Annotating 3D Content in Interactive, Virtual Worlds

Labeling or annotating 3D content is an essential task for interactive environments such as digital 3D cities or indoor models, to enable target-oriented exploration and navigation. The seamless integration of text information into the perspective view of the scene has the advantage that switching between legend and content view is avoided. However, it demands for special labeling methods that overcome common problems, in particular, ambiguous label assignment, missing labels, and label cluttering. We present a novel, image-based labeling approach that uses enhanced visibility analysis to dynamically attach labels to visible parts of 2D and 3D labeling objects. We show that this approach not only is suitable for external and internal labeling, but also explain how dynamic switching between internal and external labeling can be performed using our method. We demonstrate the applicability of our method by 2D and 3D map examples.