An Approach to Ground Moving Target Indication Using Multiple Resolutions of Multilook Synthetic Aperture Radar Images

Abstract

Ground moving target indication (GMTI) using multiple resolutions of synthetic aperture radar (SAR) images to estimate the clutter scattering statistics is shown to outperform conventional sample matrix inversion space-time adaptive processing GMTI techniques when jamming is not present. A SAR image provides an estimate of scattering from nonmoving targets in the form of a clutter scattering covariance matrix for the GMTI optimum processor. Since the homogeneity of the scattering statistics are unknown, using SAR images at multiple spatial resolutions to estimate the clutter scattering statistics results in more confidence in the final detection decision. Two approaches to calculating the multiple SAR resolutions are investigated. Multiple resolution filter bank smoothing of the full-resolution SAR image is shown to outperform an innovative approach to multilook SAR imaging. The multilook SAR images are calculated from a single measurement vector partitioned base on synthetic sensor locations determined via eigenanalysis of the radar measurement parameters.