小波信号分层中的软阈值模糊性确定分界

Soft Threshold Fuzziness Determination in Wavelet Signal Stratification

文中以小波阈值信号去噪为研究对象,提出将小波信号分层中的软阈值进行阈值精确取值分析。利用小波函数将信号从时阈变换为频域,使小波各层频域充满噪声信号的能量,使信号与小波系数形成一一对应,在各层分解后使信号层的小波函数在尺度空间中完成确定;运用小波阈值信号去噪方法,对经过处理后的信号进行小波阈值选定,通过对软阈值函数的选值,确定噪声小波函数在小于小波函数的情况下应舍弃该信号;确定软阈值的选择能够令大部分噪声的小波系数取值为零,输出处理后的有效小波信号,确定软阈值的界限。实验证明,对小波阈值中的软阈值函数进行精确取值,对叠加的噪声信号或干扰能实现有效的处理。

In this paper, wavelet threshold denoising is studied, and the soft threshold value of wavelet signal layer is analyzed. The signals from the time domain using wavelet transform threshold for wavelet function, the frequency of each layer is full of the energy of noise signal, the signal and the wavelet coefficients corresponding to the formation, in each layer after the decomposition of wavelet function in signal level identified in the scale space; wavelet threshold denoising method based on the signal, after signal processing after were selected based on wavelet threshold, soft threshold function selection,determine the noise wavelet function should give the signal in the case of less than the wavelet function;wavelet coefficients to determine the values of soft threshold selection can make the most of the noise is zero, the effective wavelet signal output after processing, to determine the boundaries of the soft threshold. The experimental results show that the soft threshold function of wavelet threshold value is accurate, and the noise signal or interference can be processed effectively.