一种基于子波变换的信号滤噪改进方法

对子波变换多尺度下信号与噪声的不同性质进行了研究,提出了一种在子波域不同尺度上选取不同的滤噪方法,该方法将经典的软阈值滤噪与子波变换的模极值传播特性在一定尺度上有机结合起来处理信号.在改善信噪比的同时,也尽可能地保持原信号的边缘信息和精细特征.通过仿真验证了该方法的实用性和优越性.

An Improved Method of Signal Denoising Based on Wavelet Transform

The characteristics of signal and noise wavelet transform are studied in this paper and an improved denoising method is proposed under different scales in wavelet domain. This method combines the classical soft threshold denoising with propagation characteristic of the wavelet transform module extremum in a certain scale to process signal. It is an effective method not only enhances signal noise ratio but also retains the edge and fine characteristics of the signal as much as possible. The experimental results show that this denoising method is practical and excellent.