Landslide recognition and monitoring with remotely sensed data from passive optical sensors

Landslide inventory mapping and monitoring are indispensable for hazard assessment and disaster management. The enhanced availability of VHR satellites, UAVs and consumer grade digital cameras offers a great potential to support those tasks at regional and local scales, and to complement established techniques such as in situ instrumentation, radar, andlaser scanning. A lack of image processing tools for the efficient extraction process-relevant information from different types of optical imagery still complicates the exploitation of optical data and hinders the implementation of operational services. This doctoral thesis is dedicated to the development and application of image processing techniques for the mapping,characterization and monitoring of landslides with optical remote sensing data. A comprehensive review of innovative remote sensing techniques for landslide monitoring shows the potential and limitations of available techniques and guides the selection of the most appropriate combination of sensors – platforms – image analysis methods according to the observed process and end-user needs. For the efficient detection of landslides after major triggering events at the regional scale, a method for rapid mapping combining image segmentation, feature extraction, supervised learning is developed. For detailed landslide investigations at the local scale, this study elaborates image processing chains for detection of surface fissures in time-series of UAV images as geo-indicators of landslide activity, the measurement of horizontal surface displacements from VHR satellite images using stereo-photogrammetric and image correlation methods, and 3D measurements from terrestrial photographs based on multi-view open-source photogrammetry.