Abstract:

Spatial clustering analysis is important method and study content of spatial analysis and spatial data mining. Based on an introduction of square plane of Map Algebra, the algorithm of distance transform is discussed. In terms of the algorithm, a new raster-based method of spatial-clustering for spatial points sets is proposed, then, this method is applied to the case of all-kind objects, weighted distance and the space having obstacles. From the examples, it is proved the algorithm is simple and reasonable.