Human surveillance system for security application

Abstract

Human surveillance is an important research activity for security concern. Due to the increasing demand of security in different domains, development of smart and efficient surveillance system has attracted immense interest in recent years. Most of the existing surveillance systems are based on monocular camera and limited by their fixed view angles and hence cannot provide sufficient three-dimensional depth information for person recognition and tracking. This paper proposes an efficient and cost-effective human surveillance system using stereo vision technique. The system uses a multi-view stereo camera pair for image capturing and analyzes the stereoscopic pictures to estimate the 3D depth information for accurate detection and tracking of the human objects. The system can provide automatic warning in case of unrecognized people and entrance in the restricted zones. Experimental results are arranged to demonstrate the robustness and efficiency of our proposed system. Our system is very inexpensive and computationally fast comparable to the existing state-of-the-art surveillance systems.