基于高斯性检验的自适应小波去噪方法

针对软阈值和硬阈值去噪算法存在的缺陷，提出了一种基于高斯性检验的自适应非线性阈值去噪方法。该方法根据信号和噪声的模极大值特性自适应确定分解层数，引入高斯性检验选择软阈值和硬阈值方法对每层小波系数进行降噪处理。仿真结果表明，该自适应滤波方法简单有效、稳定性高，去噪后信号信噪比得到很大提高，且不同仿真信号结果都明显优于经典的小波去噪算法。

An adaptive wavelet denoising algorithm based on Gaussian tests

Based on the known defects of soft thresholding and hard thresholding, an adaptive nonlinear threshold denoising method based on Gaussian tests has been proposed. The new method adaptively determines the decomposition level according to the characteristics of signal and noise for wavelet coefficients of each level, choosing soft and hard thresholding methods to deal with it through Gaussian tests. The experimental results show that the method is effective, the signal-to-noise ratio is highly improved, and the results are superior to the classical wavelet denoising algorithm for different simulated signals, resulting in much higher stability.