Topology models and rules: A 3D spatial database approach.

Abstract

Spatial databases are a key component of 3D applications that require the storage, management, and manipulation of 3D objects. Topological information that describes the containment, adjacency, and connectivity of objects within a 3D space is crucial for complex spatial analysis. The need to maintain optimal storage and computation within spatial databases recommends the use of lightweight approaches such as topology rules and models. However, current spatial databases use 2D topology mechanisms and limited 3D topology functions. This study conducted experiments on existing topology rules and models within ArcGIS and Oracle spatial database. Additional 3D topology rules were also implemented to determine topological relationships between 3D objects stored in Oracle. The 3D topology rules were based on a 36-Intersection Model (36IM) that describes intersections between objects from 0D to 3D. Based on the experiments, the 2D geodatabase topology rules within ArcGIS were able to determine topological relationships between the objects stored as 2D multi-surfaces datatype. The 2D topology rules and the topology model within Oracle were also unable to support topological relationships between 3D objects. The addition of 36IM topology rules could determine topological relationships and describe the dimensions of intersections. Evidently, the support for 3D topology within spatial databases depends on the availability of 3D datatypes. Most databases that support 3D datatypes have limited support for 3D topology, whereby existing topology mechanisms require the decomposition of 3D objects. The use of 3D topology rules supports the maintenance of topological information without breaking down 3D objects into lower dimension components.