Low-Cost, Wide-Area Tracking for Virtual Environments

Abstract

In virtual, augmented and mixed reality applications, the position and orientation of the participating people or objects is often needed. In most cases this problem is solved by using expensive - and not always that accurate - special-purpose hardware. We propose an alternative position and orientation tracking system that only uses cheap off-the-shelf components like cameras and LED ropes. The construction of the scene doesn’t require any off-line calibration or absolute positioning, which makes it easy to build and indefinitely scalable. There is also no restriction on the number of cameras participating. We will show that we can calculate the orientation independent from the translation or position by using vanishing points of parallel lines. We also propose a new parameterization of the Hough transformation to make the calculations of the line detector sufficiently fast. The proposed algorithms have been implemented and tested in a virtual set-up. The first results from our tracker are promising and can compete with many (expensive) commercial trackers. The construction of a room-sized lab set-up is in progress.