针对拖尾噪声的中值滤波—小波消噪算法分析

An Algorithm Analysis of Median Filter & Wavelet Threshold for Heavy-tailed Noise

与高斯噪声相比,拖尾噪声有更多的异常值,利用传统的小波阈值方法不能对其有效消噪。提出利用中值滤波一小波消噪方法进行处理,首先利用中值滤波抑制异常值,然后利用小波阈值方法消除残留噪声,并给出了适合拖尾噪声的消噪效果评价准则。基卜提出的准则,通过实验比较了小波阈值方法与中值滤波一小波消噪方法的消噪效果,结果表明所提出的方法能更好的消除拖尾噪声,具有较好的鲁棒性。

Compared with Gaussian noise, Heavy-tailed noise has more outliers, and traditional wavelet threshold cannot suppress outliers. A new method combining wavelet threshold and median filter is proposed. After suppressing the outliers in signal through median filter,the wavelet threshold is used and remained noise is eliminated further. A new appropriate effectiveness measure for heavy-tailed noise is given. The experiment based on the new measure shows that the new method can suppress heavy-tailed noise effectively, and it is more robust than classic wavelet threshold.