PSF estimation in SAR imagery restoration based on corner reflectors

Abstract

Point Spread Function (PSF) estimation is a key issue influencing the quality of imagery restoration. There are many PSF estimation methods for optical imagery but few for SAR imagery. This paper puts forward a new method of PSF estimation for restoration of SAR imagery. By setting corner reflectors in the field, according to the brightness and scattering features, the PSF of SAR imagery can be simulated adopting the elliptic paraboloid model; therefore, the parameters of the model can be deduced. According to the model mentioned above, SAR imagery restoration experiments were carried out. The tests show that the restoration quality using PSF that based on the elliptic paraboloid model is better than that derived from the blind deconvolution estimation. The algorithm of the model is simple and the quality of the restored SAR imagery is enhanced. Therefore, this PSF estimation method is feasible.