Research On Target Imaging Algorithm Based On Electromagnetic Inverse Scattering

Electromagnetic inverse scattering microwave imaging has been widely used in many fields,such as anti-terrorism imaging,ground penetrating radar,medical imaging,non-destructive flaw detection,weather prediction and so on.However,most of the inverse electromagnetic scattering algorithms have limitations.An algorithm may only be able to image a particular scene,which is not universal and has poor portability.When the imaging scene is changed,the original algorithm is low precision or even unable to image.Therefore,it is of great significance for target reconstruction to seek a more suitable electromagnetic inverse scattering algorithm.First,the concept,application background,practical significance and current development of electromagnetic inverse scattering imaging are presented,the progress of the research on the forward and inverse problems of electromagnetic inverse scattering,and the algorithms and regularization methods involved in solving the problems of electromagnetic inverse scattering are described.After that,the basic theory of electromagnetic inverse scattering is introduced in detail,including the common algorithms for solving forward and inverse problems,the definition and strategy of regularization.Second,the imaging method under low contrast is presented.One is TVCS(Total-Variation Compressive Sensing),which can reconstruct the object with abrupt edge and reconstruct the edge information of the unknown targets.The other is quantitative LSM(linear sampling method),which uses far-field equation to reconstruct the shape of unknown targets,and then deduces a new measurement equation to reconstruct the unknown targets quantitatively.The characteristics of the two methods are presented,and the simulation results are given.Then the high contrast imaging method,CSI(contrast source inversion)are given.Based on the advantages of adding prior information,two methods for selecting initial value of CSI are presented,the backward-propagation method and the frequency-step method,and the simulation results are given.The simulation results of the frequency-step method show that the frequency-step method can solve the problem that the backward-propagation method can't reconstruct the target accurately under the condition of high contrast and large size target.Finally,two combined imaging methods are presented.One is the TVCS-BIM,which combines TVCS and Born iterative method.It can break through the limitation of the two methods,extend the reconstruction range to higher contrast,and can be applied to more scenes.Another is LSM-CSI,which used LSM to partition the reconstructed region and then the results obtained by quantitative LSM are taken as initial values of CSI.This method can break through the limitation of quantitative LSM and is fast and accurate for large background imaging.