Cornice Detection Using Façade Image and Point Cloud

Abstract

puter vision. The Internet of things is another application for high resolution interpreted 3D scenes. According to LEBERL et al. (2012), information down to 3 cm resolution is needed for virtual navigation through urban spaces. Therefore, also the recognition of small details on building façades will be required for realistic models of the environment. LEBERL et al. (2012) derive large and highly detailed 3D models of urban spaces from various data sources. Aerial images are used to generate coarse building and landscape models, whereas laser scans and terrestrial images from mobile sensing platforms are employed for the determination of models with more detail. If these results can be merged with future results of automated approaches for the interpretation of façades such as HOHMANN et al. (2009), the combination will make automatic