Context-aware Thinning of Artificial Water Networks for Map Generalization

Research output: Contribution to journalArticleScientificpeer-review

41 Downloads (Pure)


This paper summarizes the research conducted to improve the automatic generalization of man-made water networks for topographic maps by context-dependent pruning (Altena, 2014). The aim of this study was to improve existing thinning methods for map generalization by accounting for landscape types. The results show that it is possible to improve the thinning of water networks by taking into account separate landscape types. On a more abstract level, the study delivers a methodology for the pruning of man-made networks with regard to landscape typology. In addition, it provides a method for evaluating the quality of generalization results for networks. First, previous research on both thinning and evaluation of thinning results is described. Secondly, a selection of existing algorithms are implemented and evaluated by several experiments: identification of landscape variation based on feature morphology and humidity; selection of representative test areas; and geometric network improvement. Results show that the connectivity of the network can be significantly increased. This is important to obtain better generalization results. The final experiments investigated the effectiveness on various landscape types of three different thinning algorithms. The results are evaluated in terms of the amount of thinning, the resemblance of the results to the input data, and the deviation in connectivity. The findings of this research can be used to improve the thinning of artificial networks by applying a customized thinning method to each unique landscape type. In addition, the proposed metrics to measure the effectivity of thinning algorithms – reduction, resemblance and connectivity – have been proved to be appropriate criteria for the comparison of results of alternative thinning approaches.
Original languageEnglish
Pages (from-to)12-29
JournalJournal for Geographic Information Science
Issue number1
Publication statusPublished - 2016

Bibliographical note

ISBN 978-3-7001-7988-7


  • automated generalization
  • cartography
  • evaluation metrics
  • hydrographic network
  • geography


Dive into the research topics of 'Context-aware Thinning of Artificial Water Networks for Map Generalization'. Together they form a unique fingerprint.

Cite this