Conventional two dimensional(2D)autofocus algorithms didn't take advantage of the priori information on the structure of phase error.They often suffered from lack of redundancy to accurately estimate the 2D phase error.In this paper

a knowledge aided(KA)2D autofocus approach is presented

which is based on exploiting the priori knowledge on the phase error structure.First

as a prerequisite for the proposed KA method

the analytical structure of residual 2D phase error in SAR (Synthetic Aperture Radar)imagery is investigated in the polar format algorithm framework.Then

by incorporating this prior information

a 2D autofocus approach is proposed.The new method only requires an estimate of azimuth 1D phase error

while the 2D phase errors can then be computed directly from the estimated azimuth phase error.Experimental results indicate the effectiveness of the proposed method.