摘要

A model-based adaptive synthetic aperture radar (SAR) image formation algorithm is proposed in this study. A parametric model is used to formulate the received SAR signal, each scatterer can be characterised by a set of its parameters. In the model, the instantaneous slant range is approximated by a polynomial. The parameters of each scatterer are estimated by maximising the cost function, named matching degree indicator. The root-mean-square errors of parameter estimation are close to the Cramer-Rao bounds. The image is formed by mapping the estimated parameters onto the grid formed by iso-range and iso-Doppler lines. The algorithm does not require explicit knowledge of the data collection geometry. Experimental results are given to validate the algorithm.