Hand Skeleton Estimation Based on Two Specific Gestures

Abstract

In the MEMS-based hand motion capture system, virtual hand skeleton proportion is usually calculated by using the statistical length of human hand anatomy. The effect of the difference in the proportion of the individual hand of the users, resulting in the error of the delicate activities of the hand. In this paper, a hand skeleton estimation algorithm using two specific gestures is proposed by analyzing the skeleton structure and motion characteristics of the hand and calibrating the two specific gestures respectively. In the meantime, the attitudes and positions of the virtual hand skeleton is calculated based on kinematics, and the virtual hand movement is reconstructed with high accuracy. The hand tracking and interaction works fine in a simulation system.