Exploiting hand kinematic synergies and wearable under-sensing for hand functional grasp recognition

Wearable sensing represents an effective manner to correctly recognize hand functional grasps. The need of wearability is strictly related to the minimization of the number of sensors, in order to avoid cumbersome and hence obtrusive systems. In this paper we present a wearable glove, which is able to provide accurate measurements from three joint angles. These measurements are then completed to reconstruct the whole hand posture, by exploiting a priori synergistic information on how human commonly shape their hands in grasping tasks. Results, although preliminary, show the effectiveness of the here described devices and methods and encourage to further investigate this kind of approach.