Rainforests especially those of the Congo Basin, are ecosystems rich biodiversity. The forest wealth is for countries such as Gabon an economic force based on the exploitation of species of great interest as Okoumé (Aucoumea klaineana), the Ozouga (Sacoglottis gabonensis) and Ozigo (Dacryodes buttneri). There Gabon’s forest biodiversity is threatened by direct human pressure and ongoing climate change. Improved management of this resource is needed by tools and advanced remote sensing methods.
The objective of this thesis is to test and adapt FOTO method (Fourier-based Textural Ordination) of textural analysis to discrimination and recognition of species from optical images with very high spatial resolution (6.7 Spot, Google Earth images); but also to understand and know more precisely the elements of the structure of these essences influence on the texture. In other words, how the canopy structure interacts with the physical signals to produce a given texture from which discriminate between different species. This project is part of the great general problem of deforestation and protection the environment is a priority of the Gabonese State. We hope to bring in this thesis, effective methods to better mapping of forest resources Gabon.