A study on various methods used in automatic video surveillance

Video surveillance is the most important issue in today's world as crime is increasing at an alarming rate. Moving object detection is first step in video surveillance. Analyzing video scenes is a challenging task for humans. So, efficient algorithms are deployed to do so. Detection and tracking of objects in videos is critical for wide solicitation fields such as crime detection, fight detection, person identification etc. Many methods exist to analyze the object behavior which includes background subtraction, foreground detection, cross scene abnormal detection etc. This work presents a review on available techniques for analyzing the behavior along with comparative study among various methods.