A. Braun, X. Halbecq, F. Lecordix, J.-M. Le Gallinc, F. Prigen

I.G.N. France



The French National Mapping Agency (IGN) aims at building a new flowline for its 1:25.000 and 1:50.000 topographic map series. Indeed, the full coverage of the French territory with the metric precision digital landscape model (DLM) BDTOPO is going to be achieved in 2007. The new flowline should derive the 1:25.000 and 1:50.000 digital cartographic model (DCM) from this DLM. A new project was entrusted with designing and developing this flowline. The challenges identified are :

         The use of multiple sources of data, to enrich BDTOPO with data needed for the topographic maps.

         The conflation of these heterogeneous data.

         The use of automatic process to derive the 1:25.000 and 1:50.000 DCMs from the DLM according to [TREVISAN 04].

         The capability to ensure a rapid and reliable mechanism for propagating updates stemming from the DLM into the DCM.

         A join management of the two scales (1:25.000 and 1:50.000) in all the flowline.

         A full coverage of the French territory in one database, with the open-source database management system (DBMS) PostgreSQL and the spatial cartridge PostGIS [POSTGIS 06].

         A technological compatibility with future maps on demand services (potentially web services).

         A use by about hundred operators in a user-friendly environment.

         A use in six production units dispatched in different places in France.


In order to tackle these problems, numerous research technologies have emerged and are mature enough to become industrialized: automatic conflation tools for multiple sources of data [MUSTIERE 06], automatic generalisation data process with agent-based methods [LECORDIX et al. 06], tools for propagating updates between different geographic databases [BRAUN 04]. Some of these technologies have already been embedded at IGN in a new process for the 1:100.000 topographic map series now industrialized [JAHARD et al. 03]. Taking benefits from these results, the new project is managing more ambitious targets with more production constraints.


This article will present the first results of the project, especially:

         The design of the software architecture that allows for the management of very large database (multi-scales, unbroken on the full territory) and for the decentralization of the production.

         The technologic options chosen to ensure the compatibility with future cartographic (web) services.

         Some results of automatic process (recovery of heterogeneous data, conflation, generalisation) initially stemming from research prototypes, but embedded in robust and user-friendly software environments.