Analysis of georeferenced landscape pictures extracted from public picture collections

National audienceToday photo-sharing websites allow citizen to share their pictures. Pictures taken by sensor-equipped devices such as smartphones or tablet-computers keep the information where (in terms of coordinates and azimuth angle), when and with which device a picture has been taken.We have developed a method that uses the coordinates of the camera position as well as the azimuth angle, the focal length and the crop factor in order to calculate a 3D field of view of a picture using a digital terrain model (DTM). This field of view can thereafter be used for the determination of interest in spatial features. Similar methods that are based on viewsheds have been described by Tenerelli et al. (2017) and Sherren et al. (2011).In a case study involving more than 3’000 georeferenced landscape pictures extracted from the public photo-sharing website flickr (flickr.com) we investigate the potential of the method. The results show that the method can yield interesting results for decision-making tools that take into account landscape aesthetics