At the end of the 1990s the emergence of high resolution (1. m) digital elevation models (DEMs) settled the context of high precision geomorphological analysis. These new elevation models permitted to reveal structures that remained heretofore undetectable. Earth scientists henceforth benefit from a source of data with a textural detail that was never attained before. Despite its richness, this information must be treated efficiently to extract features helping geomorphologists to analyze the processes occurring at the surface. Such processes are complex, localized and naturally multiscale. Recently, space-frequential descriptors as wavelets have been proposed successfully for the analysis of DEMs. The wavelet transform is widely used in image processing since it allows us to decompose a signal into a weighted sum of atoms with joint space-frequency localization. Such a decomposition facilitates a coherent navigation from scale to scale, but also permits to detect heretofore undiscerned phenomena at different scales. This is appealing in geomorphology, where structural components, related to a given phenomenon, are well determined in these sub-spaces specific to the scale continuum. In this paper, we propose a filtering procedure of the wavelet decomposition as an approach for the analysis of geomorphological multiscale structures. This filtering procedure enhances the high pass information contained at each scale. The proposed bottom-up approach is applied here to a case study to detect geomorphological structural elements in a valley of the Swiss Jura. It demonstrates that the proposed filtering procedure is an efficient tool for geomorphological multiscale generalization.
- Digital elevation model (DEM)
- High resolution