Vector casting for noise reduction

Abstract

We report a new method for the reduction of noise from spectra. This method is based on casting vectors from one data point to the following data points of the noisy spectrum. The noise‐reduced spectrum is computed from the casted vectors within a margin that is identified by an envelope‐finder algorithm. We compared here the presented method with the Savitzky–Golay and the wavelet transform approaches for noise reduction using simulated Raman spectra of various signal‐to‐noise ratios between 1 and 25 dB and experimentally acquired Raman spectra. The method presented here performs well compared with the Savitzky–Golay and the wavelets‐based denoising method, especially at small signal‐to‐noise ratios and furthermore relies on a minimum of human input requirements.

Refinement of spectra using a deep neural network: Fully automated removal of noise and background

“…We modeled the noise spectrum as the summation of shot noise, thermal noise, and readout noise, as described in detail in our previous work. [ 21 ] Different noise levels were generated to imitate spectra with different signal‐to‐noise ratio (SNR). Training Dataset A contains 300 000 simulated raw spectra .…”

“…Details on how the spectra were recorded and details about the equipment used can be found in our previous work. [ 21 ] The spectra cover a Raman shift range from 200 to 2000 cm −1 .…”

“…Irrespectively of the used test dataset, we compared the obtained de‐noising results with the three popular de‐noising methods, namely, SG method [ 16 ] , PS method, [ 17 ] and our vector casting method, [ 21 ] following the evaluation techniques proposed by Barton et al [ 53 ] The performance evaluation was done with respect to (a) the mean improvement of the signal quality across the entire spectral region, (b) whether or not the algorithm interacted negatively with the Raman spectral signatures, and (c) quantification of the SNR improvement relative to the SNR of the acquired original noisy spectrum. We compared the obtained background removal result with iterative fit of polynomial method (ModPoly) [ 26 ] and an iterative weighted smoother method (arPLS).…”

“…Techniques for de‐noising of raw Raman spectra include Savitzky–Golay (SG), [ 16 ] perfect smoother (PS), [ 17 ] smoothing based on wavelet transform, [ 18,19 ] finite impulse response (FIR), [ 20 ] and smoothing based on vector casting. [ 21 ] Baseline correction methods include techniques based on (i) filtering on frequency domain such as Fourier and wavelet transformation, [ 19,22,23 ] (ii) polynomial fitting, [ 24–26 ] (iii) penalized least squares, [ 27–30 ] (iv) morphological operators, [ 31–33 ] (v) multivariate analysis, [ 34 ] and shifted excitation Raman difference spectroscopy. [ 35,36 ]…”

Recent Advances in Experimental Techniques for Flow and Mass Transfer Analyses in Thermal Separation Systems

“…b shows the kinetics of the growth and the shrinkage of both peaks. With respect to the three examples provided above, the successful realization of the Raman measurements depends to a certain extend on the specific design of the Raman probe and on the processing of the obtained raw Raman spectra, which are main competences of the research team of Andreas Braeuer at TU Bergakademie Freiberg , .…”