Software architecture and calibration framework for hybrid optical IR and vision tracking system

Augmented Reality (AR) enhances the user´s perception of the real world by superimposing virtual objects into its field of view. To achieve this, AR tracks the position and the orientation of objects of the real world in real time. However, the scope of classic AR applications has been limited by the range of sensors used for tracking. Hybrid tracking approaches can provide an interesting solution to increase this range. In this paper we propose to combine two heterogeneous tracking systems: an optical IR tracker and a vision-based one in order to obtain a wide area tracking. We describe the calibration procedure of the coupled tracker and we detail how to compute the spatial relationship between the several objects used in our application. With our proposed framework, data from the two tracking systems are automatically and dynamically fused, and then transparently provided to the application. We present the software implemented to carry out this operation. Finally, we discuss the benefits of our wide-range tracker and we examine its implementation in practice.