基于时频分析的地面运动目标噪声稳健分类研究  

Noise-robust Classification of Ground Moving Targets Based on Time-frequency Analysis

作　　者：马艳艳 MA Yan-yan(The 723 Institute of CSIC,Yangzhou 225101,China)

机构地区：[1]中国船舶重工集团公司第七二三研究所,江苏扬州225101

出　　处：《舰船电子对抗》2018年第1期57-61,共5页Shipboard Electronic Countermeasure

摘　　要：在分析了地面车辆和人体目标时频谱差异的基础上,提取了三维分类特征。结合录取到的数据有可能信噪比低的实际情况,提出了正交基重构的方法对目标低信噪比回波数据进行噪声修正。基于实测数据的实验结果验证了本文所提特征提取方法和噪声修正方法的有效性。This paper extracts 3-D classification features based on analysizing the difference of time-frequency spectrum between moving vehicle and walking human on the ground,proposes a new method of orthogonal base reconstruction to correctthe noise to the echo data of targetw ith low signal to noise ratio.The experiment resutt based on measured data verifies the validity of noise extraction method and noise correcting method proposed in this paper.

关 键 词：地面运动目标 时频分析 特征提取 去噪 支持向量基 

分 类 号：TN971.1[电子电信—信号与信息处理]