3D referencing techniques for physical objects in shared augmented reality

We introduce an augmented reality referencing technique for shared environments that is designed to improve the accuracy with which one user can point out a real physical object to another user. Our technique, GARDEN (Gesturing in an Augmented Reality Depth-mapped ENvironment), is intended for use in otherwise unmodeled environments in which objects in the environment, and the hand of the user performing a selection, are interactively observed by a depth camera, and users wear tracked see-through displays. We present the results of a user study that compares GARDEN against existing augmented reality referencing techniques, as well as the use of a physical laser pointer. GARDEN performed significantly more accurately than all the comparison techniques when the participating users have sufficiently different views of the scene, and significantly more accurately than one of these techniques when the participating users have similar perspectives.