User Authentication using Leap Motion

Abstract

There is a rising interest in non-touch human-computer interaction and enhanced user authentication. We think that the hand gesture will play an important role in non-touch HCI and that it will be requested to reinforce user authentication associated with the hand. In this paper, we propose a novel method to authenticate each user by recognizing his/her hand with Leap motion. We get the reference and test data about three dimensional positions of finger joints by letting users put his/her hand over Leap motion. We identify and verify a user by comparing the intra-digit and inter-digit feature values calculated from the three dimensional coordinates of finger joints. Twenty persons have participated as examinees in our experiments and we have measured the hand of each user 30 times. The experimental results show that we can get the identification rate 84.65% based on only distances between joints as intra-digit features. This implies that hand biometrics such as distances and angles between finger joints can be used to authenticate the user who interacts with computing devices using hand gestures.