Toward Automated Façades Generation from Oblique Aerial Images

Abstract

In this paper we introduce an approach for automatic recognition and description of building façades from oblique aerial images. This approach does not use street-side images, like Google StreetView, but aerial images to derive photorealistic city models. The paper starts with a short summary of the state of the art in the domain of imagebased modeling, automated recognition and description of façades from (oblique aerial) images. Then the context of the research project is described including the hardware setup and and application requirements for complex façade analysis. Afterwards a novel approach of a grid graph model with traversers is suggested to solve the problem of façade recognition. The framework that implements the approach in a flexible manner is introduced. Finally, the paper ends with a conclusion and an outlook to future works.