Radar high‐resolution range profile recognition via geodesic weighted sparse representation

• DOI: 10.1049/iet-rsn.2014.0113
• Publication Date: 2014-06-11T19:03:37Z
• Authors: Jianqiao Wang; Yuehua Li; Kun Chen
• Keywords: Representation

One of the radar high‐resolution range profile (HRRP) recognition issues is target‐aspect sensitivity. Both theoretical analysis and real‐world data show that HRRP shows a high correlation only in very small aspect region. To overcome this problem, traditional methods, scattering centre model averaged are utilised. In study, authors present graph‐based semi‐supervised method, called geodesic weighted sparse representation (GWSR), to sensitivity problem. It assumed from different targets located on manifolds information utilised separate these manifolds. GWSR, distance calculated firstly then labelled reconstructed by weight. The nonlinear structure can be transformed into linear one through reconstruction process. Then, unlabelled sparsely weight estimate label given labels. Experiments three kinds ground target HRRPs with backgrounds demonstrate effectiveness authors’ method.