Improving SAR Tomography Urban Area Imaging and Monitoring with CAESAR

In this paper we investigate the use of an option of a method recently proposed in the literature in the context of multibaseline/multipass SAR interferometry (CAESAR) for improving the performances of SAR Tomography imaging and monitoring of urban area. The proposed solution is based on a preliminary data filtering stage available in CAESAR, based on the estimation of the covariance matrix, applied before classical tomographic processing, that is a beamforming reconstruction and Generalized Likelihood Ratio Test detection. The procedure allows effectively counteracting the decorrelation effects affecting distributed scatterers as well as coherence losses due to sparse temporal sampling. Experiments on real data acquired by the Cosmo-Skymed constellation allows validating the effectiveness of the proposed multilook tomographic approach. They demonstrate the dramatic improvement achievable in terms of spatial density of measurements with respect to classical single look SAR Tomography at the price of a limited spatial resolution loss.