Robot Manipulator Using a Vision-based Human–manipulator Interface

It is much useful of the unstructured environment which the objects in it are unfamiliar to use human-robot interaction in remote teleoperation of robot manipulator. Traditionally, contacting mechanical devices under the interactive method can restrict the operator’s motion. While camera-based tracking has the benefit of being non-contacting, previous vision-based approaches have used only few degrees of freedom for hand motion and have required hand motions that are unnatural for object manipulation tasks. Here we provide a non-contacting vision based method for robot teleoperation. By tracking by two cameras, we can get the three-dimensional position and orientation of the operator’s hands during the manipulation of the robot. We also proposed the Kalman filter to eliminate the influence of the noise. As the experimental result, our method can let the operator perform manipulating more naturally and intuitively, which laid solid foundation for effective teleoperation of robot.