Processing of remote sensing data for the estimation of rock block size distribution in landslide deposits

In rock mechanics, the size of the blocks is a fundamental parameter for several applications, i.e. rock mass classification (GSI—Geological Strength Index, RMR—Rock Mass Rating), stability analysis, blasting design and hydrogeological characterization of fractured media. Moreover, landslide deposits inventories, to be used for the improvement of rock fall predictive analyses, are not easy to obtain. Over the last decades, remote sensing techniques, e.g. photogrammetry, laser scanning, increased the availability of geospatial data for rock mechanics applications. In the present work, rock fall and rock topple deposits were analyzed by means of in-situ surveys and aerial orthophoto visual mapping in GIS environment. Furthermore, orthophotos were processed using an object-oriented classification in order to automatize the blocks dimension extraction procedure. Moreover, limits and capabilities of the objectoriented approach are discussed, in relation to the available data source.