本文由双正交共轭滤波器构造有限维内积空间的双正交多分辨分析和快速离散小波变换(FDWT),提出信号的对偶小波变换方法,并将其用于降低雷达数据存储量,提出双正交变换系数筛选方法以及基于压缩数据的小波域相关和最小距离分类方法。通过六种飞机目标的距离高分辨数据的分类实验,证明本文提出的变换、压缩和识别方法是有效的。 In this paper, biorthonormal multiresolution analysis and fast discrete wavelet transform (FDWT) of finite dimensional inner product space are constructed by biorthogonal conjugate filters. The dual wavelet transform method is proposed and used to reduce the data storage capacity of radar. The biorthogonal transform coefficient screening method and the method of wavelet domain correlation and minimum distance classification based on compressed data are proposed. The experiments of classification of high resolution data of six aircraft targets prove that the proposed transformation, compression and identification methods are effective.