Study And Realization Of Spatial Data Clustering Analysis System Based On SADBS

The spatial analysis database system, SADBS, uses "Realms" as expression base of spatial data. It adopts discrete integer coordinate system to solve topological error in spatial database. It also uses multi-index data organization and plane-sweep algorithm, which improve the performance of spatial analysis operations. SADBS can provide powerful performance of spatial analysis operations, but it has no such operations such as spatial distance and spatial position. This is very inconvenient in some typical spatial database application, for example spatial dataming. This paper mainly researches on spatial data clustering of spatial dataming. Based on density-based clustering algorithms and the spatial data's structure and index of SADBS , a RDBSCAN algorithm have been put forward.And after analyzed currently available constrained spatial data clustering algorithms ,then brought about a new algorithm, the Delaunay Hierarchical Clustering Algorithm (DHCA). In addition,this paper has probed into that how to provide operations of clustering analysis in SADBS and tries to offer the bases support for the operations of clustering analysis.Based on SADBS system, after analyzed the typical algorithms of the clustering analysis and summed up a series of basic operations of spatial clustering,then realized the operations .This expanded the application of SADBS. The new system is named Spatial Data Clustering Analysis System(SDCAS ). And this paper improved the index of spatial date in order to provide better performance for clustering analysis. At last, for some flaws of SDCAS, we give out some new directions and approaches for further research.