Advanced Research On Airborne Spotlight Sar Image Formation Algorithms

An advanced study on spotlight SAR image formation algorithms is presented in thispaper. Several processing algorithms up to now have been studied and implemented indifferent levels, and the latest results in this field are studied in detail. Some work based onreference [12], which comprehensively discussed the main old processing approaches before1995. Of course, some new methods and improvements of old algorithms are proposed in thispaper with a view to the situation of high squinted angle and computation efficiency.Firstly, a converting algorithm for getting simulated spotlight SAR echo data from widebeam stripmap SAR echo data is proposed. By using this approach, spotlight imagingalgorithms can been studied when spotlight data aren't accessible. As to Spotlight SAR polarformation algorithm (PFA), Chirp Z transform is used in the procedure of azimuthinterpolation to improve computation efficiency, and a method is introduced and implementedto correct the geometry position distortion of PFA. A new PFA using azimuth-nonuniforminterpolation is presented to adapt to high squinted angle. Theoretically, this algorithm canproduce image with the same quality for squint image geometry as for broadside. The timedomain Covolution BackProjection (CBP) algorithm is studied and the tomographicformulation of spotlight SAR is discussed. Then, a new fast CBP image formation algorithmis proposed. As to Range Migration Algorithm (RMA) and Chirp Scaling Algorithm (CSA),the implementation of them in spotlight SAR is presented in detail, and exactly mathexpressions of each signal domain are calculated. How to use subaperture extended CSA toprocess spotlight SAR data is studied and some phase functions are improved to adapt tosquinted situation. Based on reference paper [78], an improved Frequency Scaling Algorithm(FSA) is proposed to process high squinted spotlight SAR data.Lastly, the motion compensation system of SAR is simply discussed. The PGA (PhaseGradient Autofocus) technique, an important autofocusing algorithm is studied. Somecomputation questions are discussed, and some feasible approaches are proposed. Beside allof these, unfocused compression of stripmap SAR is analyzed in frequency domain and a newquick-look imaging algorithm is proposed.