Coarse, inexpensive, infrared tracking for wearable computing

Abstract

We present a novel, inexpensive, coarse tracking system that determines a person’s approximate 2D location and 1D head orientation in an indoor environment. While this coarse tracking cannot support precise registration of overlaid material, it can be used to drive user interfaces that can adapt to the quality of tracking available. Our approach uses a set of strong infrared beacons, each of which broadcasts a unique ID. The beacons are deployed in the environment such that their zones of influence strategically overlap, partitioning the area of coverage into a set of uniquely identifiable fragments. We use a compound, omnidirectional infrared receiver, composed of a set of individual, directional infrared receivers, to infer 2D position (parallel to the ground plane) and 1D orientation (azimuth), employing a Kalman-filter–based architecture for smoothing and data integration with other tracking systems available. To test our ideas, we have applied them to a prototype head tracker, and present results from our tests. 1.