Articles by Eric Desjardin

In applying an approach of dematerialization of planimetric, written and archaeological data, that should be soon processed through a GIS, it will be necessary to estimate the errors made in the compilation of the Napoleonic land registry made in order to be able to work on the construction of the urban space. The registry was started in 1811 and as a source document for morphological analysis and regressive mapping, it is geometrically quite precise. However, a review of this fiscal document (which has a twisted perspective, perhaps to tell us something about the townscape) and a comparison to correct maps have revealed a few errors depending on the level of treatment. It therefore becomes useful to characterize and to explain them, before linking them with oldest documents and archaeological data. The example of Les Coutures will allow us to put into practice these observations in order to define methods of metrological analysis and to study dynamics of urban space occupation according to various spatial and temporal scales.

By questioning the past, archaeological information is naturally prone to imperfection. Through a series of examples, we will briefly present its various aspects (inaccuracy, uncertainty, vagueness, conflict, lack) which can apply to time, space and function. The first stage consists in the identification, characterization and recording of imperfection in the archaeological information system. At the second stage, the question arises of how the imperfection of knowledge in archaeological hypotheses should be taken into account in terms of analysis, production and restitution. In the SIGRem project, we have chosen to resort to the Fuzzy sets theory. At a final stage, although the promotion through the media of results can nowadays be carried out by a 3D modeling, realistic reconstruct being very often confusingly perceived as truth, we tend to lose the richness of confidence levels we have in our knowledge. Therefore, we will also describe how visual paradigms can be used to enable dynamic perception of uncertainty in dedicated 3D virtual environments.