A Low-rank And Sparse Approach For Through-wall Radar Imaging

Through-wall radar uses electromagnetic waves to detect scenes behind walls and other visual opaque materials to reconstruct images of targets behind complex scenes or obstacles.It has been widely used in rescue operations,anti-terrorism reconnaissance and many other civilian and military fields through its ability to perceive closed structures.However,the return of the strong electromagnetic wave from the wall usually affects the reflection of the target,making the target detection difficult or even impossible,which brings great challenges to the imaging of indoor scenes.In addition,multiple reflections in complex scenes can produce multipath virtual images,which blur the radar signal when it returns from a weak target.Therefore,in this paper,considering the influence of the wall,based on several imaging methods of through-the-wall radar and combined with low rank representation algorithm,the following research contents are proposed for indoor target scene imaging:Firstly,this paper introduces the propagation model of electromagnetic wave of through-wall radar,expounds the relationship among wall,antenna and target.Then analyses the propagation path of electromagnetic wave through the interaction between transmitting antenna,wall and target and returns to receiving antenna,and three kinds of through-wall radar imaging methods are given.Based on one of the algorithms,the propagation delay of the imaging area mesh is estimated,and a fast wall compensation algorithm is proposed for the case where the computational delay is complicated.Secondly,based on the theoretical framework of compressed sensing,several clutter suppression methods are proposed to process radar echo signal from the traditional compressed sensing imaging method.The background cancellation method is used to remove the influence of the wall and image the target scene.As a comparison,for the low rank property of the wall echo signal,combined with the sparse characteristics of the target signal,the image region is subjected to low rank sparse imaging.Finally,relying on the gprMax simulation data,combined with the measured data to further improve the reliability and reference of the target imaging.Finally,the experimental scene is modeled as an indoor enclosed environment,and the antenna around the wall is designed as a single-input multiple-output antenna array to analyze the indoor multipath propagation process.A multipath virtual image suppression method based on compressed sensing is proposed for the multipath effect.Then,starting from the process of distributed radar signal processing,a distributed network model is presented.On the basis of multipath virtual image suppression,target imaging is realized by combining low rank representation and distributed compressed sensing.