Research And Implementation Of Natural Features Automatic Generalization Under The Condition Of Information Technology

Map generalization, as an important field of cartography, is also a scientific method for the representation of digital map information. Map generalization, one of the most important fields of the cartography, with many researches focused on it for many years, still has many issues existing now.With the development of cartography technology from analog map to digital map, then GIS, and digital earth, map generalization also evolves from manual to computer aided. Development of computer technology provides new methods for digital map generalization. In this paper, the author does some researches of some typical kind of natural features based on the existing results domestic and abroad. The main achievements are the follows:1. Basic algorithms and principles about natural features generalization. The first chapter summarizes the main meaning、methods and modes of map generalization. Based on this, some researches about the basic algorithms, such as Delaunay triangle irregular network, voronoi, spatial relationship and contour neighbor tree, are developed to provide the basis for map generalization. Besides, layered process principle in natural elements generalization is proposed, which is inspired by how people learn the world. According to layered process principle, spatial information and property information are both considered as equal factor in dividing the targets into several layers. By this principle the generalization results are more reasonable.2. Researches on generalization algorithm of typical point group natural features. Analysis the characteristic parameters of point group distribution, summarize the common methods for point group generalization, then apply the weighted layered process principle to the point group generalization, which makes the result more reasonable.3. Researches on generalization algorithm of typical linear natural features. Rivers and contours are researched as typical example. This paper analyses the organization characteristics of river net data in shape of arborescence, realizes layered selection and generalization of arborescence shaped river net. Two common methods of contour generalization are analyzed. One is generalization by single contour line; the other is generalization by contour line group. The concrete realization in latter method is mainly explained, and in this method neighbor contour tree is creatively proposed and applied.4. Researches on generalization algorithm of typical area-shaped natural features. Algorithm and realization of selection generalization and shape simplification of lake data are researched.The methods proposed in this paper are implemented on the New Map software developed by our own. Experimental results show that the methods created results acceptably.