Augmented reality (AR) constitutes a very powerful three-dimensional user i
nterface for many "hands-on" application scenarios. To fully exploit the AR
paradigm, the computer must not only augment the real world, but also acce
pt feedback from it. In this paper, we present an optical approach for coll
ecting such feedback by analyzing video sequences to track users and the ob
jects they work with. Our system can be set up in any room after quickly pl
acing a few known optical targets in the scene. We present two demonstratio
n scenarios to illustrate the overall concept and potential of our approach
and then discuss the research issues involved. (C) 1999 Elsevier Science L
td. All rights reserved.