Research On Linear Array Three-dimensional SAR Sparse Imaging Method

As a type of new 3D SAR imaging model,the linear array 3D SAR not only overcomes the shortcomings of traditional 2D SAR imaging,which including the shadow effect,spatial ambiguity,top and bottom inversion,but also has the advantage of easy to control movement compared with the circular SAR and Tomography SAR.So has been widely used in military and civilian fields.Due to the constraints of the Nyquist sampling theorem and the limitations of the traditional SAR filtering algorithm based on matching filtering,the linear array 3D SAR imaging has the disadvantages of huge amount of echo data and low image resolution.The compressed sensing theory states that if the signal is sparse or compressible,the original signal can be recovered with a sampling rate lower than the Nyquist sampling theorem.In 3D SAR imaging,the target scene is often sparse,so the compressed sensing theory has a great application in improving the resolution of SAR imaging and reducing the amount of SAR echo data.In this thesis,the theory of compressed sensing is applied to linear array SAR,and the method of 3D SAR imaging based on sparse reconstruction is studied.The specific research contents and innovations are as follows:1.The basic theory of linear array 3D SAR imaging and compressed sensing is introduced.Firstly,the geometric model of linear array 3D SAR is introduced,and its echo signal model is given.Secondly,based on the fuzzy function,the resolution limit of the traditional imaging algorithm is analyzed,and two kinds of traditional SAR imaging algorithms are introduced.Finally,the basic theory of compressed sensing is introduced,and the linear observation model of the linear array SAR is analyzed,which provides a theoretical basis for the application of the compressed sensing theory to the linear array SAR imaging.2.The SAR imaging method based on greedy algorithm is studied.Firstly,the process of Orthogonal Matching Pursuit(OMP)algorithm is described,which has been widely used because of its simple process,high operation efficiency and small reconstruction error.Secondly,some improvements to the OMP algorithm by scholars in recent years are summarized.These algorithms mainly improve from the criteria of element selection,coefficient updating,candidate set number and so on.Finally,the Gradient Pursuit(GP)algorithm is studied emphatically,which has the advantages of high efficiency and low spatial storage compared with the OMP algorithm.3.A Threshold-Based Gradient Pursuit(TBGP)algorithm is proposed.GP algorithm needs to set the sparsity level of the scene in advance,but the sparsity level in SAR imaging usually unknown.Aiming at the problem,the TBGP algorithm is proposed.The algorithm uses the maximum minimum scattering coefficient ratio and the change rate of the scattering coefficient as the criterion for iterative termination instead of the sparsity level.The algorithm not only retains the advantages of GP algorithm in computing time and space storage,but also overcomes the disadvantage of GP algorithm which needs to set the sparsity level in advance.Simulation and experimental data verify the above conclusions.4.Sparsity Bayesian Recovery via Iterative Minimum(SBRIM)algorithm is studied.The principle and procedure of SBRIM algorithm are analyzed.As a kind of sparse Bayesian algorithm,compared with other algorithms,it has the characteristics of high flexibility and high precision of reconstruction.The simulation results show that the SBRIM algorithm has high resolution imaging ability compared with the traditional BP algorithm.5.A Weighted Sparsity Bayesian Recovery via Iterative Minimum(WSBRIM)algorithm is proposed.Unlike the SBRIM algorithm,the WSBRIM algorithm uses a weighted constraint on L1 norm in cost function,and uses a different function to approximate L1 norm,reducing the amount of computation.The principle and procedure of the WSBRIM algorithm are described in detail.The simulation experiments and performance comparison of the related algorithms are made.The simulation results show that the reconstruction error of WSBRIM algorithm is the least,and the reconstruction precision is the best.In the efficiency of operation,the TBGP algorithm mentioned in this thesis has the advantage.Finally,the 3D SAR data results verify the validity of the WSBRIM algorithm in the real scene.