| In recent years,terrorist attacks emerge in an endless stream,security imaging has become a widespread concern.Millimeter wave has significant advantages in the near field imaging,which has the characters of both infrared and microwave,all-weather work and imaging of concealed objects.It also has non ionizing and no harm to human body.Its high frequency brings a wider available band.Based on the above advantages,millimeter wave near-field holographic imaging technology has become an effective way to study the characteristics of targets.This technology is increasingly popular and has broad research prospects.The fundamental goal of millimeter wave holographic imaging is to locate the near-field object in one dimension,two-dimensional coronal plane and two-dimensional cross-section image,and finally get the high resolution target 3-D image.Based on the application of millimeter wave in near field imaging,this paper has carried out related research and discussion.Three imaging algorithms have been analyzed and improved to solve the problem of phase interference caused by spherical wave in near-field imaging process,and further improve the imaging quality.Firstly,the thesis discusses the spatial phase compensation algorithm,which gain the analysis of echo signal,and then according to the specific expression of echo signal to construct spatial phase compensation factor,to compensate for the echo after Fourier transform.To solve the problem of the side lobe dragging and decreasing of the target image,the improved CLEAN algorithm is modified.Secondly,analys the reasons for the inaccuracy of compensation according to the imaging results of spatial phase compensation,and then discuss the wavenumber domain imaging algorithm based on phase compensation.To improve image resolution,the algorithm extends the wave number vector in all directions to construct the wavenumber domain compensation factor(as matched filter more accurate).For the uneven distribution of frequency data in cross section imaging,we use improved cell interpolation or NUFFT to solve the problem,so that the target image point will further converge.Furthermore,in order to break through the limitation of the Nyquist sampling frequency,the compression sensing reconstruction algorithm is discussed.The theoretical basis of compressed sensing is analyzed.The reconstruction algorithm is mainly discussed.Based on the multi extremum search algorithm,the search probability is improved,and the searching probability of the target scattering area is increased.The imaging time is greatly saved.Lastly,the comprehensive comparative analysis of the three imaging algorithms is performed,mainly from the imaging distance,the sampling requirement,the complexity of the algorithm,the imaging resolution and other aspects to get a comprehensive comparison of the advantages and disadvantages of the three algorithms,analys their application scope based on their advantages and disadvantages. |
