Security enhancement in open and closed environment with Video analytics and Deep learning

Abstract

The real-time requirements demand that there is need to ensure safety of the human beings as well as to ensure the safety of the properties when the owner is not available at the site or country itself. Hence this paper proposes a framework that could make the CCTV surveillance into a smart device that could inform the owner or inmates in case of any unwanted intruders. Intrusion detection is highly possible with the existing tools and techniques like machine learning and deep learning. Many algorithms do exist for scene identification in images. This paper further refines these algorithms so that it can be used for object detection in night-vision by utilizing the strength of programming languages like Python and Java that has numerous supporting packages.  This smart cam would be a boon to both the common people to ensure safety of the loved ones and the city police who could use this technology to place the criminals behind the bars on timely intimation.