2P1-N08 Grasp Posture Database and Its Application to Grasp Posture/Motion Analysis(Digital Human)

Abstract

This paper proposes statistical approach for creating grasp taxonomy and analyzing process of grasping motion. We constructed a grasp posture database by conducting principal component analysis for joint angles of hand measured by optical motion capturing system. We can obtain grasp taxonomy by conducting clustering analysis for the distribution of principal component scores for the first and second principal component. We can also analyze the process of grasping motion by plotting corresponding grasp posture of each time step on the same two-dimensional space.