SNAPVis and SPANVis: Ontologies for recognizing variable vista spatial environments

Abstract

This paper gives the SNAP and SPAN ontologies relating to recognizing variable vista spatial environments, namely, SNAPVis and SPAN Vis. It proposes that recognizing spatial environments is a judgment process of whether the perceived environment is compatible with the remembered one. Their compatibility is based on both their spatial changes and the commonsense knowledge of objects' stabilities. The recognition result is determined by whether such changes are due to possible movements of related objects or not. This paper presents six SNAPVis ontologies: fiat boundary, near region, fiat parts (the three are fiat regions), classic topologic relations, qualitative orientations, and qualitative distances (the three are spatial relations) and one SPANVis ontology: the commonsense knowledge of stability of spatial objects. The paper briefly presents a cognitive map of vista spatial environments and the process of recognition. © Springer-Verlag Berlin Heidelberg 2005.