Automated detection of permanent scatterers in radar interferometry: Algorithm and testing results

The time-series differential radar interferometry with Permanent Scatterers (PS) can overcome the limitation of temporal decorrelation, and thus enhance the reliability of radar interferometry when applied in regional deformation mapping. How to detect automatically PS from SAR images is one of crucial procedures in data reduction of the PS interferometric system. With theoretical and experimental analysis, this paper proposes a Dual-Thresholding Algorithm (DTA) for automated identification of PS. Since the DTA is of the combination of both correlation-coefficient thresholding and amplitude dispersion index thresholding, it can save computation time and keep reliable for PS identification. Using 26 C-band SAR images collected from 1992 to 2002 over Shanghai by satellites ERS-1 and ERS-2 of European Space Agency, the DTA has been validated for PS detection. The testing results show that the proposed DTA can identify PS from SAR images effectively and reliably.