基于二进小波变换的实信号的多尺度Hilbert变换和瞬时频率提取

DYADIC WAVELET TRANSFORM BASED REAL SIGNAL MULTISCALE HILBERT TRANSFORM AND EXTRACTION OF INSTANTANEOUS FREQUENCY

该文着重讨论了使用二进小波变换方法实现数字Hilbert滤波,特别是使用滤波器组实现多尺度Hilbert滤波所要考虑的各尺度滤波器的设计问题。将多尺度Hilbert方法应用到求信号的瞬时参数中,并利用小波变换的去噪思想,取得了比使用DWT方法去噪更好的去噪效果。

In this paper , the method for designing digital filters and their implementation and application via filter banks are introduced. The design of digital Hilbert filters of each scale in dyadic wavelet filter banks is particularly discussed. Moreover, this method, which based on multiscale Hilbert filtering methods, is applied to the extraction of the instantaneous frequency and obtains a better denoising result compared with the traditional DWT denoising method.