摘要
针对基于毫米波雷达多普勒频谱的旋翼、固定翼识别难题,从理论上分析了直升机旋翼探测的必要条件,结合中心矩、统计量等特征提取方法,提出了利用组合特征进行分类的识别思路,并选取支撑矢量机作为分类器。外场试验数据表明该方法具有良好的识别性能。
In this paper, the necessary condition for helicopter rotor blade radar echo detection is analyzed, aiming at the difficult i- dentification problem of helicopter and fixed-wing aircraft based on Doppler spectrum of the millimeter-wave radar. Then a method based on combined features of central moment and statistics is suggested, using a Support Vector Machine （SVM） classifier. Sim- ulation results with real data show the performance and applicability of the method is excellent.
出处
《现代雷达》
CSCD
北大核心
2009年第4期28-33,共6页
Modern Radar
基金
国防预研项目
关键词
毫米波段识别
组合特征
旋翼
多普勒谱
支撑矢量机
millimeter-wave identification
combined feature
rotor blades
Doppler spectrum
SVM
