Utilisation De L’image Multispectrale Pour L’exploration Et La Recherche Des Ressources Minérales : État Des Connaissances Et Proposition D’un Modèle De Traitement

Abstract

Multiband space remote sensing is an indirect tool for prospecting the Earth's surface. It is very powerful especially in its applications related to the field of geology including geological mapping, mining and oil exploration. It can also significantly reduce the cost of exploration, reach inaccessible areas, guide mining research to favorable regions and reach a large surface. In this article, we highlight in details the state of knowledge in this field of research by citing the different methods and approaches carried out by several specialists who generally define the use of remote sensing for lithostructural and mineralogical mapping and particularly for the exploration and research of mineral substances. We also create methods derived from the aforementioned methods of treatment by means of a logical analogy between the different bands of several satellites of observation of the terrestrial globe, particularly between : Landsat 7 ETM +; Landsat 8 OLI / TIRS; Aster and Sentinel 2A. At the end, we synthesize these results by proposing a multispectral image-processing model that can be applied directly. This model starts with the calculation of Optimum Index Factor (OIF), which allows us to detect only the most important colored composites; and the reports of the bands, rations, the principal component analysis, ACI and the classification that allow the realization of a lithological and mineralogical mapping as well as maps of lineaments by means of directional filters. The validity of the models is tested by comparison with field data and geological maps of the studied site.