Automatic generation of thematic maps using Multi-Agent Systems

International audience Thematic maps are tools that allow various stakeholders to track the spatial distribution of a phenomenon, at a given moment, and assess its evolution. However, it is difficult to construct real-time thematic maps that reflect spatial evolution and follow digital mapping standards. The aim of this study is to create a tool capable of generating thematic maps that evolve according to the behavior of the actors in their environment. To do this, we use multi-agent systems (MAS), where agents interact with their environment. This environment is initialized using a satellite image, which is classified by remote sensing. Additionally, Geographic Information System (GIS) vector data is employed to delimit different zones. We chose the GAMA development environment to implement this approach. These solutions offer a tool capable of generating maps of resources applied to forests that reflect reality, at different times during the simulation. We applied the study to the Vohibato district in the Haute Matsiatra region of Madagascar. The results enabled us to generate various types of thematic maps in conformity with digital mapping standards, showing the evolution of forest cover in each commune of the district. This shows that the district lost around 22% of its forest cover between 2018 and 2038. Generating thematic maps automatically will facilitate understanding and explanation of changes. In the context of forest cover, it allows us to assess forest conservation strategies and facilitate discussions and decision-making among the various stakeholders involved in forest ecosystem conservation.