Spectroscopic data correction using parametric modeling

Spectroscopic data are prone to instrumentation noise. Traditional methods for noise removal require a priori knowledge of the spectra. One does not have such information when analyzing chemical components in the environment. In this paper, parametric methods for performing spectral analysis of spectroscopic data are investigated. The corresponding model parameters used to describe the spectrum as well as the model order are examined for both synthetic and real data for efficient peak detection accuracy and noise removal. A comparison with the results obtained to the Savitzky-Golay fitter is also made to demonstrate the applicability of parametric methods.