Geometric Information Makes Spatial Query Processing More Efficient.
Volker Gaede · 1995
In order to index complex and heterogeneous cartographic data by means of spatial access methods, one typically uses single, simple geometries to approximate the given geometries. Traditionally, no information regarding the quality of the approximation is stored. In this paper, we show that the availability of such information allows us to decide at an early stage that numerous objects fulfill the spatial predicate. This leads to significant performance improvements for the spatial selection as well as for the spatial intersection join. We present two techniques enabling such an early decision for the z-ordering method, and we demonstrate their efficiency by means of experiments. 1 Introduction Fast access to spatial data stored in databases is essential for answering spatial queries efficiently. As a result, numerous spatial access methods have been proposed and studied in the past. The complexity and heterogeneity of spatial data makes it impossible to store cartographic data direc...