Side-Lobe Cancelation in DInSAR Pixel Selection With SVA

Abstract

Synthetic aperture radar (SAR) systems are inherently band limited in both range and azimuth, and hence, the point spread function (PSF) has the shape of a bidimensional sinc function. In addition, all SAR images are slightly oversampled, and as a consequence, the contribution of a single target extends to more than a single cell. The main lobe and the side lobes of strong scatterers are sometimes clearly visible in the images. This characteristic of the SAR images must be considered when applying differential interferometric synthetic aperture radar (DInSAR) pixel selection algorithms. For persistent scatterers, the properties, for instance, the amplitude stability, are preserved in both redundant information around the main lobe and side lobes. For this reason, a cluster of pixels rather than just the pixel position corresponding to the exact location of the target will be detected. Spatially variant apodization (SVA) is a nonlinear filter based on cosine-on-pedestal weighting functions able to achieve a total side-lobe cancelation without degrading the original image resolution. When working with complex data under complex scattering scenarios, the PSF moves away from the ideal bidimensional sinc, and the SVA performance worsens. The amplitude and phase of the original images could be distorted by the SVA filtering compromising the pixel selection and the quality of the final DInSAR results. In this letter, SVA is used to method locate in the image the side lobes of high-power scatterers and generate a mask while preserving the amplitude and phase of the original images.