基于联合稀疏性的多视全极化HRRP目标识别方法

该文考虑利用连续获取的多视全极化高分辨距离像(High Range Resolution Profile, HRRP)进行目标识别的问题。多视全极化HRRP样本包含了3个层次的先验信息：样本内各分量来自同一目标；单视内4种极化组合方式下的HRRP均对应相同的目标姿态；相同极化方式下的多视观测是相关的。为有效利用上述信息进行目标识别，该文提出一种基于联合稀疏表示的多视全极化HRRP目标识别方法。该方法约束各分量对应的稀疏表示系数共享原子级的稀疏模式。原子级稀疏约束使得从各极化字典中选择来自相同姿态的字典原子对样本中各分量进行稀疏表示，可以有效利用上述3个层次的先验信息进行目标识别。利用目标电磁散射数据对所提方法进行了验证，结果表明，该方法具有较好的识别性能，并且对噪声具有良好的鲁棒性。

Multi-view Polarization HRRP Target Recognition Based on Joint Sparsity

The issue of automatically recognizing a target from its Full-Polarization High Range Resolution Profiles (FPHRRPs) with consecutive observations is considered. The prior information contained in a multi-view FPHRRP sample is hierarchical: all the entries contained in the sample are originated from the same target; the entries within a single view are associated with the same target pose; the multiple views under the same polarization mode are correlated. To utilize efficiently the prior information for target recognition, a novel joint sparse representation based multi-view FPHRRPs target recognition method is proposed. The presented method assumes all the entries within a multi-view FPHRRP sample share a common sparsity pattern in their sparse representation vectors at atom-level, which has the advantage of exploiting the aforementioned information to enhance recognition performance. Experiments are conducted using a synthetic vehicle target dataset. The results show that the proposed method achieves promising recognition accuracy and it is robust with respect to noisy observations.