# Detection of Salient Crowd Motion Based on Repulsive Force Network and Direction Entropy

^{1}

^{2}

^{3}

^{*}

## Abstract

**:**

## 1. Introduction

## 2. Calculation of Crowd Velocity Vector Field

## 3. Construction of Repulsive Force Network

#### 3.1. Establishment of a Network Node

^{w}node set $Q=\{{q}_{1},{q}_{2},\cdots ,{q}_{n}\}$ can be generated, where n is the total number of nodes. The number of network nodes is equal to the number of particles in the crowd velocity field.

#### 3.2. Establishing the Network Edges Using Repulsive Force Model

#### 3.3. Calculation of Node Strength

## 4. Optimizing Node Strength Field Using Direction Entropy

#### 4.1. Establishment of Vector Direction Entropy Matrix

_{i}in all direction ranks. According to the definition of Shannon entropy [21] and [23], we can assign the velocity direction entropy between the central particle ${q}_{xo}{}_{yo}$ and other particles ${q}_{xy}$$(x\ne {x}_{0},y\ne {y}_{0})$ neighboring the central particle. The calculation is determined by the following formula:

_{11}, H

_{12}……H

_{MN}is the entropy at the corresponding position of the crowd particle field. In order to facilitate the node strength field optimization operation in later stage, the direction entropy matrix is normalized as follows:

