Abandoned object detection with video surveillance

Abstract

We are presenting a new framework for smart threat detection with the help of video surveillance which captures live streaming from the camera and determine if any abandoned object has been set up. It then raises an alarm right after the carrier of the bag leaves the scene. The abandoned object is detected with the help of background subtraction and brought more accuracy on the object’s image with the help of morphology. Every individual person on the frame is being detected by their body with the help of Kalman filter and brought relation of the body with the bag, so that the alarm is raised only when the body of the abandoned object carrier leaves the frame.