Generalization Technology in Three Dimension Scaleless GIS Based on SR-Tree

Scaleless GIS is one of the key technologies of digital earth and WebGIS. The ability of deriving small scale map from large scale map automatically is very important in modern GIS. How to make the spatial data in three dimension GIS automatically adjust with the change of scale and how to make full use of the current database have been very important problems to be resolved. In this paper, aiming at the application in three dimension GIS, the authors first present an improved amount rule of selection expanded from two dimension algorithm. Second, the authors introduce a new quality principal of selection considering the mobile viewpoint. Third, the authors state a novel dynamic combine principal which can deal with elements combining in real time. Fourth, based on the description of the above definition and algorithm, the authors present the data organization strategy SR Tree, which is used for creating and organizing spatial elements in the current three dimension GIS database with LOD mechanism. Fifth, the authors describe the corresponding generalizing algorithm based on SR Tree strategy. The last section shows the experiments of SR Tree strategy. Compared with other data organization algorithm, SR Tree strategy can not only finish generalizing spatial objects in three dimension scaleless GIS more effectively, but also make full use of the current geographic information database. It satisfies the current need in three dimension scaleless GIS applications.