Real-Time SuspiciousActivity Recognition in Examinations

Abstract

Using unfair means in the examinations is very common nowadays, and the examinee who tries to cheat and ace the examinations is quite a big problem for the student who had done hard work to appear in the exam. Recognition of the suspicious human activity in video surveillance is an effective tool for today’s business such as in examination surveillance where the real need is to be able to analyze suspicious human activity. Manual checking of examination conducting in a computer lab might be inclined to errors during human supervision. So there is a need for an automated suspicious activity detection system which will help in distinguishing suspicious activities yet additionally helps in limiting such activities. Besides the error probability will be a lot lesser. For the detection of the suspicious activities in examinations, our system will be able to detect the face and the eyes. Although the case of installing cameras in the lab, would be much help to reduce the human error and the human cost used in the invigilation. The system automatically detects suspicious activities and generates alert. Such systems have been developed previously but have some certain limitations but our system aims to solve these problems. Also it provides a good automated surveillance system.