A taxonomy for information linking in augmented reality

Abstract

A key challenge in augmented reality is the precise linking of virtual information with physical places or objects to create a spatial relationship. The visual presentation of these links can have many forms like direct overlays or connection lines. In spite of its importance, this topic has never been systematically addressed by existing approaches. As a first step in this direction, we suggest a taxonomy for such visualizations to facilitate their detailed analysis in terms of graphical properties. It consists of the three artifact types spatial anchor, information object and information connection as well as the three dimensions reference system, visual connection and context. Additionally we surveyed literature to collect the knowledge on how these dimensions and their combinations affect user performances. To explain the application of our taxonomy, we classified user interfaces from literature. We also conducted an empirical experiment regarding the effects on task performance of different classes from our dimension visual connection, i.e. the type of visual connection that is presented to the user. The outcomes give important guidance for augmented reality interface design in a part, which has not been researched before. The results show that the preferred method for visualizing information linking is using a close spatial proximity, followed by a continuous visual connection, a color coded symbolical connection and a shape coded symbolical connection.