基于小波变换的信号消噪

Signal Denoising Based on the Wavelet Transform

采用一种新的阈值确定方法 :二次平方损失函数法 (QSLF法 ) ,通过对待分析信号进行小波变换 ,在小波变换域中 ,把小波系数进行阈值处理 ,即 :增强属于信号的小波系数 ,减弱或去除属于噪声的系数 .然后进行小波变换的逆变换 ,虽然会造成在小程度上丢失了一些细节 ,但只要考虑到抑制干扰噪声与所保留信号细节之间的权衡问题 ,仍能恢复出所期望的待分析信号 .计算机模拟实验结果表明

A novel method of thresholding for denoising was put forward:Quadratic Square Loss Function (QSLF) Based on the wavelet transform to the measured-signal,we disposed the Wavelet Coefficient with thresholding in the field of the wavelet Transform,such as strengthening the wavelet coefficient to the signal and weakening the coefficient to the noise, Then renewed the original signal through the inverse transformation A@pitch(0 9)lthough some details were lost to some extent,the original signal was renewed when you thought over the b@pitch(2)alance between the restrain of disturbing-noise and the details of the signal After computer simulation experiment,@pitch(0)it proved that this method was helpful to the signal denoising