Classification of four wood species using terahertz time-domain spectroscopy THZ-TDS

Terahertz time-domain spectroscopy (THz-TDS) combined with chemometric techniques was investigated as a non-destructive method for species classification in wood samples from four different species. For this purpose, samples were introduced into the THZ-TDS spectrometer chamber purged with dry nitrogen, and spectra were collected at wave intervals from 0,1 THz to 10 THz to evaluate the samples. It was obtained that in the range between 0,1 THz and 1,7 THz, there is higher differentiation of the samples, being this the best range to generate classification models. After the chemometric analysis, the Support Vector Machine algorithm achieved an accuracy rate of 91 % in classifying the four wood species. Finally, it was demonstrated that THz spectroscopy can be used for quantitatively complex natural organic materials such as wood.