Radar Target Imaging Methods Based On Sparse Representation

The electromagnetic scattering mechanism indicates that the backscattering field of the radar target in the high frequency region can be regarded as generated by several scattering centers with different structures.According to the radar echo data,it is important to extract the parameters of the scattering center and generate the corresponding radar image,which is an important link in the analysis and recognition of high resolution radar target characteristics.Based on the sparse distribution of the target scattering centers,the sparse representation-based parameter extraction and imaging methods are studied in order to obtain higher-resolution radar images from finite observational data.Firstly,the basic principle of radar imaging is expounded.The relationship between radar imaging and signal sparse representation is explained by radar observation signal model,which provides a theoretical basis for subsequent parameter estimation.Aiming at the problem of one-dimensional range profile imaging,a method of parameter extraction and imaging based on FOCUSS and minimization residuals in geometric diffraction model is proposed.The method uses the estimation result of FOCUSS algorithm as the initial value of the parameter and the minimum residual as the objective function to optimize the type parameter of the scattering center,and improve the precision of parameter estimation.Aiming at the problem of two-dimensional radar target imaging.Based on the sparse reconstruction in the frequency domain,a new method for estimating the scattering center parameters based on the time domain reconstruction is proposed,and the estimated parameters are retrieved into the imaging scene.The method utilizes the sparsity of the attribute scattering center response and parameter dictionary in the time domain to transform the whole process of frequency domain reconstruction to time domain,which improves the stability of parameter estimation and reduces the computational complexity of sparse reconstruction.Aiming at the problem that the parameters are not on the grid points in the process of sparse reconstruction,an adaptive refinement grid method based on rough estimation of parameters is proposed by using iterative optimization.The idea of auto-adaptive meshing is combined with sparse domain reconstruction.The proposed method improves the estimation accuracy of the parameters and can control the iterative computation well.For the three-dimensional target imaging problem,the sparse reconstruction problem is considered as a sparse reconstruction of the full-angle data from multi-angle data recovery.By using the distribution of the two-dimensional images in the image domain,the frequency-domain data of different angles in the three-dimensional data are transformed to the sparse reconstruction in time domain.This sub-angle processing can convert large-scale and complex high-dimensional reconstruction problems into small-scale problems that can be solved in parallel.The proposed method can effectively reduce the computational cost and ensure the accuracy of the estimation.