An efficient algorithm for single-/multi-channel SAR superresolution imaging of large scenes

Amplitude and Phase EStimation (APES) has extensively applied in single-/multi-channel SAR/ISAR superresolution imaging. It is superior to traditional Fourier imaging techniques in improving resolution and suppression sidelobe artifacts. However, even implementing with fast algorithms, that spectral estimation for two-dimensional signals, e.g., SAR images of large scenes, is a computational intensive method. The chipping and mosaicing strategy, which is proposed to greatly decrease the computation cost, leads to some spectral estimation errors and redundant computation burden. With quantitative analysis of real measured SAR images and theoretical derivation, we discuss the spectral estimation error that is originated from spectral overlapping and energy leakage. To eliminate the error, we introduce the subband decomposition-based spectral estimation method, and this method has the ability of suppressing spectral overlapping. The effectiveness and efficiency of the new algorithm is demonstrated by real spaceborne polarimetric SAR data processing.