EVALUATION OF A PSI-BASED CHANGE DETECTION REGARDING SIMULATION, COMPARISON, AND APPLICATION
- Institute for Photogrammetry, University of Stuttgart, Stuttgart, Germany
Keywords: Synthetic Aperture Radar (SAR), Persistent Scatterer Interferometry (PSI), Change Detection, Time Series Analysis, Urban Monitoring
Abstract. Persistent scatterer interferometry (PSI) detects and analyses strong, stable, and coherent radar signals throughout a time series of SAR images. Such coherent signals are reflected from corner-reflector-like substructures in built-up cities, which are regarded as so-called PS points. Certain PS properties such as deformation velocity and topography height can be derived for scene monitoring. Previously, we introduced a PSI-based change detection to detect disappearing and emerging PS points along with their occurrence times. Such temporary PS points existing only during a certain period correspond to change events, e.g., mostly constructions in cities. The tests using TerraSAR-X images successfully identified where and when the construction events in Berlin took place in 2013. The results were compared and all agreed with the ground truth. In this study, we evaluate our method more deeply. A simulation test is conducted to evaluate the theoretical accuracy in space and time. We also compare our method with two classical approaches: image rationing and amplitude-based semi-PS detection. The computational requirements are revealed afterwards. Finally, potential applications are proposed and discussed. All of these works help us to better characterize our technique and learn the pros and cons.