This study is concerned with a general model of satellite data interpretation. For this interpretation, the user has to abstract a particular dimension from the landscape considered as a system. The review of some applications in remote sensing has revealed that all the algorithms used in image processing for land surveys consist of operations on three kinds of information extracted from satellite data. In cartographic applications and in land use inventories, the emphasis is laid on the spectral information. When the aim of the study is to model the geographical properties of the landscape, the analysis of spectral information is combined with an analysis of the spatial information. Finally, when evolutionary studies are conducted, the temporal information is introduced.