Automatic generalization of cartographic data for multi-scale

Abstract

The multiscale base map compiled from the official 1:10 000 framework data is served as the background in the national geoportal map browser. High expectations of the users of this map – both-up-to-datedness and comfort of use – are pressing to search for more efficient methods to generate it preserving highest cartographic quality. There are two ways towards that: (a) automated generalization of the georeference base dataset into smaller scale datasets that are then used as sources for the multiscale web map and (b) automated cartographic generalization of the single source dataset into multiscale map layers (used in Lithuanian geoportal). As it is commonly believed that generation of Web map layers from separately generalised data sources is more appropriate, the authors performed a research in order to compare the two methods in terms of precision of representations, efficiency of update and communicative quality of the resulting maps. Some procedures that allow for improvement of visualization quality when the second method is used are discussed in the paper. The main conclusion drawn from the research is that a multiscale map generated by means of cartographic generalization can for many applications successfully replace multiscale map generated from separately generalized data sources.