基于目标分解的极化雷达飞机识别法

Target decomposition-based polarimetric radar aircraft recognition method

尝试将目标分解算法用于高分辨率极化目标的识别。基于 Krogager目标分解法 ,将各距离分辨单元的极化散射矩阵分解为 3个简单散射体的散射矩阵 ,分别代表球形散射体、二面角散射体和螺旋体散射体的散射机理。并利用求得的 3个幅度参量随径向距离的变化波形 ,提取反映目标散射中心位置和物理结构特性的特征。进一步采用 Fisher可分性准则作判据 ,从原始特征集中选取最佳特征组。对 5种飞机模型的识别结果表明 。

This paper applies target decomposition theory to high resolution polarization target classification. Krogager's target decomposition scheme is used to decompose the backscattering matrix of each range cell into three elementary scattering matrices, which represent sphere, diplane and helix the scattering mechanisms. The features describing scattering center locations and physical characteristics are then extracted from the range profiles of the three component amplitudes. An optimum classification feature vector is built using Fisher's separability criterion. This method was applied to the classification of five aircrafts with the results demonstrating that the target decomposition based classification scheme is quite effective.