Sar Imaging Based On Compressed Sensing

Synthetic Aperture Radar (SAR) is carried on a platform like a plane or a satellite to image the ground. Due to the low computational resources of the acquisition platforms, the data cannot generally be processed on board and must be stored or transmitted to the ground where the image formation process is performed. With the imaged region larger and larger and the resolution higher and higher, more and more data need to be acquired and processed. This makes the implementation of SAR systems more and more difficult. Recently, a theory known as Compressed Sensing (CS) is proposed. It indicates that a sparse signal can be recovered with a high probability from the measurements less than usual. Currently, the theory has been widely used in SAR imaging and other areas to release the burden of system on huge amount of data sampling, storage and transmission.Although the research of SAR imaging based on CS has made some progress, there is still lack of systemic research on SAR imaging based on CS, and no two-dimensional CS based SAR imaging algorithm. In the thesis, the theory and algorithms of CS based SAR imaging is discussed and combined. The major works include the following three parts: the CS based echoes acquisition methods, the CS based SAR imaging algorithms and the application of CS in SAR imaging. Firstly, we establish the sparse representation models of echoes after analyzing the signals'features and propose a phase-reservation two-dimensional CS based SAR imaging algorithm. The effectiveness of the proposed algorithms is tested through processing both simulation and real data.The major contributions of the thesis are summarized as follows: In the study of CS based echoes acquisition methods, we firstly analyze the echoes, and then establish the sparse representation models of the processed signals.In the study of CS based SAR imaging algorithms, we firstly select the sparse signal reconstruction algorithm suitable for SAR imaging and improve it, then propose a phase-reservation two-dimensional CS based SAR imaging algorithm combined with sparse representation of echoes and non-correlation measurement matrix.In the study of CS based SAR imaging algorithm, we realize SAR imaging based on CS.