Show Your Hands: A Vision-Based Approach to User Identification for Interactive Surfaces
Dominik Schmidt, Hans W. Gellersen · Lancaster EPrints (Lancaster University) · 2009
User identification opens up new interaction possibilities on interactive surfaces. Yet many current multi-touch systems only detect isolated touches and cannot identify users. This paper presents a low-cost, biometric method for user iden-tification for vision-based interactive surfaces. To identify users, we extract characteristic contour features from a flat hand posture and use Support Vector Machines (SVM) for classification. Our evaluation shows the method’s robustness together with high true and low false positive rates of 96% respectively 0.5%. We further outline possibilities to inte-grate this method with surface interaction techniques, taking into account that users have to perform distinctive hand pos-tures to afford identification. ACM Classification Keywords H.5.2 Information interfaces and presentation (e.g., HCI):: User Interfaces—Input devices and strategies (e.g., mouse, touchscreen) (a) (b) (c) (d) Figure 1. Extraction steps: (a) raw camera image, (b) extracted con-tours, (c) high curvature points, (d) extracted hand features