A resolution enhancement method for SAR images with adaptive speckle reduction

Abstract

Due to the capability of producing a high-resolution image under all weather conditions, SAR (synthetic aperture radar) is a very powerful and attractive tool for remote sensing. Since SAR systems generate images by coherent processing of the scattered signals, they are susceptible to speckle noise. Speckle noise is multiplicative so the presence of speckle noise hinders a human observer from analyzing features of SAR images. Also, speckle noise makes automatic interpretation of SAR images extremely difficult. Various methods have been proposed to suppress speckle noise. Although conventional speckle filtering algorithms can effectively reduce speckle noise, but also smear edges and blur images to some degree. Because SAR system has a limited bandwidth, even if no speckle suppression filter is applied for obtaining fine-detail images, there still exist slow responses to sudden changes, i.e. sidelobes. We propose a new SAR image enhancement method, which reduces sidelobes in fine details as well as suppresses speckle noise effectively in homogeneous areas. To verify the efficacy of the proposed method we apply it to an ERS-1 SAR image. The resulting image shows that the proposed method suppresses speckle noise effectively without any blurs or sidelobes in fine detail areas.