摘要
本文运用于波变换和付立叶-梅林变换两种不同的方法对三种飞机缩比模型进行了分类识别的研究与实验,详细比较了两种方法的正确识别率以及对方位角的敏感程度,并给出了比较结果,结果表明,在高频区高分辨率雷达目标识别中,于波变换方法的总体识别率高于付立叶-梅林交换,而在稳健性方面不如付立叶-梅林变换,实验结果与理论相符.
Wavelet transform and Fourier-Mellin transform are applied to classify threekinds of model planes. The correct rate of classification of the two methods is compared,so is thesensitivity to azimuth of the two methods. The comparison results are presented. The results indicate that radar target recognition in HF domain wavelet transform method is better than FourierMellin transform method, while the stability of Fourier-Mellin transform is better than wavelettransform. The experimental result is accord with theory.
出处
《电子学报》
EI
CAS
CSCD
北大核心
1997年第9期95-97,共3页
Acta Electronica Sinica
关键词
子波变换
神经网络
雷达
目标识别
MFMT
Wavelet transform,Fouier-Mellin transform,Feature extraction,ANN