基于多测量向量模型的极化探地雷达成像算法

针对极化探地雷达( GPR)工作过程中目标成像空间的联合稀疏性，提出了一种基于多测量向量模型的极化探地雷达成像算法。在建立极化探地雷达回波信号模型的基础上，利用各极化通道测量数据的联合稀疏性将各个极化通道的测量数据等效成多测量向量( MMV )，通过多任务贝叶斯压缩感知( MT-BCS)算法对各个极化通道的测量数据进行联合处理从而实现各个极化通道对应的探测场景反射率的重建。基于时域有限差分( FDTD)法的仿真数据处理结果表明所提成像算法在目标位置重建的准确性和背景杂波抑制能力上均优于单测量向量( SMV)模型的极化探地雷达成像算法。

An imaging algorithm based on multiple measurement vectors model for polarimetric ground penetrating radar

By exploiting the joint sparsity of the target image space, a novel imaging algorithm based on multiple measurement vectors( MMV) model for polarimetric ground penetrating radar( GPR) is proposed. With the established polarimetric signal model of targets echoes,the multiple polarimetric channel measure-ment data can be treated as the measurement vectors by exploiting the joint sparsity between different polar-imetric channel signals. Then the measurement data is jointly processed by the multi-task Bayesian com-pressive sensing( MT-BCS) algorithm to reconstruct the image of the investigated scene. The processing re-sults of simulation data generated by finite-difference time-domain ( FDTD ) method have demonstrated that the proposed imaging method is superior to the traditional single measurement vector( SMV) based im-aging method both on the accuracy of target location and the suppression of clutter.