Adaptive Spatial Multi-looking of Differential SAR Interferograms Sequences using Circular Statistic

摘要

In this work, we discuss the potential of a new adaptive multi-looking technique that operates on a sequences of multitemporal differential SAR (DInSAR) interferograms. The core of the method is focalised on the use of circular (directional) statistics. Unlike other methods recently proposed in the literature, which identify homogenous distribuited scatterers (DS) that are present into a resolution cell by analysing statistics of the complex valued single-look-complex (SLC) SAR images, the proposed method is exclusively based on the analysis of hig resolution interferograms. It does not need of any other type of assumption, and it uses the two sample Kuiper test to check the statistical homogeneity of neighbouring SAR pixels. The technique can be applied to generate adaptive multi-look interferograms both at the native grid of full-resolutions images and at the multi-look resolution scale. Also, the method can be simply complemented with other space/time noise-filtering techniques, such as the one adopted within the Extended Minimum Cost Flow (EMCF) based Small Baseline Subset (SBAS) processing chain, based on the optimal combination of multi-look interferograms. The experimental results demonstrate the validity of the method.