Dynamic Moving Savitzky-Golay Smoothing Algorithm for the Raman Spectrum Automatic Baseline Correction

Zhen Ma,En Ma,Feibing Xiong,Guopeng Xu,Chong Li

DOI: https://doi.org/10.14128/j.cnki.al.20173704.614

2017-01-01

Applied laser

Abstract:The baseline drift is an unavoidable phenomenon in the collection of Raman spectrum.In the process of qualitative and quantitative analysis of Raman spectrum,the baseline drift can adversely affect spectrum analysis.To improve the accuracy and stability of the analysis results,baseline correction pretreatment of Raman spectra is required.In order to correct the base-line drift,a dynamic moving Savitzky-Golay algorithm(DMSG)is proposed to fit the background baseline to automatically identify the spectral peak region of the original spectrum and the subinterval of the non spectral peak region,thus the number of smooth iterations and the window width are automatically controlled in each Raman spectral subinterval,and the Raman peak is gradually reduced to the same order as the baseline,finally the baseline of the entire spectrum is estimated.The algorithm is applied to the simulation and real Raman spectrum for baseline correction.The results show that the algorithm has a better baseline correction effect both for the simulated Raman spectrum and the real Raman spectrum.It is a feasible baseline correction method to provide more accurate information for further analysis of the spectrum data.