基于改进阈值法的小波去噪算法研究

Wavelet denoising algorithm based on improved thresholding method

为了克服传统小波阈值去噪算法中存在的缺陷,采用小波系数放大法,并对阈值函数进行改进。由于信号中存在奇异点,会在奇异点处产生Pesudo-Gibbs现象,采用基于改进小波阈值法的平移不变去噪方法可以有效抑制Pesudo-Gibbs现象,对去噪效果进行强化。仿真实验结果表明:基于平移不变的去噪方法可以明显提高信噪比,其去噪效果优于传统的软、硬阈值去噪方法,具有较高的实用价值。

In order to overcome flaws of the traditional wavelet threshold denoising algorithm,the method of wavelet coefficients magnified was used,and threshold function was improved.Because singular points were existed in the signal,the Pesudo-Gibbs phenomenon was produced in the singular points.Pesudo-Gibbs phenomenon was effectively inhibited and the effectiveness of the denoising was strengthened by using translation invariant denoising based on improved wavelet threshold method.Simulation experiment results show that the signal to noise ratio can be significantly improved by using the denoising based on translation invariant method,and effectiveness of the denoising is better than the traditional soft and hard threshold denoising method.It is applicable in engineering.