Simultaneous Localization and Mapping for Augmented Reality

Recently, the methods of Simultaneous Localization and Mapping (SLAM) have received great interest in the field of Augmented Reality. Accurate tracking in unknown and new environments promises to reduce the initial costs of building AR systems which often require extensive and accurate models of the environments, interaction objects and virtual annotations. However, it is still an open question how interesting and useful annotations can be created, attached and stored for unknown and arbitrary locations. In this paper, we discuss possible uses of SLAM in the different components of typical AR systems to provide meaningful applications and go beyond current limitations.