针对舰船目标高分辨距离像的特征选择方法

摘要

In order to improve the target recognition rate of ships based on high resolution range profile (HRRP),the target region of a ship is detected firstly to solve the issues of HRRP magnitude sensitivity and shift sensitivity. Next, according to the traditional feature extraction methods,15 features for ship target recognition are extracted. Finally,to select an optimal subset of features,a feature selection method is proposed based on the Fisher criterion,which makes use of complementary among features. Simulation results shows effectiveness of the proposed method and a classifier is chosen to match the selected feature set.%为了提高基于高分辨距离像(HRRP)的舰船目标识别率,首先通过目标区域提取来解决HRRP的幅度敏感性和平移敏感性的问题;然后根据传统的特征提取方法,提取出15个较好的特征进行多特征综合识别;最后在基于Fisher准则的特征选择方法上进行改进,提出了一种基于特征互补性的特征选择方法,选择一个最优特征子集.通过仿真实验验证了提出的特征选择方法,同时选择出一个最佳分类器.