Visual Geo-Localization and Camera Pose Estimation for Computational Photography
Current digital cameras do not only capture light, they in fact compute pictures. In this talk I will introduce the field of computational photography, which evolved from computer graphics, image processing, computer vision, and optics to extend the capabilities of photography and display technology. Then, I will focus on our recent work on automatic visual image geo-localization and camera pose estimation for computational photography applications. Our aim is to accurately find the location and orientation of the camera which captured the photograph. We introduce a system for automatic alignment of the query photo with a geo-referenced 3D terrain model. Sufficiently accurate match between a photograph and a 3D model offers new possibilities for image enhancement. Furthermore, it can be used to transform photographs into a realistic virtual 3D experience, e.g. to automatically highlight elements in the image, such as the travel path taken, names of mountains, or other landmarks. The synthetic depth map, and/or the whole 3D model of the queried photo, can be used for novel view synthesis, image relighting, dehazing or refocusing.
Martin Čadík is an associate professor at Brno University of Technology (Czech Republic) where he heads the computational photography research group – CPhoto@FIT. His fields of competence cover computer vision, computer graphics, human perception and image processing. His current research interests are on computational photography, visual geo-localization, camera pose estimation, and image quality assessment. Martin Čadík obtained a PhD in computer science from the Czech Technical University in Prague in 2008. He worked as a post-doc research fellow at Max-Planck Institute for Informatics in Saarbrücken (Germany) from 2009 to 2013. Martin Čadík joined the Brno University of Technology in 2013.