SAR image enhancement using noninteger Nyquist SVA technique

In SAR and inverse SAR (ISAR) imaging, conventional Fourier transform (FT) based image reconstruction techniques result in images with limited resolution. The down-range and cross-range resolutions of these algorithms are inversely proportional to the radar signal waveform bandwidth and to the synthetic aperture size, respectively. On the other hand, when modem spectral estimation methods are applied to radar imaging, these nonlinear techniques, usually called super resolution algorithms, offer improved resolution, better contrast, and reduced speckle. Spatially variant apodization (SVA) is a nonlinear filtering operation which significantly reduces the sidelobe levels without degrading mainlobe resolution of the sinc impulse response. In this work, we propose a modified version of noninteger Nyquist SVA and develop an iterative super SVA procedure for SAR and ISAR image enhancement. The proposed technique was successfully applied to various SAR/ISAR images.