Flohr, DanielFischer, JanBernd Froehlich and Roland Blach and Robert van Liere2014-01-312014-01-312007978-3-905673-64-7https://doi.org/10.2312/PE/VE2007Short/059-064The estimation of the position and orientation of the digital video camera is a central challenge in video seethrough augmented reality. Many augmented reality applications solve this problem with the help of markerbased methods, which analyze artificial fiducials in the images of the real environment, e.g., using the widespread ARToolKit library. Among the drawbacks of conventional marker tracking is the necessity to manually define marker patterns. Badly chosen patterns have a negative impact on tracking performance. Although improved methods for automatic marker generation have been described, manually controlled marker tracking is still widely used in many applications for practical reasons. In this paper, we describe a lightweight drop-in extension for IDbased marker tracking. Our system makes it possible to automatically generate a large number of tracking fiducials identified by unique numerical IDs. The created marker patterns consists of large monochrome patches, which improves the recognition rate and tracking performance compared to typical manually defined fiducials. Due to the design of our extension, only minimal adaptations are required in order to add ID-based tracking to existing augmented reality software. We discuss experimental results demonstrating the improved pattern recognition and describe an example application.Categories and Subject Descriptors (according to ACM CCS): H.5.1 [Information Interfaces and Presentation]: Artificial, augmented, and virtual realities