Research On Near Field Array Imaging In The Terahertz Band

Near field array imaging system in the terahertz band can achieve detection and imaging of near field targets with high frame rate and high resolution,so it is applicable to the fields that require high real-time performance such as safety inspection,missile guidance and military reconnaissance.However,the range migration in the near field condition is serious and the wavelength is short in terahertz band which increase the difficulty of array system integration.In this paper,the key issues of terahertz near field array imaging system design and imaging method are studied in detail.Aiming at the design of terahertz near field array imaging system,the signal frequency band,array scan mode,array element configuration mode,calibration method for phase inconsistency of element channels and experimental system structure are determined.The key parameters such as resolution,sampling rate and signal to noise ratio of the imaging system are analyzed,and the specific parameter formula is obtained.Aiming at the terahertz near field array imaging method,the imaging algorithm and the influence of the element position deviation on imaging is studied emphatically.Firstly,the range migration algorithm is deduced.In order to solve the problem of large amount of computation and possible inaccuracy of interpolation,a range shift migration algorithm based on the range migration factor with dynamic compensation distance adjustment is proposed.Then,the influence of random array element position deviation and fixed element position deviation introduced by array mechanical scanning is analyzed,and a correction method based on two-dimensional non-uniform fast Fourier transform is proposed for fixed element position deviation.In order to solve the problem that the huge amount of antennas and large amount of data which leads to higher system cost,larger storage space and longer acquisition time of echo signals under the array imaging system,a terahertz near field sparse array imaging method based on random sampling frequency is proposed.Based on the three-dimensional sparse data,we adopt the combination of orthogonal matching pursuit algorithm in the compression sensing theory and the two-dimensional non-uniform fast Fourier transform algorithm to reconstruct the image of target.