A New Wearable System for Head Gesture Recognition Designed to Control an Intelligent Wheelchair
Monica-Claudia Dobrea, Dan-Marius Dobrea, Ionuţ-Cristian Severin · 2019
This paper proposes and evaluates a new simple method of recognizing head movements in order to empower a tetraplegic to easily control his wheelchair. The input for the recognition system is one based on the capacitive signals acquired from a wearable system consisting of four capacitive sensors embedded in a necktie. The proposed non-contact wearable system is a very simple and cheap one, being accessible, and, above all, it is very easy to use. All processing and analyses were done offline and the obtained competitive results make our approach a very promising and feasible solution from both, economic and constructive point of view. A key advantage of the system resides in its high simplicity which considerably reduces the computational burden, making it suitable for real-time processing. Also, due to the capacitive sensors' characteristics and their arrangement, the head gesture recognition process becomes a straight one, leading to very high recognition accuracy.