Abstract

A novel authentication mechanism relied upon multi-touch gestures on touch sensitive device are examined in this paper. We defined an approved set of 22 gestures was characterized using movement features of fingers and hand. A study was performed to evaluate all of the gestures that combined biometric data collection with usability options. After this, a gesture matching algorithm was developed invariant to rotation and translation. A good performance was achieved in distinguishing different users who perform gesture authentication. High accuracy rate is achieved when using single gestures and found a substantial improvement when multiple gestures are used in sequence. From above studies, we conclude that multi-touch gestures show great assurance as an authentication mechanism.

Authors and Affiliations