ABNORMAL-HUMAN-ACTIVITY-DETECTION-SYSTEM

Abnormal-Human-Activity-Detection With the increase in the amount of anti-social activities taking place in the environment, security has been given the utmost importance lately. Therefore, organizations require a constant monitoring of people and their interactions. Since this constant monitoring of data by humans to judge if the events are abnormal is a near impossible task as it requires a lot of workforce and constant attention. Therefore, the challenge that comes up is the demand for an automatic and intelligent analysis for such video sequences. Our project comes forward as an attempt to provide solution to such a problem as the model developed is a smart surveillance system which can detect unusual or abnormal activity automatically. A method for representing the motion characteristics is described for detection and localization of unusual activities in the crowd scenes on a generalized framework which includes both a local and global range for detection of such activities.