The project “SafeEarth” deals with the broad spectrum of exploiting remote sensing data in terms of disaster management. We focus in particular on the automatic detection, reconstruction and tracking of infrastructure and relevant objects like roads, buildings, vehicles and people due to their importance for the immediate planning and implementation of emergency actions. Special focus is dedicated to the fast, integrated, geometrically and semantically correct interpretation of airborne and spaceborne images. The exploitation of complementary information from multi-sensorial images, i.e. optical and radar data, leads to synergistic effects enhancing the interpretation of objects. The analysis of multitemporal images or image sequences is important to draw conclusions of the temporal development of disasters concerning several days (e.g. flooding, earthquakes, landslides) or to monitor special events and infrastructure continuously (people tracking, event detection, vehicle tracking), respectively.
To this end, three essential components of image understanding science will be brought together, which reflect also the expertise of the cooperating partners: i) Sensor technology and physical sensor modeling (DLR); ii) Mathematical and statistical modeling in image processing (DTU); iii) automatic image understanding and object extraction (TUM). The comprehensive and complementary expertise of the partners involved in this project will serve as basis for an intensive exchange with the goal to integrate the different research activities in a synergistic fashion under the thematic umbrella of the project.