Posted by Kenneth Burch On November - 1 - 2016 Comments Off on Automatic Spike Removal Algorithm for Raman Spectra

Raman spectroscopy is a powerful technique, widely used in both academia and industry. In part, the technique’s extensive use stems from its ability to uniquely identify and image various material parameters: composition, strain, temperature, lattice/excitation symmetry, and magnetism in bulk, nano, solid, and organic materials. However, in nanomaterials and samples with low thermal conductivity, these measurements require long acquisition times. On the other hand, charge- coupled device (CCD) detectors used in Raman microscopes are vulnerable to cosmic rays. As a result, many spurious spikes occur in the measured spectra, which can distort the result or require the spectra to be ignored. In this paper, we outline a new method that significantly improves upon existing algorithms for removing these spikes. Specifically, we employ wavelet transform and data clustering in a new spike-removal algorithm. This algorithm results in spike-free spectra with negligible spectral distortion. The reduced dependence on the selection of wavelets and intuitive wavelet coefficient adjustment strategy enables non-experts to employ these powerful spectra-filtering techniques.

Applied Spectroscopy  PDF_IMG

Comments are closed.