雷达目标一维像识别中的自适应特征子空间法

针对雷达目标一维距离像识别中常规特征子空间法的缺点，提出了一种自适应特征子空间法。该子空间既能自适应子空间维数的变化，又能自适应目标类别数的变化。同时，与常规特征子空间相比，建立自适应特征子空间的运算量大为减少。对三类和四类目标的仿真实验结果表明了该方法的有效性。

Recognition of radar target based on adaptive feature subspace using range profile

Aiming at the shortcoming of general statistical feature subspace for range profile recognition in radar target, an adaptive feature subspace approach was proposed. When the number of subspace's dimension or the number of target classes was changed, the adaptive feature subspace could be easily obtained by iteration. Compared with general statistical feature subspace, less operation was needed in forming the adaptive feature subspace. The simulated experimental results of three targets and four targets demonstrate the efficiency of the proposed approach.