Matching of Persistent Scatterers to buildings

• Main Authors: Schunert, A., Schack, L., Soergel, U.
• Format: Conference Proceeding
• Language: English
• Subjects: Buildings; Google; Iterative closest point algorithm; Monitoring; Radar interferometry; Solid modeling; Synthetic aperture radar; Urban areas
• ISSN: 2153-6996; 2153-7003
• DOI: 10.1109/IGARSS.2012.6352309

Persistent Scatterer Interferometry (PSI) aims at estimating the topography and deformation for a set of sufficiently stable radar targets, referred to as Persistent Scatterers (PS). Thereby, a stack of SAR images is exploited to distinguish between signal of interest and disturbances caused by, for instance, changes of the atmospheric conditions over time. The achievable sampling density mainly depends on the resolution of the employed sensor, and the characteristics of the scene at hand (i.e. the number of sufficiently stable reflectors). In case space-borne SAR data of the highest resolution is used, very high PS densities, potentially enabling the mapping of structures at sub-building level, can be achieved. This offers a big and completely new range of possible applications. For instance, it is possible to monitor single buildings using PSI. But with this also new questions and challenges arise. Two of those would be: which buildings are sampled dense enough for monitoring purposes, and to which actual building structure does the PS correspond? In order to investigate those questions, the PS results are aligned with building outlines. This easily enables to map the PS density of buildings, which partly answers the first question. Admittedly, also the distribution of the PS and the shape of the building as well as its structure play a role. But a map of the PS density would be definitely helpful to assess which structures can be monitored adequately. The correspondence between actual building structures and PS is a good deal harder to answer. In this work, we attempt to improve the geocoding of the PS with the help of the GIS data, in view of the fact that this may be very helpful for the given task.