Reducing Hand Grasp Classes Based on Grip Geometry: Grasp Simulation Using a 20-DOF Anthropomorphic Hand

The human hand is very complex, and it is not easy to provide an exact biomechanical representation. Therefore, available prosthetic hands have limited and intricate control requirements resulting in high user rejection. Reducing the hand grasp classes and subsequently lowering the degrees of freedom (DOF) without compromising on the different graspable objects could cater to significant improvement. In this research, 14 hand grasps have been reduced to 6 grip geometry-based classes. Performing simulation using a 20-DOF anthropomorphic hand, it is shown that the proposed method is successful in grasp completion for all transformed 14 grasps, and grasp stability is achieved for 12 out of 14 cases. Moreover, it significantly decreases the control load of the 20-DOF hand into 14-DOF.