Deriving Functionality from 3D Shapes: Ontology Driven Annotation and Retrieval

3D shapes are of crucial importance in several applications but capturing, representing and processing the complexity of the knowledge a shape may reveal is not an easy task. Therefore it becomes crucial to study which form characteristics may have a specific meaning in order to make explicit, sharable and, above all, machine understandable, the knowledge embedded in 3D shapes. This may be done from different perspectives due to the increasing diversity of potential users. This paper illustrates a knowledge-based (KB) system supporting the functionality driven annotation and retrieval of 3D models. Kernel of the system is the Form, Functionality and Behavior ontology, which provides a representation of the knowledge needed to infer object functionality and behavior from the shape of models and their components.