We introduce SL1MMER, a spectral estimation algorithm based on L1-norm minimisation, model order selection and final parameter estimation. It combines the advantages of compressive sensing with the amplitude and phase accuracy of linear estimators. Our target application is differential Synthetic Aperture Radar (SAR) tomography. We will also show that by means of our proposed time warp method the tomographic imaging equation with an M-component motion model can be rewritten as an M+1-dimensional spectral estimation problem.     «

We introduce SL1MMER, a spectral estimation algorithm based on L1-norm minimisation, model order selection and final parameter estimation. It combines the advantages of compressive sensing with the amplitude and phase accuracy of linear estimators. Our target application is differential Synthetic Aperture Radar (SAR) tomography. We will also show that by means of our proposed time warp method the tomographic imaging equation with an M-component motion model can be rewritten as an M+1-dimensional...     »