Gesture Recognition with Sensor Data Fusion of Two Complementary Sensing Methods

This paper presents the further development of a Dielectric Elastomer Sensor (DES) based gesture recognition system by sensor data fusion of two complementary sensor systems. The combination of two independent sensor systems with different physical principles enables a reliable recognition of hand and arm movements, which can be used to distinguish the origin of the movement, which means whether it is initiated actively by the nervous system or passively by external forces. The voluntary movements of the hand with the corresponding activity of the forearm muscles are registered with a noninvasive electromyography sensor system of the Myo gesture control armband. The steady-state positions of the passively positioned arm are detected with flexible DES stretching over the arm joints. In this paper, the solution approach as well as the experimental setup for a wearable gesture recognition system based on sensor data fusion of the sensors is presented. Promising results show the capability of combining the advantages of each sensor by fusion of the two different sensor data. This system can be used in various applications such as rehabilitation monitoring or intuitive control of robot systems.