Motion mapping between human hand and HIT/DLR dexterous robot hand based on fingertip positions

This paper constructs a human hand model and a HIT/DLR dexterous robot hand model respectively, then analyzes their structural differences and the motion mapping problem due to these differences. The dataglove collects the angles of human hand finger and the fingertip positions can be calculated based on the human hand model. The human hand motion is mapped to the dexterous hand through the modified fingertip positions mapping method. This method solves the mapping difficulty caused by the workspace difference between the two models. The simulation results show that this mapping method is effective and intuitive, and can satisfy the demand of teleoperation tasks.