Research On Super-resolution Electromagnetic Imaging Method

With the advance of society and computer technology,microwave imaging technology is increasingly used in our daily life and military fields.However,the limitations of traditional electromagnetic imaging methods are becoming more and more prominent.Therefore,it is of great significance to carry out related research,which is to obtain higher resolution and better performance imaging technology,whether it is the military modernization of national defense or the further development of social industry.Electromagnetic inverse scattering imaging is an important branch of the inverse problem,which is one of the most challenging topics in the field of academic research,and its can break through the existing electromagnetic diffraction limit and achieve super-resolution imaging of the detection area,which has a wide range of applications in many fields,but at the same time it encounters two major challenges: morbidity and non-linearity,therefore,how to effectively improve the image quality and invert the parameter target more quickly is the main research content of this dissertation.Firstly,in order to solve the problems of poor imaging quality in a inhomogeneous background,in this dissertation,we propose a novel method based on the differential integral equation model,namely the Difference Lippmann-Schwinger integral equation(D-LSIE),and this equation can effectively separate the prior information of the known background from the information of the unknown scatterer.In the process of optimizing the solution,the improved Levenberg-Marquardt algorithm(modified enhanced Levenberg-Marquardt,ME-LM)is used.In particular,in order to enable the algorithm to stably invert,we use a hybrid regularized technique,i.e.,generalized cross-validation(GCV)and truncated singular value decomposition(TSVD),and this strategy can make the algorithm more stable and improve the ability of the algorithm to solve high nonlinear electromagnetic inverse scattering.Secondly,based on the optimization algorithm of subspace-based optimization method(SOM),in this dissertation,we propose a novel multiplicatively regularized iterative updating background method(MR-IUBM).By adopting the multiplicative regularization technique,the regularization function and the loss function are organically combined,which avoids the difficulty of manually selecting the regularization parameters separately,and also makes the algorithm more stable,has strong anti-noise ability and makes the reconstructed image edge is more clear.The results of numerical simulation data and the experimental data verify the effectiveness of the algorithm.Finally,in this dissertation,based on the existing two-dimensional electromagnetic imaging theory,we systematically study the three-dimensional electromagnetic inverse scattering imaging problems.By combining the subspace optimization technique and the Distorted Born Iterative method(DBIM),we propose an imaging method based on subspace regularization and DBIM algorithm in a three-dimensional scene.Through different numerical simulation tests on various scenarios,we realize the three-dimensional imaging in different scenarios and verify the effectiveness of method.