基于多通道小波变换的化学信号滤噪方法

A filter method of Chemical Signal based on m-band wavelet transform

基于多小波变换的理论与算法,提出了多小波软阈值去噪算法。用模拟高斯信号对多小波软阈值滤噪方法与单小波软阈值滤噪方法进行了比较,实验结果表明,多小波滤噪方法去噪效果优于单小波。将多小波软阈值滤噪方法用于黄连提取物的5种组分毛细管电泳信号的滤噪,进行滤噪处理后,噪音基本上被消除,峰位置十分清晰,峰的位置、面积及高度基本不变,基线平稳,有利于进一步进行定量计算。

A novel de-nosing method--multiwavelet soft threshold filtering based on the theory and arithmetic of muhiwavelet transform was founded. The fundamental principal of the multiwavelet soft threshold method and scalar wavelet soft threshold method was compared by simulated Gauss signal. The experimental results showed that the method of muhiwavele soft threshold filtering was better than the method of scalar wavelet soft threshold filtering. The method of muhiwavelet soft threshold filtering was used to de-osing processing of the capillary eleetrophoresis signals of five components of hydrastis pick-ups. After the de-nosing processing,the noise can be eliminated basically, the peak position is very clear, the peak position, proportion, and height are fixed ultimately. The baseline is stabilized, which was availed to the next quantified calculation.