Study on Improvement of Phase Gradient Autofocus Algorithm

Based on the theory of phase gradient autofocus on airborne SAR, an improved algorithm for reducing the phase gradient autofocus computation is provided in this paper. According to the image contrast criterion, some isolated stronger scatters for estimation phase errors are chosen, and then a new way to select windows width is developed, which can accelerate the convergence of the autofocus. Point target simulation is carried out, which verify the feasibility and the validity of this algorithm. The present algorithm can estimate discretional rank phase errors and doesnpsilat require strong reflection point, which is up to accuracy. After decreasing the computation of this algorithm, it is more suitable for real-time processing.