This paper enables us to do 24x7 video surveillance a specified area and alert the end user in real time. conventional surveillance system requires a person to monitor the video, It also takes huge storage capacity and lack computation capability during monitoring. We aim to overcome these cons of the existing system by creating a real time video surveillance system. This system is used in the area where on one is allowed without permission. Our methodology aims to detect the moving object and segment objects supported motion information and alert the end user. For this we planned a pixel wise background modeling which compares the background image with the foreground image. Background image Which is initially stored once the webcam is turned on Foreground image Which are captured by the webcam after the background image is stored and these are compared with the Background image to get the status. It uses pattern recognition algorithm for the complex comparison process between two images. After the comparison operation if an moving object is detected and if the threshold value is over the specified limit then interrupt will be send to the end user’s mobile device using a GSM modem
展开▼
