A Surveillance System for the Recognition of Intent within Individuals and Crowds

Abstract

Differentiating between normal human activity and aberrant behavior via closed circuit television cameras is a difficult and fatiguing task. The vigilance required of human observers when engaged in such tasks must remain constant, yet attention falls off dramatically over time. In this paper we propose an automated system to monitor video sensors and tag aberrant human activities for immediate review by human monitors. From the psychological perspective, isolated human motion depicted by point-light walker (PLW) displays have been shown to be salient for recognition of action [17] and determination of emotional state [18]. We propose that by using the motion data that immediately precedes hostile behavior, it may be possible to classify hostile intent before destructive actions take place. These hostile intent gestures can be used to assign individuals a threat assessment level and improve remote sensor monitoring. Such assessments are useful for monitoring human activities and could potentially provide early warning of lED emplacement activities.