SAR imaging of dominant scatterers using cascading StOMP

We propose an algorithm for synthetic aperture radar (SAR) imaging of dominant scatterers based on the theory of Compressive Sensing (CS). The SAR raw data is randomly sampled in two-dimensional, and the pulse compressing and azimuth focusing are implemented by stagewise orthogonal matching pursuit (StOMP). In order to preserve all energy of lower scatterers, two stage thresholds are designed while implementing pulse compressing. The experiments on real SAR data show that wide-scene image can be constructed using less data, without the loss of resolution.