The radar targets identification based on feature extraction with wavelet transformation and nearest - distance classer is studied in chapter 3 . discrete orthogonal wavelet transformation and wavelet transformation based on beylkin algorithm are applied to the samples of the targets to test their performance . the former has a better antinoise performance and the latter has a better identification rate 離散正交小波變換中提取的特征為信號的低頻部分,在特征提取過程中已實現了消噪,具有較好的抗噪性能;而beylkin算法利用樣本序列所有圓周移位的小波變換的高頻部分,提取的信號特征雖具有移不變性,有較好的識別率,但抗噪性能較前者差。