Study Of High Resolution Spotlight Synthetic Aperture Radar Imaging

High resolution imaging is very important for airborne synthetic aperture radar (SAR) application. To achieve high resolution SAR imaging, spotlight mode is widely utilized in real SAR activities. Spotlight SAR can achieve very high resolution in azimuth by illuminating an interesting area over very long observing time, and thus it provides very wide Doppler bandwidth. High resolution provided by spotlight mode is essential in SAR real applications, such as battlefield surveillance and reconnaissance, and high resolution targets feature extraction and recognition. Polar format algorithm (PFA) is frequently utilized in spotlight SAR imaging, due to its easiness to implementation and low computational cast. Focusing on the implementation of PFA, we study the application of Chirp-Z transform to transfer polar format to rectangular format. With the help of Chirp-Z transform, PFA can be much more efficient. We also study the wide swath spotlight SAR imaging algorithm with spectral analysis (SPECAN) in detail. SPECAN convolution is utilized in azimuth to suppressing Doppler ambiguity first, then conventional high resolution SAR imaging algorithm can be used to generate imagery with high accuracy, such as range migration algorithm (RMA) and chirp scaling algorithm (CSA). This algorithm is free of the limitation of swath size, thus suitable to high resolution and wide swath spotlight SAR imaging applications. The main content of this dissertation is summarized as follows:1,We study the processing flow of PFA and deduce its limitation on imaging swath size mathematically;2,To enhance the efficiency of PFA, the implementation of PFA with Chirp-Z transform is analyzed and discussed in detail, and simulated and real data experiments are given to confirm its validity;3,In terms of overcoming the imaging area limitation of polar format algorithm, we present processing algorithm for wide-swath spotlight SAR data based on SPECAN. SPECAN is used in azimuth to eliminate azimuth spectrum folding effect and migrate and scale the azimuth data support. Then we compensates azimuth quadratic phase and recovers the ideal data support via appropriate interpolation. Finally well focused imagery can be achieved with a two-dimensional inverse Fourier transform.