SPATIAL STRUCTURES TO SUPPORT AUTOMATIC GENERALISATION

Abstract

Automating the cartographic generalisation process has been the focus of much research. It has repeatedly been reported that one critical task is to model the spatial relationships between map features to understand their meaning and importance for the target map. This paper will review the different spatial structures that are usually used by researchers to support their automatic generalisation solutions. This paper then proposes a model for building and storing a proximity graph between features. This model has been designed for supporting generalisation. This means that the graph works as a logical structure that links the features together by references, allowing fast analysis. Different types of graphs are presented. An efficient technique is described to build a triangulation on a set of features, as well as the tools to derive the other types of graphs from it. These tools have been implemented, the results are presented in the paper.