Crowd Behavior Analysis: A Review

Crowd Behavior Analysis: A Review

Author Name : Dr Jay Shankar Prasad, Divya Singh, Shashi Yadav, Navneet Yadav

ABSTRACT

Crowd is a unique group of individual or something involves community or society. The phenomena of the crowd are very familiar in a variety of research discipline such as sociology, civil and physic. Nowadays, it becomes the most active-oriented research and trendy topic in computer vision. Traditionally, three processing steps involve in crowd analysis, and these include pre-processing, object detection and event/behavior recognition. Meanwhile, the common process for analysis in video sequence of crowd information extraction consists of Pre-Processing, Object Tracking, and Event/Behavior Recognition. In terms of behavior detection, the crowd density estimation, crowd motion detection, crowd tracking and crowd behavior recognition are adopted. In this paper, we give the general framework and taxonomy of pattern in detecting abnormal behavior in a crowd scene. This study presents the state of art of crowd analysis, taxonomy of the common approach of the crowd analysis and it can be useful to researchers and would serve as a good introduction related to the field undertaken.This project is a cross-platform application for analysis of   crowd. The system is capable of recognition of human faces, counts distinct people in a dynamic scene, and identifies the   expressed emotions of the people present in the scene. The system is capable to capture the image and video and present the analysis.  The convolution neural network (CNN) used here to obtain the fast and accurate response, true classification of the expressions. The positive traits of the person present in the crowd is analyzed e.g.   Whether the person is attentive or not by checking their expressions and giving the output accordingly. After identification and analysis of the people’s expressions, the label of correct expression displayed on the screen for the user. We obtain approx. 90 percent correct classification rate of the facial expressions through this project.

Keywords: Crowd analysis, pre-processing, object tracking, event behavior recognition