Scattering centre extraction for extended targets

In this paper, we consider the estimation of scattering centre parameters for the case of extended targets. In general, models are tested on synthetic targets made up of few scatterers. This choice is imposed by the maximum model-order assumed by the method. It is limited by the rank of the observation matrix and by the noise which corrupts the signal subspace. However real-world systems have to deal with complex targets such as aircrafts. For this reason, we revisit the theory for the case of targets made up of a large number of scatterers. We present a novel method for selecting poles corresponding to true physical scatterers. Finally, we propose a technique for reducing the model order based on a modification to the ISAR technique. The resulting 2D-model results in lower noise compared to traditional techniques and can be used for pole selection, data compression, image enhancement and high resolution ISAR-imaging.