An improved technique for distortion removal in Raman spectra using impulse response filter banks

Abstract

For accurate analysis and characterization of materials using Raman spectroscopy, it is often necessary to remove signal distortions, including the noise from the spectra, using a filter such as the Savitzky-Golay (SG) spectral filter. However, distortion filtering accuracies are limited for SG filters due to their polynomial-fitting-based distortion removal process and the inherent inability to select required distortion removal frequencies. In contrast to the SG filtering techniques, Impulse Response (IR) based stable digital filtering techniques using Filter Banks (FB) should be more accurate in removing only the unwanted distortions, especially, due to their ability to control the filtering process based on the signal frequency sub-bands. Thus this paper describes a new and more efficient Impulse Response Filter Bank (IRFB) band-pass filtering technique to remove the distortions in Raman spectral signals using two-band filter banks.