The Application of Mexican Hat Wavelet Filtering and Averaging Algorithm in Raman Spectra Denoising

Random noise is inevitable in Raman spectral analysis. It has a negative influence on the accuracy of analysis and detection limit. There are many disadvantages for the present methods of wavelet filtering to spectral signals. We achieved separation noise from signal by using Mexican Hat wavelet filtering combined with the averaging algorithm. This method can be used in both low frequency and high frequency cases. It denoised excellently for strong noise signals with a low signal-to-noise ratio even for less than one, and good result can be gotten during spectral processing. The spectral processing technique we propose is relatively simpler and quicker, and the result is reliable. After processing, the relative errors of the position, the height and the area of the peak are less than 0.01%, 1.13% and 0.01%, respectively. Experiments show that this method can effectively raise the accuracy of spectral analysis.