Research On 3D Imaging Algorithm Of Multi-baseline Tomography Sar

Multi-baseline tomography SAR(MB-TomoSAR)3D imaging technology is an effective method to solve the layover problem of 2D SAR imaging.Compared with other imaging systems that perform 3D processing in the echo data domain such as Linear Array SAR(LASAR)and Curvilinear SAR(CLSAR),it has the advantages of simple implementation,small amount of calculation,and no need to change the existing imaging system and imaging algorithm,which has a wide range of applications in terrain mapping,military reconnaissance,resource exploration and vertical inversion of surface vegetation.MB-TomoSAR applies the synthetic aperture principle into the elevation direction to achieve 3D imaging.However,how to achieve high-precision imaging under nonuniform flight distribution and sparse flight distribution is the main problem.Therefore,this paper studies the high-precision MB-TomoSAR imaging algorithm based on BP and CS theory.The main work and innovation of the paper are as follows:1.Study the basic theory of MB-TomoSAR imaging.First,the 3D imaging theory of MB-TomoSAR is elaborated,including the construction of geometric models,the derivation mathematical model and the introduction of the processing flow;Secondly,the factors that affect the elevation resolution and the maximum range of unambiguous imaging are analyzed,and the mutual restriction between the number of flights and the distance between elevation resolution and imaging blur is explained;Then,aiming at the diversity of the flight distribution,the MB-TomoSAR imaging under the non-uniform flight distribution is discussed.Finally,simulation experiments are used to illustrate the effects of flight interval,number of flights and flight distribution on imaging,and point out the problems caused by traditional non-uniform distribution and sparse distribution on traditional TomoSAR imaging algorithms.2.Study the MB-TomoSAR imaging algorithm based on BP theory aiming at the problem of realizing 3D imaging under non-uniform flight distribution,First,the defects of the imaging algorithm based on the interpolation idea are analyzed with low accuracy,and then the array 3DBP algorithm is studied.It is pointed out that although the algorithm can perform high-precision imaging,its operation efficiency is not high,and is not suitable for processing the image sequences;Secondly,to solve the problems of low imaging accuracy and low operation efficiency,a fast MB-TomoSAR imaging algorithm based on BP is proposed.The algorithm introduces the idea of backward projection,and the image sequence acquired by multiple flights is used for image registration,and then the phase compensation is performed on each flight data to perform coherent accumulation to achieve 3D high-precision imaging.Finally,the problem of geometric shift caused by MB-TomoSAR image is analyzed,and the correction method of geometric shift is given.The processing results of simulation and measured data verify the effectiveness of the proposed algorithm.Compared with the traditional TomoSAR imaging algorithm,this algorithm can not only realize the imaging under the non-uniform flight distribution,but also have higher imaging accuracy.3.Aiming at the problem of high-precision imaging of MB-TomoSAR under sparse flight distribution,a MB-TomoSAR imaging algorithm based on CS is studied.First,the basic theory of compressed sensing is briefly introduced,and the linear measurement model of MB-TomoSAR imaging is derived and analyzed.Second,based on the optimization solution of the linear measurement model,the classic layer based on Orthogonal Matching Pursuit(OMP)is studied.The algorithm combines greedy iteration and orthogonal matching ideas to achieve the optimal estimation of signals by iterating orthogonally.It has the advantages of simple process implementation and high operating efficiency,but at the same time,its realization requires preset sparseness.And the elevation reconstruction accuracy is not high.Compared with the OMP algorithm,the Sparsity Bayesian Recovery via Iterative Minimum(SBRIM)algorithm does not require preset sparseness,and the reconstruction accuracy is higher.Therefore,this paper proposes a MB-TomoSAR imaging algorithm based on SBRIM.Based on Bayesian theory,the method uses the prior probability and other information of the elevation signal to model,and then iteratively estimates the optimal solution of the elevation signal.The simulation and measured data verify the effectiveness of the proposed algorithm,and show that the proposed algorithm can achieve high-precision imaging,and is not sensitive to noise.