Abstract: In computer vision and cognitive science field an automated scene analysis has been a topic of great attention. The techniques used for behavior analysis in computer vision, are mostly targeting individuals behavior. In recent times, crowd phenomena gained the popularity in the real world, thus crowded scene analysis has attracted much attention. Most of the people get panic when unexpected events occur in public as well as private places. The modeling and analysis of the crowd surveillance videos is an issue that exists due to the unconstrained behaviors observed. In the last few years, research on crowded scene analysis is going on, covering various aspects such as crowd motion pattern learning, crowd behavior and activity analysis, and anomaly detection in crowds. In various papers the problem of crowded scene analysis is addressed. To automatically identify crowd behaviors in surveillance videos, many computer vision methods have been applied to crowd behavior analysis.
Keywords: Crowded scene analysis, Video Surveillance, Crowd Behavior, Crowd Motion, escape and non-escape activity.