Development of Finger Motion Reconstruction System Based on Leap Motion Controller

Abstract

In order to study the process of finger movement through visualization, this paper presents a Leap Motion Controller-based finger motion reconstruction system. With the Leap Motion Controller, finger motion data is easily generated for the calculation of angle of finger joints. Additionally, the calibration functions for three types of joint are derived to improve the measurement accuracy. The calibrated joint angular data is segmented and normalized in time, and then the finger frames are calculated to reconstruct the finger motion. The experimental results showed that the calibration functions had a clear positive effect on improving the accuracy of joint angle measurement. Besides, the system has been proved to be able to identify the feature points of the motion effectively and achieve the recognition of sub-motions. Consequently, the process of sub-motions was reconstructed by plotting all motion frames on a figure. Through the finger motion reconstruction graph, the kinematic information on multiple joints and at various time points can be obtained visually and directly, which is helpful in understanding of the biomechanics of finger motion.