Research On Airborne SAR Imaging Algorithm Based On Measured Data

Synthetic aperture radar (SAR) enhances radar's information acquisition capability and information awareness ability by obtaining high resolution images. Now, airborne SAR has already become a common method for acquiring ground images. In this paper, imaging formation algorithm has been researched and implemented via two sets of airborne SAR data in laboratory. Their application conditions and advantages have been summarized based on the comparison of every algorithm effect. It is of realistic significance.This paper is divided into six chapters. Firstly, we go over the basic theory of airborne SAR, including the principle of SAR, range and azimuth resolution, range migration and principle of the stationary phase. Then, started with traditional RD algorithm, imaging formation mechanism was explored and point-object simulation and real data imaging have been realized. Thirdly, each imaging formation has been accomplished after making a thorough study on standard CS algorithm, nonlinear CS algorithm, RMA algorithm, multi-viewport algorithm and super-resolution algorithm which is a frontier technology at present. From imaging formation process, due to the little range migration for side-looking mode, high quality two-dimensional target imaging can be acquired by using each algorithm. In the case of small squint angle, RD algorithm with second range compression (SRC) also has good performance, but it requires large amount of computation for interpolation. For large squint angle case, RD algorithm may be unable to image and meanwhile ideal RMA algorithm which has best performance in theory brings huge amount of calculation for high interpolation precision in frequency domain. As a result, nonlinear CS imaging effect can be regarded as the best result and RMA algorithm has similar imaging effect with standard CS algorithm.Although the two sets of measured data belong to side-looking mode scene image, airplane will be inevitably influenced by many factors during the flight, which can cause imaging quality decline. For better focus on azimuth, after researching the effects of various phase error on point-object impulse response, spatial spectrum energy equalization method and aperture correlation method have been applied for the estimation of the Doppler centroid frequency and Doppler rate. Finally, airborne velocity and squint angle of antenna phase centre deviation can be calculated by Doppler parameters and movement error compensation can be compensated.In conclusion, we briefly discuss evaluating criteria of image quality and analyze each imaging effect form many aspects and draw conclusion according to criteria value. This paper makes systematic radar imaging method study on two sets of real data and accumulates relevant experience at large scale imaging formation application.