Smart Tablet Monitoring by a Real-Time Head Movement and Eye Gestures Recognition System

Abstract

Different research works are studying to integrate new ways of interaction with mobiles devices such as smartphone and tablets to provide a natural and easy mode of communication to people. In this paper we proposed a new system to monitoring tablets through head motions and eye-gaze gestures recognition. This system is able to open browser application with simples motions by the head and the eyes. For the face detection we used a viola and jones technique, and for the eyes detection and tracking we used Haar classifier to the detection part and the template matching for tracking. We develop this system on android based tablet. Through the front-facing camera the system capture a real-time streaming video on which we implemented the detection and recognition module. We tested the system on 10 persons and the experiment results shows that this system is robust and it is invariant for the lightness and the moving state of the user. We introduce also in this system, a module to estimate point of regard on the tablet’s screen based only on the front-facing camera. As known the tablets resources are limited, in consequence the response time of our system is not satisfied for the real time condition. That is why we apply the cloud computing to solve this problem.