The Study Of Imaging Methods For Sparse Targets Through Walls With Fuzzy Wall Geometry Parameters

Through-the-Wall Radar can locates and tracks hidden targets,it plays an important role in military and rescue applications,but when the targets behind the wall are detected,multipath signals will be generated,which will affect the imaging quality.This dissertation focuses on the imaging algorithms of Through-the-Wall Radar with fuzzy position parameters and the multipath effect as follows:（1）In view of the problems that there are problems such as high requirements for the initial estimate of the side walls position and harsh operating conditions when using Through-the-Wall Radar Imaging（TWRI）algorithms with fuzzy side wall position parameters.This dissertation proposes a new image algorithm of sparse targets through the wall with fuzzy wall position parameters,the proposed algorithm combines the multipath effect with the Back Projection（BP）algorithm,and it takes advantage of the fact that when the position of the side walls deviate,the position of the real targets in the reconstructed image also deviate to realize the estimation of the position parameters of the side walls,and then,Block Orthogonal Matching Pursuit（BOMP）algorithm is used to reconstruct the detection scenes behind the wall.The simulation results show that the proposed algorithm can not only accurately recover the location parameters of the side walls,but also achieve high resolution imaging of the detection scenes behind the wall.Compared with the Limited Broyden Fletcher Goldfarb Shanno-Particle Swarm Optimization（LBFGS-PSO）algorithm,the proposed algorithm can estimate the position of the side walls accurately within a rougher initial estimation range,and has a wider application range.（2）In view of the problems such as a large number of false targets in reconstructed images when using the BP algorithm to image the scene behind the wall,in this dissertation,the BP algorithm is improved by using multipath effect.First,the reconstructed images of electromagnetic waves in different propagation paths are obtained.Then the real targets and the false targets are distinguished according to the difference of the mesh pixel values in the corresponding positions of different reconstructed images.The simulation results show that the reconstruction quality of the improved BP imaging algorithm is better than that of the traditional BP imaging algorithm.