Direction, Velocity, Merging Probabilities and Shape Descriptors for Crowd Behavior Analysis

Rakhshenda Javid,Naveed Iqbal Rao,M Mohsin Riaz,Abdul Ghafoor

doi:10.1109/access.2019.2929242

Rakhshenda Javid, Naveed Iqbal Rao + Show 2 more

Open Access

PDF Available

https://doi.org/10.1109/access.2019.2929242

Copy DOI

Export

Save

Cite

Abstract
Highlights/Summary
Full-Text PDF
Similar Papers

Abstract

Listen

Descriptors are important for quantifying crowd behavior. The existing descriptors generally provide information about crowd density, i.e., number of people/objects present in a defined spatial area. However, other properties of crowd like speed, direction, shape, and merging probabilities (of different crowds at group level) are also important for crowd analysis. In this paper, crowd descriptors (by mitigating the effects of outliers) are introduced which can be used for crowds having various densities. The simulations on various datasets show the applicability of the proposed descriptors.

Highlights

The most prominent characteristic of moving objects is their coherent motion
Overall the existing techniques use velocity correlations, background subtraction or synthetic data, whereas we focus on analyzing the crowd at group level
The research closest to our work is proposed in [12] but it is not directly comparable to our work since they focus on the crowd as a whole and finding the dominant motions in the crowd where as we focus on the direction of different crowd groups in a crowded scene

Summary

Introduction

The most prominent characteristic of moving objects (people, animals/birds, insects, micro-organism and vehicles etc.) is their coherent motion. Some examples of people coherent motion include human crowds at stations, shopping malls, zebra crossings etc. Crowd direction and density descriptor can locate the source and sinks of different crowd groups in a scene. The motion of different crowd groups in a scene helps in estimating their speed which in turn helps to prevent any disastrous situation e.g. road accidents. The probability of one group merging with other group can help in crowd management and disaster control. Quantitative analysis of coherently moving groups is useful in different applications (like crowd scene understanding, video classification, group segmentation, visual surveillance, homeland security, disaster prevention and behavior analysis [2])

Objectives

Results

Conclusion