Design Of Near Field Microwave Imaging System Based On Compressed Sensing

Electromagnetic wave imaging is widely used in detecting and imaging long-range objects or environments in microwave and millimeter wave bands.In recent years,microwave synthetic aperture radar imaging has been widely used in vehicle navigation,safety shielding and medical imaging.We infer the physical information of the object to be measured by comparing some features between the received signal and the transmitted signal,and solve the inverse problem to obtain such information as shape information and location information.Building a more efficient and accurate microwave sensing system has always been the goal of scientific research and learning.It can combine different structures and algorithms.This paper mainly combines surface wave structure and compressed sensing algorithm.Firstly,the theory of inverse scattering imaging and the compressed sensing algorithm are described.Then,the surface wave structure Goubau line sur face wave transmission line as a microwave sensing system is introduced.Surface wave has a high resolution-sub-wavelength,in addition to a great advantage is that the space binding is also very high.Therefore,Goubau line surface wave transmission lin e is used as a sensing system for microwave imaging.In this sensing system,different from the traditional mechanical scanning and electronic scanning,mechanical transformation is used to move the data.It scans the data with frequency as the dispersion degree.It is a more efficient and fast scanning method,which lays a foundation for real-time microwave imaging.Based on the above principles,transmission lines and data detection methods,a microwave imaging sensing model is designed.From the basic transmission theory of electromagnetics,combined with the weak scattering characteristics of surface wave transmission lines,that is,this paper only focuses on the weak scattering situation.Combining with the mathematical tools such as Green's function a nd Bonn approximation,the mathematical relationship between the reflection coefficients of ports and the local reflection coefficients is obtained.Finally,the relationship between the disturbance impedance caused by scatterers and the local reflection coefficients is obtained.The best advantage of compressed sensing is to recover unknown global data sets by using far fewer measurement samples than traditional ones.In this paper,classical two-step iterative threshold algorithm(TwIST)is used.Based on the analysis of singular value decomposition and the optimization of compressed sensing algorithm,the microwave sensing system and reconstruction algorithm are simulated and tested in this paper.Firstly,the perception ability between the proposed transmission line and the used transmission line is compared,and the structure used in this paper has great advantages.In one-dimensional simulation,it is proved that the algorithm can reconstruct one-dimensional single small scatterer and two small scatterers effectively and accurately.At the same time,for two near-range scatterers,the perception system can distinguish more than 5 mm apart.Small scatterers.The system also proves that three small scatterers can be located effectively.In the reconstruction of the scatterer with a width,the reconstruction phase is introduced to help distinguish,that is,the interval between the positive and negative phases is close to the width of the scatterer with a width.In the process of validation,the actual test of the perception matrix and reconstruction results are consistent with the simulation.In addition,the two-dimensional expansion of microwave imaging is also carried out,and the relevant mathematical optimization formulas are deduced.Due to the limitation of the current hardware structure,frequency scanning can only be used in one direction,but not in the other direction,so it is mechanical scanning.Because the input and output of the compressed sensing algorithm can only solve the problem of one-dimensional reconstruction,an improved TwIST algorithm suitable for two-dimensional matrix operation is proposed,which reconstructs two-dimensional simulated Y-shaped and measured S-shaped,respectively.The results show that it is good.Finally,this paper also extends some of the far-field.