3 D Synthetic Aperture Leidachao Resolution Imaging Methods

The linear array synthetic aperture radar (LASAR) is a new three dimensional (3D)SAR system, which uses a linear array antenna mounted in traverse direction tosynthesize a2D virtural antenna array with the motion of the platform to achieve thealong track and the cross track resolution; combing with the pulse compressiontechnique in the range direction, the3D resolutions are obtained. At present, the widelyused methods in3D SAR system are3D-BP method and3D-RD method. Thesemethods are based on coherent accumulation. Besides, to traditional method, restrictedby the Rayleigh limit, the resolution of along-track direction is difficult to rise. Toovercome these limitations, this dissertation applies super-resolution algorithm incross-track direction and along-track direction instead of traditional matched filtermethods.The main content is:(1) This dissertation introduced different modes of LASAR system, and the3D spatialambiguity function (AF) is analyzed. Then the3D spatial resolutions are derived.Lastly, matrix model of echo is seted up.(2) The new imaging method based on MUSIC algorithm is applied. Two drawbacksneed to be taken into account for the MUSIC algorithm applied in LASAR imaging.The first, in some cases, the reflection coefficients of the scatter points arecorrelated and the covariance matrix is usually singular, so it leads to the failure ofMUSIC algorithm. The second, MUSIC algorithm is only usable to localize thetarget location but not to measure its backscattering power, so this method can notobtain the scattering information of targets. To overcome the first of theaforementioned drawbacks, this paper applies an improved spatial smoothingtechnique to reduce the relativity of the reflection coefficient and constructcovariance matrix, and to overcome the second drawback, by using the least square(LS) algorithm to obtain the final scatter backscattering information.(3) The new imaging method based on IAA algorithm is applied. The presented newimaging method can greatly improve imaging resolution and suppress the side-lobefor LASAR3D imaging in cross-track direction and along-track direction. Thesimulated results confirm the validity of the method and show that it can improvethe imaging quality. The method is used in measured data and proved advantagesand feasibility.(4) According to sparse characteristics of3D scene, the fast3D IAA method is applied.This dissertation proved the new method can greatly reduce the computational cost.(5) The new imaging method based on CS algorithm is applied. The presented newimaging method can greatly improve imaging resolution and suppress the side-lobe for LASAR3D imaging in cross-track direction and along-track direction. Thesimulated results confirm the validity of the method and show that it can improvethe imaging quality. The method is used in measured data and proved advantagesand feasibility.(6) At last this dissertation compared performance of the several super-resolution3Dimaging method.In a word, the research results are useful for the design of LASAR system andimaging algorithms.