Context-based urban terrain reconstruction from uav-videos for geoinformation applications

摘要: Urban terrain reconstruction has many applications in areas of civil engineering, urban planning, surveillance and defense research. Therefore the needs covering ad-hoc demand performing a close-range with miniaturized relatively inexpensive sensor platforms are constantly growing. Using (miniaturized) unmanned aerial vehicles, (M)UAVs, represents one most attractive alternatives to conventional large-scale imagery. We cover this paper four-step procedure obtaining georeferenced 3D models from video sequences. The four steps – orientation, dense reconstruction, modeling geo-referencing robust, straight-forward, nearly fully-automatic. two last namely, almost-nadir videos co-registration 6ndash; represent main contribution work will therefore be covered more detail. essential substeps third step include digital model (DTM) extraction, segregation buildings vegetation, as well instantiation building tree models. is subdivided into quasi- intrasensorial registration Euclidean reconstructions intersensorial geo-referenced orthophoto. Finally, we present results real data-set outline ideas for future work.