#### 4.2. Optimizing the Node Strength Field

## 5. Experimental Results and Analysis

#### 5.1. Crowd Retrograde Behavior Detection

#### 5.2. Crowd Motion Instability Region Detection

#### 5.3. Detection Results Using Different Neighborhood Size

#### 5.4. Performance Evaluation and Comparison

## 6. Conclusions

## Author Contributions

## Funding

## Conflicts of Interest

## References

- Challenger, R.; Clegg, C.W.; Robinson, M.A. Understanding crowd behaviours: Supporting evidence. In Understanding Crowd Behaviours; Leigh, M., Ed.; Emergency Planning College: Easingwold, UK, 2009; pp. 1–326. [Google Scholar]
- Zhang, X.; Shu, X.; He, Z. Crowd panic state detection using entropy of the distribution of enthalpy. Phys. A Stat. Mech. Appl.
**2019**, 525, 935–945. [Google Scholar] [CrossRef] - Wixson, L. Detecting salient motion by accumulating directionally-consistent flow. IEEE Trans. Pattern Anal. Mach. Intell.
**2002**, 22, 774–780. [Google Scholar] [CrossRef] - Lim, M.K.; Chan, C.S.; Monekosso, D.; Remagnino, P. Detection of salient regions in crowded scenes. Electron. Lett.
**2014**, 50, 363–365. [Google Scholar] [CrossRef] [Green Version] - Lim, M.K.; Kok, V.J.; Loy, C.C.; Chan, C.S. Crowd Saliency Detection via Global Similarity Structure. In Proceedings of the 22nd International Conference on Pattern Recognition, Stockholm, Sweden, 24–28 August 2014; pp. 3957–3962. [Google Scholar]
- Loy, C.; Xiang, T.; Gong, S. Salient motion detection in crowded scenes. In Proceedings of the 5th International Symposium on Communications, Control and Signal Processing (ISCCSP), Rome, Italy, 2–4 May 2012; pp. 1–4. [Google Scholar]
- Jian, M.; Zhag, W.; Yu, H.; Cui, C.; Nie, X.; Zhang, H.; Yin, Y. Saliency detection based on directional patches extraction and principal local color contrast. J. Vis. Commun. Image Represent.
**2018**, 57, 1–11. [Google Scholar] [CrossRef] [Green Version] - Zhou, B.; Tang, X.; Wang, X. Coherent Filtering: Detecting Coherent Motions from crowd clutters. In Proceedings of the 12th European Conference on Computer Vision (ECCV), Firenze, Italy, 7–13 October 2012; Volume 7573, pp. 857–871. [Google Scholar]
- Solmaz, B.; Moore, B.E.; Shah, M. Identifying behaviors in crowd scenes using stability analysis for dynamical systems. PAMI
**2012**, 34, 2064–2070. [Google Scholar] [CrossRef] [PubMed] - Zhang, X.; Yu, Q.; Yu, H. Physics Inspired Methods for Crowd Video Surveillance and Analysis: A Survey. IEEE Access
**2018**, 6, 66816–66830. [Google Scholar] [CrossRef] - Hao, Y.; Xu, Z.J.; Liu, Y.; Wang, J.; Fan, J.-L. Effective Crowd Anomaly Detection Through Spatio-temporal Texture Analysis. Int. J. Autom. Comput.
**2019**, 16, 27–39. [Google Scholar] [CrossRef] - Kanade, L. An Iterative Image Registration Technique with an Application to Stereo Vision. In Proceedings of the 7th International Joint Conference on Artificial Intelligence, Vancouver, BC, Canada, 24–28 August 1981; pp. 674–679. [Google Scholar]
- Bouguet, J.Y. Pyramidal implementation of the Lucas Kanade feature tracker description of the algorithm. OpenCV Doc.
**1999**, 22, 363–381. [Google Scholar] - Chen, G.; Wang, X.; Li, X. Introduction to Complex Networks: Models. Structures and Dynamics; Higher Education Press: Beijing, China, 2012. [Google Scholar]
- Joldos, M.; Technical, C.C. A parallel evolutionary approach to community detection in complex networks. In Proceedings of the IEEE International Conference on Intelligent Computer Communication and Processing IEEE, Cluj-Napoca, Romania, 7–9 September 2017; pp. 247–254. [Google Scholar]
- Boccaletti, S.; Latora, V.; Moreno, Y. Complex networks: Structure and dynamics. Phys. Rep.
**2006**, 424, 175–308. [Google Scholar] [CrossRef] - Dzwinel, W.; Yuen, D.A.; Boryczko, K.B. Diverse physical scales with the discrete-particle paradigm in modeling colloidal dynamics with mesoscopic features. Chem. Eng. Sci.
**2006**, 61, 2169–2185. [Google Scholar] [CrossRef] - Mohcine, C.; Seyfried, A.; Schadschneider, A. Generalized Centrifugal Force Model for Pedestrian Dynamics. Phys. Rev.
**2010**, 82, 046111. [Google Scholar] [CrossRef] - Zou, Q.; Sun, W.; Xing, L. Interest Point Detection in Images Based on Topology Structure Features of Directed Complex Network IEEE. In Proceedings of the 36th China Control Conference, Dalian, China, 26–28 July 2017; pp. 1507–1511. [Google Scholar]
- Wu, Z.; Lu, X.; Deng, Y. Image edge detection based on local dimension: A complex networks approach. Phys. A Stat. Mech. Appl.
**2015**, 440, 9–18. [Google Scholar] [CrossRef] - Shannon, C.E. A mathematical theory of communication. Bell Syst. Tech.
**1948**, 27, 379–423. [Google Scholar] [CrossRef] - Shannon, C.E.; Wyner, S.A. (Eds.) Collected Papers; IEEE Press: New York, NY, USA, 1993. [Google Scholar]
- Tribus, M.; McIrvine, E.C. Energy and information. Sci. Am.
**1971**, 225, 179–188. [Google Scholar] [CrossRef] - Mehran, R.; Oyama, A.; Shah, M. Abnormal crowd behavior detection using social force model. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, Miami, FL, USA, 20–25 June 2009; pp. 935–942. [Google Scholar]
- Goutte, C.; Gaussier, E. A Probabilistic Interpretation of Precision, Recall and F-Score, with Implication for Evaluation. Int. J. Radiat. Biol. Relat. Stud. Phys. Chem. Med.
**2005**, 51, 952. [Google Scholar] - Ali, S.; Shah, M. A Lagrangian Particle Dynamics Approach for Crowd Flow Segmentation and Stability Analysis. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, Minneapolis, MN, USA, 17–22 June 2007; pp. 1–6. [Google Scholar]

