DESIGN AND DEVELOPMENT OF A SURVEILLANCE ROBOT

Abstract

Conventional surveillance which was done by human, is a dull job and prone to many mistakes. Additionally, manpower required for monitoring is expensive and is not suitable for weltering in remote places. Surveillance camera-based monitoring system is a temporary solution in this scenario. However, surveillance cameras are fixed to a certain position with limited coverage area. Furthermore, surveillance cameras defunct during and after natural disasters such as earthquakes,storms, etc., while there is a necessity for search and rescue of dwindling survivors. We propose a surveillance robot that overcomes the surveillance systems’ restricted coverage area problem with a light weight and low-cost robot structure, attached with a movable surveillance camera. Instead of recording video footage in passive, the system actively telecast visual information to a RaspberryPi system connected with Wi-Fi. A face detection system based on Viola-Jones algorithm is used to detect faces on real time basis. A user-friendly and easy to manage Graphical User Interface (GUI) is introduced using PyQt5. The experimental results show that the robot can detect singleto multiple faces either eyes open or closed on real time basis.