超宽带雷达建筑物结构稀疏成像

超宽带雷达具备穿透墙体获得建筑物内部结构布局的能力，为建筑物内人员探测定位提供更丰富的信息。传统成像常存在较为严重的旁瓣，而且墙后目标成像位置也会受墙体影响而产生偏移。为提高成像质量，稀疏重构技术被引入穿墙成像领域，但传统方法对弱散射目标的重构概率较低。该文提出结合相干因子(Coherence Factor, CF)加权的稀疏重构方法，在稀疏重构提取支撑集的过程中，利用CF增强成像的结果来提高支撑集原子的正确性，降低稀疏重构过程中强散射目标旁瓣的影响，最终提高场景中弱散射目标的重构概率。同时建立了多层墙体位置校正模型，将场景校正放到稀疏重构之后进行，从而以较低的计算复杂度降低墙体定位误差。实测数据处理结果表明，相比于传统的稀疏成像方法，相同的数抽取比例下，该文提出的方法能够有效提高场景中弱散射目标重构概率，并将建筑物内部墙体定位误差降低至10 cm以内。 

Sparse Imaging of Building Layouts in Ultra-wideband Radar

Ultra-WideBand (UWB) radar can reconstruct the layout of a building, providing rich information for detecting and locating humans in buildings. Traditional imaging methods suffer from serious sidelobes and location displacement of behind-the-wall target because of the influence of walls. Sparse recovery is introduced into the field of through-the-wall imaging to improve the imaging quality. However, the reconstruction probability of weak scattering targets is low in traditional methods. In this study, the combination of sparse recovery method and Coherence Factor (CF) weighting is proposed to improve the reconstruction probability of weak scattering targets inside a room. The quasi-establishment of the support set can be improved during sparse imaging by reducing the effect of the sidelobes of strong scattering targets with CF, ultimately enhancing the robustness of the sparse imaging of the building layout. A location correction model for multiple walls after sparse imaging is established, based on which the locating error of walls can be reduced with a low amount of calculation. The results of the measured data reveal that compared with the traditional generalized orthogonal matching pursuit method, the proposed methods can improve the reconstruction probability of weak scattering targets and reduce the locating error of the inner layouts of buildings to less than 10 cm.