Abstract
It is important to estimate the number of people in fast moving crowd scenarios for the surveillance systems. Regression-based techniques achieved promising results for counting the number of people in crowded scenes. However, appearance features used in the most existing techniques are not able to mirror the motion state of crowds. The motion state of crowds is important to solve the counting problem of fast moving crowds, since it is decided by the amount of people. In this study, we propose a novel method to address this problem from three perspectives: (1) train a crowd counting estimation model suited to all of the crowded scenes; (2) combine motion states with multiappearance features for crowd counting; and (3) count fast moving crowds in unconstrained videos. These ideas are implemented in a fast object segment framework, which can segment fast moving crowds in the unconstrained videos. Extensive experiments validate the effectiveness of our proposed method.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Loy CC, Chen K, Gong S, Xiang T (2013) Crowd counting and profiling: methodology and evaluation. In: Modeling, Simulation and Visual Analysis of Crowds, vol 11. Springer, pp 347–382
Ma Z, Yu L, Chan AB (2015)Small instance detection by integer programming on object density maps. In: CVPR
Papazoglou A, Ferrari V (2013) Fast object segmentation in unconstrained video. In: Proceedings of the IEEE International Conference on Computer Vision (ICCV), p 1777–1784
Solmaz B, Moore BE, Shah M (2012) Identifying behaviors in crowd scenes using stability analysis for dynamical systems. IEEE Trans Pattern Anal Mach Intell 34(10):2064–2070
Chan Antoni B, Vasconcelos Nuno (2012) Counting people with low-level features and Bayesian regression. IEEE Trans Image Process Publ IEEE Signal Process Soc 21(4):2160–2177
Chen K, Change Loy C, Gong S, Xiang T (2012). Feature mining for localised crowd counting. In: BMVC
Change Loy C, Gong S, Xiang T (2013) From semi-supervised to transfer counting of crowds. In: IEEE International Conference on Computer Vision, pp 2256–2263
Hughes RL (2000) The flow of large crowds of pedestrians. Math Comput Simul 53:367–370
Wang Lu, Lin Gao, Lei Yan et al (2012) Key-frame retrieval method based on optical flow and entropy statistic for blooming video. Trans Chin Soc Agric Eng (Trans CSAE) 28(17):125–130
Chan AB, Liang Z-SJ, Vasconcelos N (2008) Privacy preserving crowd monitoring: counting people without people models or tracking. In: IEEE Conference on Computer Vision and Pattern Recognition, pp 1–7
Keerthi S, DeCoste D (2005) A modified finite Newton method for fast solution of large scale linear SVM. JMLR 6(1):341–361
Melacci S, Belkin M (2011) Laplacian support vector machines trained in the primal. JMLR 2(4):1149–1184
Acknowledgments
This work was supported by National Basic Research Program of China (973 Program) 2012CB821200 (2012CB821206) and the National Natural Science Foundation of China (No. 61320106006, No. 61532006, No. 61502042).
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2016 Springer Science+Business Media Singapore
About this paper
Cite this paper
Wei, X., Du, J., Fan, D., Lee, J. (2016). Fast Moving Crowd Counting for Unconstrained Videos. In: Jia, Y., Du, J., Zhang, W., Li, H. (eds) Proceedings of 2016 Chinese Intelligent Systems Conference. CISC 2016. Lecture Notes in Electrical Engineering, vol 404. Springer, Singapore. https://doi.org/10.1007/978-981-10-2338-5_46
Download citation
DOI: https://doi.org/10.1007/978-981-10-2338-5_46
Published:
Publisher Name: Springer, Singapore
Print ISBN: 978-981-10-2337-8
Online ISBN: 978-981-10-2338-5
eBook Packages: Computer ScienceComputer Science (R0)