Reconstructing humans' hand motion: Preliminary results and applications in the design of mechanical fingers for anthropomorphic tasks

This paper reports the development of a low-cost sensor-based glove device using commercially available components that can be used to obtain position, velocity and acceleration data for individual fingers of the hand. Optical tracking of the human hand and finger motion is a challenging task due to the large number of degrees of freedom (DOFs) packed in a relatively small space. We propose methods to simplify the hand motion capture by utilizing accelerometers and adopting a reduced marker protocol. The preliminary results show that the use of relative position, velocity and acceleration homogeneous transformations enable us in getting improved finger motion data with respect to those obtained from a Vicon system. This data is directly related to contact and curvature constraints between the fingers and the grasped object. Once obtained from the glove, the higher derivative specifications are used in formulating the synthesis task for the design of robotic fingers.