Active/passive Compounding Terahertz Imaging Algorithm Based On Fast Human Security Imaging System

Terahertz and millimeter-wave technologies have been gradually applied in human security inspections.Human security screening images can be obtained through active radar scan imaging or passive radiometer imaging and the types of dangerous explosives carried can be identified.A simple active or passive imaging system cannot take into account both the imaging frames and the detection rate of suspicious objects,and it is difficult to work in crowded places.The imaging system of active-passive security screening is proposed for the characteristics of active/passive compounding security imaging,and studied the key issues of the algorithm in this paper,.The core of the active-passive compounding imaging system is the linkage of the active and passive systems.For the passive imaging system with poor image quality and serious noise pollution,this paper studies the related passive image denoising algorithm.The total variation model is introduced to preserve the image texture features,the result is compared with wavelet domain denoising method.In order to remove the background texture of the image,the RTV algorithm is proposed to segment the suspicious object.Experimental results show that the method can effectively extract the position and contour of the target.In order to obtain tomographic images of suspicious objects,active imaging is performed by a terahertz band FMCW system.The key indicators of the active imaging process have been analyzed and validates the system’s one-dimensional imaging capabilities has been well verified.For the problem of resolution degradation caused by the nonlinearity of the transmitted signal in the linear frequency modulation system,the error is estimated using the delay line and the homomorphic deconvolution method.The RVP algorithm is re-derived and the large approximation problem in the compensation process is performed modify.Experimental results show that the improved linearity correction method has a good effect on simulation and measured data.The imaging defocusing problem caused by walking in the active/passive compounding imaging process is analyzed and modeled.Azimuth and range data distortion caused by defocusing Gaussian beams is derived and then the SRCNN algorithm is proposed to reconstruct the image.