期刊名称：ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences
印刷版ISSN：2194-9042
电子版ISSN：2194-9050
出版年度：2006
卷号：XXXVI Part 4
出版社：Copernicus Publications
摘要：The High Resolution Stereo Camera (HRSC) on Mars Express covers the surface of our neighboring planet in full color and 3D. Altogether, nine channels are obtained in stereo angles between ±18.9°. These data sets are well suited to derive Digital Terrain Mo- dels (DTM) and, based on that, color orthoimages. Both, photogrammetric standard processing of HRSC data as well as alternative and/or supplementary approaches are usually based on image matching or shape-from-shading. In contrast to classical matching between images – and aiming for the combined derivation of orthoimage, DTM, and reflectance properties of the Martian surface – HRSC data are processed using the Facets Stereo Vision approach, which is a flexible method for matching in object space. Transfer functions connect the object surface (surfels) with corresponding pixels in each image; the entire image content is treated as observables. The unknowns, in particular an orthoimage (brightness values) and a DTM (height values), are estimated in a combined least squares adjustment with respect to spatially regular grids in object space, i.e. the facets. Thus, existing interconnections between these models are implicitly regarded. In this context, an indirect approach of Facets Stereo Vision is utilized for the processing of HRSC data. It is implemented with MATLAB and applied for small Martian regions with varying surface characteristics. Exemplarily, the results for a steep crater rim and a comparatively flat region are presented and judged
