Storing and using scale-less topological data efficiently in a client-server DBMS environment

In this paper we present a data structure that stores the results of a generalisation procedure efficiently as a scale-less map inside a spatial DBMS. This structure makes is possible to interactively visualize polygonal subdivisions on any scale efficiently. This is done by maintaining a topological structure from which a map can be reconstructed. The reconstruction of a polygonal subdivision for a given scale is done in two steps. The first step retrieves the necessary boundary lines from the database together with information on how these boundaries should be combined to for a subdivision. The second step reconstructs a topologic layer from these boundaries. The two steps in the process are modelled in such a way that the first step can be efficiently implemented on top of a standard spatial DBMS (with three simple SQL queries). The second part of the process, which is more iterative, can be either performed at the client side or on an applications server. An important feature of the data structure is that the data is stored topologically in such a way that as much of the geometry of an object is re-used. This makes the storage very compact and ensures that only little data needs to be shipped from the database.