Markerless Human-Manipulator Interface with Vibration Feedback Using Multi-sensors

Abstract

This paper proposes a novel markerless human-manipulator interface, which tracks the posture of the operator's hand for robot teleoperation and provides vibration feedback in the course of operation. In this system, five Leap Motions (LMs) and a wireless wearable GEAK Watch were integrated to track the postures of the human hands so that the workspace was enlarged, and the data was more reliable by fusing measured data from multiple sensors. However, due to the native noises and tracking errors of the sensors, the measurement errors were found increasing over time, which reduces the accuracy of the system. To solve this problem, a hybrid filter algorithm was integrated for the posture estimation of human hands in this paper. Kalman Filter (KF) was used to estimate the position, and the particle filter (PF) was used to estimate the orientation. To improve the fault tolerance of the system, the GEAK Watch was also used to provide vibration feedback when the robot was about to collide with obstacle or the operator's hand moved outside the workspace. Finally, a series of experiments were conducted with this human-manipulator interface and the results indicate that the system works effectively and exactly in the process of robot teleoperation.