Sage Journals: Discover world-class research

Abstract

The cluster behavior has the characteristics of large spatiotemporal scale, multi-source heterogeneous information, and complex target features. Most of the existing cluster behavior understanding methods are based on the analysis of monitoring images, and their scene definition is narrow. The real scene not only contains the data information collected by the surveillance camera but also the knowledge information outside the scope of monitoring, such as the experience of police officers and relevant rules. Simultaneously considering the data and knowledge not only reduces the misjudgment but also reduces the processing task of the machine and improves the recognition speed. In this paper, data and knowledge are integrated to conduct scene modeling, which can describe the scene more comprehensively and accurately. Based on the scene information element constructed by multi-source heterogeneous information, the cluster behavior understanding system is built to improve the speed and accuracy of cluster behavior identification.

Keywords

cluster behavior scene information behavior understanding public safety

Get full access to this article

View all access options for this article.

References

Xue

Liu

, et al. Crowd scene analysis encounters high density and scale variation. IEEE Trans Image Proc 2021; 30: 2745–2757.

Rezaei

Yazdi

. Real-time crowd behavior recognition in surveillance videos based on deep learning methods. J Real-Time Image Proc 2021; 18: 1669–1679.

Yan

Xie

Tang

, et al. HiGCIN: Hierarchical graph-based cross inference network for group activity recognition. IEEE Trans Pattern Anal Mach Intell 2020; 45(6): 6955–6968.

Yuan

Wang

. Spatio-temporal dynamic inference network for group activity recognition. In: Proceedings of the IEEE/CVF International Conference on Computer Vision, Montreal, BC, Canada, 11–17 October 2021, 2021, pp. 7476–7485.

Mao

Jin

Ping

, et al. Modeling multi-scale sub-group context for group activity recognition. Appl Intell 2023; 53(1): 1149–1161.

Xiaobin

Weishi

Jianguo

. Learning person-person interaction in collective activity recognition. IEEE Trans Image Process 2015; 24(6): 1905–1918.

Ramanujam

Perumal

Padmavathi

. Human activity recognition with smartphone and wearable sensors using deep learning techniques: A review. IEEE Sensor J 2021; 21(12): 13029–13040.

Cui

, et al. Progressive relation learning for group activity recognition. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, Seattle, WA, USA, 13–19 June 2020, 2020, pp. 980–989.

Tran

Gala

Kakadiaris

, et al. Activity analysis in crowded environments using social cues for group discovery and human interaction modeling, Pattern Recogn. Letture 2013; 44: 49–57.

10.

Chellappa

Zhou

. Recognizing interactive group activities using temporal interaction matrices and their Riemannian statistics. Int J Comput Vis 2012; 101(2): 305–328.

11.

Liu

Wang

Gong

, et al. Deep fully connected model for collective activity recognition. IEEE Access 2019; 7: 104308–104314.

12.

Cui

, et al. Progressive relation learning for group activity recognition. In: CVPR, Seattle, WA, USA, 16–18 June 2020, 2020, p. 19874668.

13.

Direkoglu

. Abnormal crowd behavior detection using motion information images and convolutional neural networks. IEEE Access 2020; 8: 80408–80416.

14.

Bendali-Braham

Weber

Forestier

, et al. Recent trends in crowd analysis: A review. Machine Learning with Applications 2021; 4: 100023.

The understanding of cluster behavior based on scene information element

Abstract

Keywords

Get full access to this article

References