Research On SAR Target Enhancement Imaging Method Based On Prior Information

As an active microwave detection system,synthetic aperture radar(SAR)has the capability of all-weather and all-weather imaging,and has important application value in aerospace,geographic mapping,military guidance and other fields.With the development of radar technology,the performance of SAR imaging is getting higher and higher,and the demand for target detail features of SAR image is also increasing.The traditional matched filtering imaging method does not use any prior information and cannot adapt to the characteristics of the target and environment,so its imaging performance is difficult to be broken through.In recent years,compressed sensing SAR imaging methods can enhance the features of target scattering points by utilizing the sparsity prior information about the target.However,it is difficult to enhance the edge,texture and other structural features of the target only by using sparsity priors,and the imaging results are easily interfered by strong clutter scattering points of the target.Therefore,it is of great significance to research a more effective SAR target feature enhancement imaging method.SAR target enhancement imaging based on prior information is based on compressed sensing sparse SAR imaging.By adding the representation of the prior information about the target of the reconstruction model,it adapts to the characteristics of the target and the environment of the imaging scene and achieves the feature enhancement of the target for the imaging process.In view of the existing problems of SAR imaging method based on sparse prior,this thesis carries out research on SAR target feature enhancement imaging method by utilizing more prior information about structural features of SAR images.The main research contents are as follows:1.The echo signal model of airborne SAR imaging is established.Aiming at the problem that the traditional matched filtering SAR imaging method cannot adapt to the characteristics of the target and the environment,the compressed sensing SAR imaging method using sparse prior information is studied,which lays a theoretical foundation for the study of the target enhanced SAR imaging method.2.Weighted and hybrid total variational(TV)regularization SAR target enhancement imaging method is proposed.In this method,the statistical prior information is used to segment the initial imaging results,and the target region and background region are weighted respectively.On this basis,a hybrid TV regular term representing the target structure prior information is constructed to enhance the target feature and suppress the background clutter.3.The SAR target enhancement imaging method based on structure clustering dictionary for complex scenes is proposed.In this method,the sample set of image blocks is clustered according to the structural features,and the sparse dictionary learning of all kinds of image blocks is carried out by K-SVD algorithm to obtain the super-complete dictionary containing the scene prior information,which completes the sparse reconstruction of the imaging scene and achieves the object enhancement of complex scenes that do not meet the sparse characteristics.The SAR target enhancement imaging methods mentioned above are verified by simulation experiments.Experimental results show that the proposed imaging methods can effectively enhance the features of the target.