Study Of Contour Generalization Based On Structured Algorithm

With the continuous development of computer technology and the deepening of research in GIS application, map generalization has already become importantly in study of GIS application. Landform element generalization is an important corpus of map cartographic generalization, and the landform elements are the difficult in the form of contour generalization, so contour generalization is an important research content in map generalization.It is difficult to achieve the contour automatic generalization because of its complex and ever-changing in real landform. The contour generalization can be divided into two categories:based on a single curve and group-generalization, and the structured generalization and the landform elevation generalization are principal methods of contour group-generalization. The existing method, to some extent, can complete contour generalization, but these methods have its limitations and relative merits. In addition, the contour is a group of similar curves to express geomorphologic shape, so group-generalization of contour has more practical meaning and can ensure coordination of generalization results. The main subjects of this paper are as follows:1、Semi-automatic pretreatment of the original contour, including broken contour lines, breakpoint and flying-spot topological relationship processing.2、Combining with simplified contour topology relation model, design a contour topology relation model of incomplete binary tree. In the model, only traverse a contour can complete the establishment of the model.3、By D-P algorithm of single line, researching contour characteristic point extraction and structural generalization of based on characteristic points, and discusses the characteristic point extraction and determination of D-P threshold in the composite.4、Arming at the methods in process of structured contour generalization carried on the detailed discussion and research, such as screening characteristic points, connecting characteristic lines, evaluating valley lines, identifying small valley, small valley deletion, intersect processing, etc. Put forward the hierarchical structured idea of valley lines, improve the efficiency of identifying small valley.5、Realize a simple contour generalization system, with comparing the two methods experimental results on two kinds of structured contour generalization and analyzing the existing problems of the two methods and the advantages and disadvantages, provides more ideas for the study of structured contour generalization.