Research On SAR Autofocus Algorithms

For SAR, imaging is based on the movement that can cause problems meantime. Airborne plane should go in ideal line in flight in order to get high resolution. But it is difficult to catch the exact way even if high-precision IMU system is installed. Algorithm errors and airflow disturbance also affect imaging quality.In the preface to this dissertation, the basic idea and development trend of SAR are introduced briefly. The main contribution of this dissertation is summarized.Chapter 2 introduces the basic principle of synthetic aperture briefly, exposes SAR geometric relationship and RD imaging algorithm, the result of point target simulation is shown.The position in SAR data processing and classification of motion compensation are introduced and autofocus technique is educed. Autofocus algorithms which are based on range-compressed data and complex image are separately detailed and algorithms using imaging rule are included.In chapter 3 autofocus based on range compressed data is studied. MD algorithm and PD algorithm are analyzed and compared. MAM and DPT algorithm are introduced and simulated.Chapter 4 discusses autofocus based on complex image.The principle of PGA algorithm and its advanced are detailed and classified; experiments are processed to verify the conclusion.Besides, ROPE algorithm is introduced and simulated.In chapter 5,autofocus using imaging rule is analyzed, an autofocus algorithm based on exponential function which can satisfy different scenes by means of adjusting the exponent index is proposed. We assess the performances of the new algorithm using imageries of different contrast; the results are compared with PGA algorithm, the fast minimum entropy phase compensation algorithm and COPA algorithm, respectively. In the end of the chapter, how to use the autofocus algorithms in different scenes is summarized. Chapter 6 summarizes the whole work of this thesis and points out the problems to be solved in the future.