Millimeter Wave Three-Dimensional High-Resolution Imaging Technology For Dielectric Targets

Millimeter wave and terahertz wave have moderate penetration to most dielectric materials,which makes millimeter wave and terahertz wave high-resolution near-field three-dimensional radar imaging technology play a more and more important role in human security,hidden target detection,nondestructive testing and other fields.These application scenarios are in line with the free space hypothesis.However,the free space hypothesis will no longer meet the requirements of imaging scene when detecting media targets.Therefore,this paper studies the three-dimensional high-resolution imaging technology of layered media targets.Firstly,the imaging algorithm of multilayer dielectric targets based on frequency domain matched filtering is proposed,and the imaging resolution is analyzed.Aiming at the problem that the improved back project(IBP)algorithm and time reverse(TR)algorithm suitable for medium target imaging need to solve the refraction point of electromagnetic wave,resulting in a significant increase in the amount of imaging calculation,it is proposed to convert the radar echo signal to the frequency domain,and make use of the continuity of tangential component when electromagnetic wave propagates in layered medium to avoid the calculation of refraction point,the computational complexity is reduced.Then,based on the radar range resolution formula,the imaging resolution of medium target is deduced,and the resolution formula is verified and explained by numerical simulation.Secondly,the layered medium target imaging algorithm based on non-uniform fast Fourier transform(NUFFT)is studied.At the first,the imaging algorithm of half space medium for non-uniform Single-Input-Single-Output(SISO)array radar is studied.NUFFT technology is used to transform the spatial echo of non-uniform SISO array radar to frequency domain.At the same time,NUFFT technology is used to realize range imaging,which avoids Stolt interpolation in range direction.Then,the imaging algorithm suitable for uniform SISO array radar for multilayer dielectric targets is studied.After the spatial echo of array radar is transformed into wavenumber domain,the processing method based on interpolation and fast Fourier transform(FFT)cannot be used because the electromagnetic wave has different wavenumbers in different media.Therefore,NUFFT is proposed to image layer by layer in multi-layer media,which is compared with PSM which is also suitable for imaging the scene,it can greatly shorten the imaging time.Finally,the fast imaging algorithm based on half space medium green’s function is studied.The radar echo is modeled by using the half space medium green’s function,and the imaging algorithms suitable for one-dimensional Multiple-Input-Multiple-Output(1D-MIMO)scanning array radar and Single-Input-Multiple-Output(SIMO)array radar for half space medium targets are derived.The effectiveness of the algorithm is verified by numerical simulation and experimental measurement imaging.The imaging results show that the medium target imaging algorithm based on half space medium green’s function has better imaging quality than the radar imaging algorithm based on free space assumption.