A Physiology-Based Flexible Strap Sensor for Gesture Recognition by Sensing Tendon Deformation

Gesture recognition using machine-learning methods has been widely studied for human-machine interaction including advanced cybernetics, virtual reality, and healthcare systems. In this paper, we propose a physiology-based flexible strap sensor attached to the back of the hand. By sensing the tendon deformation on the back of the hand, the proposed sensor can recognize hand gestures with high accuracy. The proposed sensor contains six pressure sensing units connected by a flexible strap substrate. The graphene aerogel (GA) serves as the sensitive material of the sensing unit, which is sealed with two polyethylene terephthalate (PET) films. The size of the proposed sensor is 130 mm (L) × 6 mm (W) × 3 mm (H), which is flexible and stretchable for fitting different hands and different gestures. The sensing units can cover the main tendons on the back of the hand, and the data collected from the sensing units can provide distinguishing information of different hand gestures. Experimental results confirmed that the proposed sensor could achieve excellent linearity, repeatability, and resolution. A machine learning method was utilized to recognize twelve typical precision-grasping gestures, and the results showed that the proposed sensor and machine learning method could classify the precision-grasping gestures with a recognition accuracy of 84.7%. The technology is expected to provide a promising path for gesture recognition.