Intrusion Detection using Camera and Alert Management

Abstract

The ability of a closed circuit television camera to detect and raise an alert is the future of surveillance systems. The traditional way to provide security to any area is by placing a CCTV camera but this way is not sufficient for handling the frequently occurring illegal activities. In order to update these traditional methods, this research study has analysed the techniques that automate the system using image processing algorithms such as Sum of Absolute Difference (SAD). The technology used for implementation is OpenCV (Open source computer vision) with Histogram of Gradients for performing human detection from images and video streams. The image that indicates the difference between the first stage and the time-stamped image where the intruder may be identified is acquired as the result. The accuracy is measured using the Mean Squared Error (MSE) of the image pixels.