Research On Automated Selection Of Hash-style Habitation

Cartographic generalization has always been one of the most creative and challenging field of cartography, is also a need to address but not well resolved. With the development of GIS, we often need to produce small-scale map data from a large-scale map data, map data is made to achieve a multi-purpose libraries, while the automatic cartographic generalization is the key to solve this problem.Habitation is the one of the most important elements on the map , is also the one of the largest proportion elements. As the generalization contents of the habitation is relatively more, and its graphics processing is complicated, so the habitation element is a key and difficult problems in cartographic generalization.Hash-style habitation is a important part of habitation, but the study on the its cartographic generalization is very weak, which directly affected the entire effect of the habitation element cartographic generalization. we try to introduce the idea of"breaking up the whole into parts"into the hash-style habitation of automatic selection in its deep research, and carry out a series of useful discussions.The main contents of this article are as follows:(1) Based on the idea of"breaking up the whole into parts", we create the selection operational process for the hash-type habitation automatically selection.(2) We work out the new clustering algorithm suitable for the hash-style habitation. This algorithm does not need pre-determined cluster centers, and full simulation of nuclear fission chain reaction can quickly find the hash-style habitation for clustering.(3) Improve the algorithm based on Delaunay triangulation for finding out the boundary points, it uses the best contours to determine boundary points, enhances accuracy of the boundary points, and does not miss the boundary points.(4) Improve the selection model, make it suitable for hash-style habitation selection.(5) Create the selection model for the boundary points and internal points. The model based on the idea of"breaking up the whole into parts", constructs dynamic Voronoi polygon for internal points, cancel the dense points, while fixing the adjacent points, ensures the dynamic real-time reflection of the points'density and uniformity of selection. It arranges the importance of the border points, and select the more important points first to maintain the basic outline.(6) Experimentations for the above algorithms are carried out, and the good results show effectiveness of these cartographic generalization algorithms.