Interactive art system for multiple users based on tracking hand movements

We propose a camera-based human computer interface where hand movements directly influence the creation and manipulation of art in an interactive system. This unobtrusive, low-cost interface addresses limitations in previous interactive art systems by supporting multiple users simultaneously in an unconstrained, uncontrolled environment. We chose Markov Models for our computer vision algorithm to track multiple hands across multiple camera views in real time. We show how our algorithm can be applied as a tool for a team of users to interact with Strange Attractors and Hidden Paintings. Strange Attractors is an interactive art system that displays beautiful fractal structures whose location on a large screen reflects the users' hand movements. Hidden Paintings is an interactive game in which two almostidentical versions of an artwork are displayed on a large screen and the movements of both hands of the users are superimposed as bounding boxes onto the display, allowing users to identify image regions in one version of the artwork that differ from the regions in the other. Experiments with twelve subjects show successful adoption of the proposed interactive art system and its two applications.