Algorithms of Chemicals Detection Using Raman Spectra

Abstract

Raman spectrometers are devices which enable fast and non-contact identification of examined chemicals. These devices utilize the Raman phenomenon to identify unknown and often illicit chemicals (e.g. drugs, explosives) without the necessity of their preparation. Now, Raman devices can be portable and therefore can be more widely used to improve security at public places. Unfortunately, Raman spectra measurements is a challenge due to noise and interferences present outside the laboratories. The design of a portable Raman spectrometer developed at the Faculty of Electronics, Telecommunications and Informatics, Gdansk University of Technology is presented. The paper outlines sources of interferences present in Raman spectra measurements and signal processing techniques required to reduce their influence (e.g. background removal, spectra smoothing). Finally, the selected algorithms for automated chemicals classification are presented. The algorithms compare the measured Raman spectra with a reference spectra library to identify the sample. Detection efficiency of these algorithms is discussed and directions of further research are outlined.

Determination Of Gas Mixture Components Using Fluctuation Enhanced Sensing And The LS-SVM Regression Algorithm

“…We should underscore that the method does not require any additional data processing (e.g., background noise removal, spectra smoothing, etc. ), which is common in other chemo-metric methods [25]. This is a benefit of the LS-SVM method, because any additional pre-processing method demands optimal selection of some additional parameters before its use.…”

Non-Gated Laser Induced Breakdown Spectroscopy Provides a Powerful Segmentation Tool on Concomitant Treatment of Characteristic and Continuum Emission

“…Since the continuum background is broad and featureless, previous investigators have employed several numerical post-processing schemes to approximate and remove it from the acquired spectral data [17], [18]. Here, the background of the LIBS spectra was removed by application of an iterative least squares-based curve-fitting algorithm that uses a polynomial (6 th order for our dataset) with non-negativity constraints [19].…”

Combined analysis of whole human blood parameters by Raman spectroscopy and spectral-domain low-coherence interferometry

“…We conclude, that a quantitative analysis may be carried out with grater dataset, which will allow statistical approach by employing multivariate models and classification algorithms [18][19][20][21], such as principle component analysis (PCA), partial least-square (PLS) regression, or support vector machines (SVM). Such approach will enable to determine the quantities of specific compounds in blood.…”