Research action on Cartography, graphic semiology, generalisation, legend conception


This research topic aims at improving map conception processes, on the one hand by automating cartographic generalisation processes, on the other hand by proposing methodologies to help various users to make customized maps.

The action research in cartographie has been active from 2013 to 2014

Leader of the research action :
Guillaume Touya


Key words : legend, map, colour, graphic semiology, cartograohic generalisation, knowledge, conception, customisation, cartographic representation, geovisualisation, multi-agent systems

Research Topics

  • Automation of cartographic processes
  • Previous research at COGIT lab led to the automation of map production lines at IGN. COGIT still aims at the improvement of these automated productions, but also at the contribution to new automated cartographic processes. The integration of self-evaluation systems inside the cartographic processes is one the challenges to face for the automation. Historically, cartographic generalisation is a big part of the cartographic processes we seek to automate. The research on automated cartographic processes also contributes to the automation of non-cartographic processes like the generalisation of geographic databases, or the conflation of different geographic datasets. Artificial intelligence techniques, such as multi-agents systems, are used to automate cartographic processes.

  • Customised map design
  • Besides the production of traditional map series, the cartography team is also interested in on-demand customised mapping, such as stylised maps, maps generalised at any scale, or maps with more or less abstract or realistic representations. Such cartographic services should integrate measures of the cartographic output chosen by the user (fitness for use, legibility, effectiveness, aesthetic).

  • Interactions with the cartographic representation
  • People more and more interact with geovisualisation tools or geoportals, rather than with paper maps. The issue is to provide navigation capabilities that allow the user to switch between scales and representations without being lost because of abrupt representation changes. We also seek to propose methods to help the conception of on-demand quality maps, with interaction and dialogue tools.

  • New cartographic representations
  • We aim at the conception of automatic tools to derive new cartographic representation that can be adapted to different user needs and preferences, and to diverse supports: more realistic maps (e.g. by insertion of parts of orthophotos), or more aesthetic maps with adapted textures and rendering (e.g. for sea or mountain representation), or maps that represent the quality of displayed data, for instance.

  • Cartography with user data and Volunteered Geographic Information
  • We want the automatic cartographic processes developed by the research team, to be able to handle either user data (e.g. derivation of a hiking map with tracks imported by the user), or Volunteered Geographic Information (VGI) (e.g. make a large scale topographic maps with OpenStreetMap data). Regarding VGI, we focus on the heterogeneity of the level of detail.

  • Software development
  • As part of these research objectives, we seek to develop advanced cartographic tools on the open source research platform of the COGIT lab, namely GeOxygene, in order to experiment our methodologies. The GeOxygene software modules CartAGen (for generalisation) and semiology (for legends) are examples of implemented research on this platform.

Thèses en cours

  • Sébastien Biniek :
    Systèmes typographiques paramétriques et programmatiques en cartographie dynamique et multi-représentations
    Supervised by Guillaume Touya
    Co-supervision : Gilles Rouffineau

COGIT PhD on this topic

IGN Web Site