自适应小波的构造及其在信号处理中的应用

Self-Adapted Wavelet Construction and Its Application to Signal Processing

基于带限正交小波基的特性,本文采用匹配方法构造了一种新的带限尺度函数,它与信号的主要能量所在频带具有相似的特征.匹配尺度滤波器是优化滤波器,它锁定能量聚集的频带,使得所需频率分量的输出增大.原始信号能够用优化的小波实时处理.通过对脑电图信号的处理,证明本文算法是有效的.

Based on the property of bandlimited wavelet, a matching method is developed for constructing a new bandlimited scaling function, which has the similar frequency spectrum as the original signal in the band occurring main energy. The scaling filter is an optimizing filter, which locks the frequency band in that energy is concentrated, so that the filter output of the interested frequency components increases. The original signal can be processed with optimize wavelet in real time. The experimental results with electroencephalogram signals demonstrate the method is effective.