Distributed user tracking concepts for augmented reality applications

Augmented reality (AR) constitutes a new user interface paradigm. Using light head-sets and hand-held or worn computing equipment, users can roam their daily working environment while being continuously in contact with their computer systems. For AR to work properly in a large factory, new system architectures have to be designed with consideration of the special requirements imposed by AR. In particular, AR requires real-time facilities to track the user's position and viewing direction. In the past, various carefully calibrated sensing devices have been used for this purpose, including magnetic trackers and active LED systems. Research is now focusing on computer vision-based methods. It is our hypothesis that, in the future, the most successful indoor approaches will combine local (user-worn) vision-based tracking methods with global user tracking schemes from fixed (wall-mounted) sensors, using mobile wireless networking technology to allow the user-worn system to communicate with the globally available sensing infrastructure in an intelligent building. We propose the concept of "AR-ready" intelligent buildings which provide built-in tracking services via different sensing modalities.