Object Identification Based on the Automated Extraction of Spatial Semantics from Web3D Scenes

We present a web-based methodology for the extraction of semantic information and object identification in poorly annotated Web3D scenes. Our approach is based on a set of rules that mimic human spatial cognition, backed by an indexing mechanism for the automated spatial correlation between objects in terms of linguistic predicates. The proposed methodology is comprised of real-time techniques with broad application areas, ranging from CAD and 3D room planning tools based on architectural designs, to medical and e-learning environments where their objects' relative position is closely associated with identification attributes.