The selection of scattered settlements based on ant colony optimization algorithm

Scattered settlements are composed mainly of individual buildings that can be seen as point cluster. The selection properties of points include selectable, disposable and selectable-or-disposable. According to the traditional basic theory relative to selection of scattered settlements, the basic idea to select scattered settlements is put forward. The selection of point cluster is transformed gradually into the simplification of linear cluster, and ant colony optimization algorithm is used to simplify the linear objects finally. The example on how to accomplish selection is taken: firstly subclass regions are partitioned by Delaunay Triangulation, and reclassified based on minimum spanning tree (MST); then MST is divided to some simple linear objects; finally, ant colony optimization algorithm is applied to simplify these linear objects. The experiment results indicate the proposed method is feasible and effective.