PROBE Spatial Data Modeling in an Image Database and Query Processing Application

Abstracf-The PROBE research project has produced results in the areas of data modeling, spatial/temporal query processing, recursive query processing, and database system architecture for “nontraditional” application areas, many of which involve spatial data and data with complex structure. PROBE provides the point se? as a construct for modeling spatial data. This abstraction is compatible with notions of spatial data found in a wide variety of applications. PROBE is extensible and supports a generalization hierarchy, so it is possible to incorporate application-specific implementations of the point set abstraction. PROBE’s query processor supports point sets with the geometry filter, an optiniizer of spatial queries. Spatial queries are processed by decomposing them into 1) a set-at-a-time portion that is evaluated efficiently by the geometry filter and 2) a portion that involves detailed nianipulations of individual spatial objects by functions supplied with the application-specific representation. The output from the first step is an approximate answer, which is refined in the second step. The data model and the geometry filter are valid in all dimensions, and they are compatible with a wide variety of representations. PROBE’S spatial data model and geometry filter are described, and it is shown how these facilities can be used to support image database applications.