**Figure 1.**Framework of salient crowd motion detection based on repulsive force network and direction entropy.

**Figure 4.**Expressing the effect of optical flow reversal: (

**a**) original sample frame; (

**b**) detection result using original optical flow; (

**c**) detection result using reversed optical flow; (

**d**) detection result after the integration of optical flow.

**Figure 5.**Retrograde motion detection in train station scene: (

**a**) input frame; (

**b**) node strength field of repulsive force network; (

**c**) detection result using direction entropy; (

**d**) salient region detection after optimized; (

**e**) overlap the salient region with input frame.

**Figure 6.**Retrograde motion detection in single retrograde scene: (

**a**) input frame; (

**b**) node strength field of repulsive force network; (

**c**) detection result using direction entropy; (

**d**) salient region detection after optimized; (

**e**) overlap the salient region with input frame.

**Figure 7.**Salient crowd instability motion detection in marathon scene: (

**a**) original video frame and ground true (red box); (

**b**) node strength field of repulsive force network; (

**c**) detection result using direction entropy; (

**d**) salient region detection after optimized; (

**e**) overlap the salient region with original video frame.

**Figure 8.**Salient crowd instability motion detection in pilgrimage scene: (

**a**) original video frame and ground true (red box); (

**b**) node strength field of repulsive force network; (

**c**) detection result using direction entropy; (

**d**) salient region detection after optimized; (

**e**) overlap the salient region with original video frame.

**Figure 9.**Retrograde motion detection in train station scene using different neighborhood size: (

**a**) original video frame; (

**b**) detection result using 5 × 5 neighborhood; (

**c**) detection result using 13 × 13 neighborhood; (

**d**) detection result using 23 × 23 neighborhood.

**Figure 10.**Retrograde motion detection in single retrograde scene using different neighborhood size: (

**a**) original video frame; (

**b**) detection result using 5 × 5 neighborhood; (

**c**) detection result using 15 × 15 neighborhood; (

**d**) detection result using 23 × 23 neighborhood.

**Figure 11.**Instability motion detection in marathon scene using different neighborhood size: (

**a**) original video frame and ground truth region; (

**b**) detection result using 5 × 5 neighborhood; (

**c**) detection result using 11 × 11 neighborhood; (

**d**) detection result using 23 × 23 neighborhood.

**Figure 12.**Instability motion detection in pilgrimage scene using different neighborhood size: (

**a**) original video frame and ground truth region; (

**b**) detection result using 5 × 5 neighborhood; (

**c**) detection result using 15 × 15 neighborhood; (

**d**) detection result using 25 × 25 neighborhood.

Crowded Scenes | Symbol of Parameter | The Value |
---|---|---|

Train station scene in Figure 5 | ε M × N | 13 480 × 360 |

Single retrograde scene in Figure 6 | ε M × N | 15 480 × 360 |

Marathon scene in Figure 7 | ε M × N | 11 640 × 480 |

Pilgrimage scene in Figure 8 | ε M × N | 15 640 × 480 |

Crowded Scenes | Statistics | Size of Neighborhood | Results |
---|---|---|---|

marathon | Pr | 5 × 5 | 0.862 |

11 × 11 | 0.910 | ||

23 × 23 | 0.531 | ||

R | 5 × 5 | 0.841 | |

11 × 11 | 0.909 | ||

23 × 23 | 0.877 | ||

pilgrimage | Pr | 5 × 5 | 1 |

15 × 15 | 1 | ||

25 × 25 | 0.684 | ||

R | 5 × 5 | 0.244 | |

15 × 15 | 0.867 | ||

25 × 25 | 0.656 |

© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

## Share and Cite

**MDPI and ACS Style**

Zhang, X.; Lin, D.; Zheng, J.; Tang, X.; Fang, Y.; Yu, H.
Detection of Salient Crowd Motion Based on Repulsive Force Network and Direction Entropy. *Entropy* **2019**, *21*, 608.
https://doi.org/10.3390/e21060608

**AMA Style**

Zhang X, Lin D, Zheng J, Tang X, Fang Y, Yu H.
Detection of Salient Crowd Motion Based on Repulsive Force Network and Direction Entropy. *Entropy*. 2019; 21(6):608.
https://doi.org/10.3390/e21060608

**Chicago/Turabian Style**

Zhang, Xuguang, Dujun Lin, Juan Zheng, Xianghong Tang, Yinfeng Fang, and Hui Yu.
2019. "Detection of Salient Crowd Motion Based on Repulsive Force Network and Direction Entropy" *Entropy* 21, no. 6: 608.
https://doi.org/10.3390/e21060608