1 Human Activity Tracking and Analysis for

Abstract — Human activity tracking is an important area of research in the field of video surveillance by past 10 years. It is very difficult task to recognize the activity of the object performed in the video. It is very difficult to track the moving target in real time because there are so many factors involved in it. There are so many methods and algorithms available to detect and track the moving object. There is lot of research going on in the field of video surveillance. In the last decade, due to the increase in terrorist activities and general social problems, providing security to citizens have become the top most priorities for almost all the nations and for the same, a very close watch is required to be kept in the areas that needs security. Keeping human watch 24x7 is not possible as we all know that humans can easily be distracted and a small distraction in very sensitive and highly secure area can lead to big loses. To overcome this human flaw in the area of monitoring, the concept of making monitoring automatic came into existence. Since, video surveillance has came in the market, researches have been taking place in order to make to more easy, accurate, fast and intelligent. The goal of visual surveillance is not only to put cameras in place of human eyes, but also to accomplish the entire surveillance task as automatically as possible. In one statement we can say that video surveillance is nothing but taking the video, identifying unwanted entities, tracking their actions, understanding their actions and raising an alarm. In this paper, we will be study the phases of the video surveillance system. We will see the 3 main methods of human detection. Further, we will see most salient region method for tracking and in this paper we propose a method of handling occlusion using velocity and direction information