Dieter Fellner, Fraunhofer IGD, Darmstadt, Germany, Section Informatics
Abstract:
This presentation argues that in contrast to traditional scanning approaches of 3D objects / models the resulting
models should not be based on billions of triangles but on generative structures which accurately (and
hierarchically) encode semantic structures thereby esembling building structures rather than discrete surface
approximations.
Due to the semantic encoding technology the resulting models will then be by several magnitudes more compact,
compared to traditional triangle-based surface approximations thereby lending themselves for an internet-based
access and dissemination.
Maybe even more important, this new approach will support traditional library services like searching, indexing and abstracting in libraries of 3D objects -- all of this only possible in an automated way by new and innovative types of metadata.
| operated by | 
|
