2023). Determining the orientation of circularareas in video images. Artificial Intelligence, 28(3), 131—138. https://doi.org/10.15407/jai2023.03.131 [in Ukrainian].

YOLOv4: Highspeed and accurate object detection. URL: https://docs.ultralytics.com/ru/models/yolov4/ (Last accessed: 12.06.2023).

Tensor Flow. URL: https://uk.wikipedia.org/wiki/TensorFlow (Last accessed: 23.12.2021).

Dataset PETS 2006. URL: https://www.researchgate.net/figure/PETS-2006-Dataset-Sequence-1-a-original-image-bbackground-patches-that-can-not-be_fig3_236630140 (Last accessed: 23.12.2021).