Visual marker detection and decoding in AR systems: a comparative study

Abstract

Visual markers are widely used in existing augmented reality (AR) applications. In most of such applications, the performance of an AR system depends highly on the tracking system for visual marker detection, tracking, and pose estimation. Currently, there are more than one marker based tracking/calibration systems available. It is thus desirable for the user to know which marker tracking system is likely to perform the best for a specific AR application. For this purpose, we compare several marker systems all using planar square coded visual markers. We present the evaluation results, both qualitatively and quantitatively, for the usability, efficiency, accuracy, and reliability. For a particular AR application, there are different marker detection and tracking requirements. Therefore, the purpose of this work is not to rank existing marker systems; instead, we try to analyze the strength and weakness of various aspects of the marker tracking systems and provide AR application developers with this information.