The Morphing of Area Features Based on Skeleton Line Endpoint Matching

A method to morph area features based on skeleton line matching is presented in this paper for dealing with the continuous map generalization questions using spatial data. Scale interpolation is based on two key representations; thus this algorithm can dynamically generate multi-scale represen-tations in real time. Skeleton lines of two area representations of same entity at double scales are extracted by Delaunay partitioning. The skeleton lines of two area representations are matched by Optimal Subsequence Bisections, of the result subdivides the polygon boundary into two series of points. The subdivision boundaries are interpolated by the linear approach to obtain multi-scale representations of vector polygons between the original and target polygons. Experiments show that the method can produce smooth and continuous multi-scale representations of area features. The algorithm can be used for continuous map generalization with spatial data in multi-scale representations.