Spaceborne tomographic SAR techniques for analyzing garbled urban scenarios: Array processing advances and experiments

摘要

SAR Tomography (Tomo-SAR) is an experimental “coherent data combination” mode allowing full 3-D imaging of complex urban and infrastructure scenarios with layover (“garbled”) scatterers, exploiting multibaseline interferometric SAR data stacks. Also deformation monitoring in garbled scenarios is possible, with the more general recently introduced Differential Tomography (Diff-Tomo) framework. Various approaches have been proposed to improve Fourier-based Tomo-SAR which is affected by unsatisfactory height sidelobe behaviour and resolution, due to the typical low number of baselines with irregular distribution. Among these approaches, superresolution multilook beamforming techniques proved to posses interesting capabilities, at the cost of operation with reduced horizontal resolution. In this work, the recently proposed knowledge-based baseline interpolation and a superresolution (Capon) method are integrated in a new Tomo-SAR processor able to offer height superresolution and sidelobe cleaning with single-look data, allowing full resolution operation, which is important in urban areas. Tests are reported with real ERS data.