Study Of The Tree-Like River Map Generalization Based On Knowledge

Map is not only a graphic description of spatial information, but also an artwork. There are many creative works in a map, such as map design and map generalization. Today, the digital age, GIS is a main tool for spatial analysis. Map is the best description form of spatial information yet. Getting and updating map data quickly is the key to extending GIS' application field. It could not avoid map generalization to how fully uses existing map data, Automating map generalization has been the bottleneck of GIS' development.Now, map generalization turns to special geo-feature oriented generalization research from general geometry oriented generalization research on model algorithm, then map generalization has different semantic oriented essential generalization, this paper which emphases on linear spatial features generalization and thinks drainage features for the example designed a simple generalization system. The main contents of paper include:1. Discuss the map generalization theory. Analyze the challenges of the automated map generalization. Discuss the automated map generalization strategy based on knowledge. Including knowledge expression and knowledge formalization method of map auto generalization.2. Discuss the knowledge formalization method of river selection; sum up the rules of river selection, according to the characteristics of dendritic river systems, The structuring river data model facing the river system spatial knowledge was built.3. Based on the structure characteristic of river network and the principle of river selection, the river was auto-selected using genetic multi-object optimization algorithm. .The indicators including length, interval and importance of river etc. and the colltivity structure characteristic of river-network were taken into account when selecting the river.4.design a simple rivers generalization system, and paper takes example for the river generalization from 1:100,000 to 1:250, 000.