A NOVEL SAR SIDELOBE SUPPRESSION METHOD BASED ON APPROXIMATE GREATEST COMMON DIVISORS

Sidelobe suppression is a challenging but crucial issue for SAR image quality enhancement. Inspired by the blind image restoration method based on the approximate greatest common divisor (AGCD) in natural images, a novel SAR sidelobe suppression method is proposed in this paper. Sidelobe suppression is interpreted as image deconvolution with a known PSF. Another SAR image with different PSF is generated by injecting the phase error at first. Then the PSF estimation is introduced into SAR, which is realized by solving the AGCD of two noised polynomials associated with two SAR images. Finally, the sidelobe suppression reduces to the solution of the inverse problem. Experiments on a TerraSAR-X SAR image show that improved results are qualitatively realized in sidelobe suppression and detail preservation in comparison with two existing algorithms.