Blinowska K J, Durka P J
Laboratory of Medical Physics, Warsaw University, Poland.
Br J Audiol. 1997 Dec;31(6):449-59. doi: 10.3109/03005364000000039.
Wavelet transform and multiresolution decomposition are described. Examples of the application of orthogonal wavelet transform to acoustic evoked potentials and otoacoustic emissions (OEA) are given and basic features of wavelet packets and wavelet network methods are characterized. An approach that enables the identification of local signal structures--a generalization of wavelet transform called Matching Pursuit--is presented. In the framework of this method the signal is decomposed into time-frequency 'atoms', which offers a possibility of determination of an 'instantaneous frequency' with the accuracy close to the theoretical limit. The method is illustrated by application to OAE signals. The advantages and limitations of the methods presented are discussed.
描述了小波变换和多分辨率分解。给出了正交小波变换在听觉诱发电位和耳声发射（OAE）中的应用示例，并对小波包和小波网络方法的基本特征进行了描述。提出了一种能够识别局部信号结构的方法——一种称为匹配追踪的小波变换推广方法。在该方法框架下，信号被分解为时频“原子”，这使得能够以接近理论极限的精度确定“瞬时频率”。通过将该方法应用于OAE信号进行了说明。讨论了所提出方法的优缺点。
