Abstract

A Fourier type filtering method is proposed for the pretreatment of near-infrared (NIR) spectra of thin (<100 μm) transparent plastic foils before their identification by means of multivariate calibration methods. The interference of multiply reflected beams from the boundary surfaces of the foil causes a disturbing signal component in the spectrum and the identification becomes impossible. The purpose of the filtering is to eliminate the interference pattern from the spectrum. In the Fourier transformed NIR spectrum against the wavenumber there appears a discrete spectral component caused by the interference. This component can be recognized and cut off. After inverse Fourier transformation of such pretreated spectra, absorption peaks are free from interference modulation, so application of multivariate calibration methods is much more effective. With principal component analysis (PCA) on cluster plots, visual distinction between different plastics becomes possible. Correct class membership is provided by use of the Mahalanobis distance.

Multi-spectral infrared spectroscopy for robust plastic identification

Abraham Vázquez-Guardado, Mason Money, Nathaniel McKinney, and Debashis Chanda
Appl. Opt. 54(24) 7396-7405 (2015)

Identification of overlapping plastic sheets using short-wavelength infrared hyperspectral imaging

Keisuke Ozawa
Opt. Express 31(8) 12328-12338 (2023)

Near-infrared finger vein patterns for personal identification

Miyuki Kono, Hironori Ueki, and Shin-ichiro Umemura
Appl. Opt. 41(35) 7429-7436 (2002)

Identification of the interference spectra of edible oil samples based on neighborhood rough set attribute reduction

Shijun Xu, Wenbo Wu, Chuanxing Gong, Jinjian Dong, and Caifei Qiao
Appl. Opt. 62(6) 1537-1546 (2023)

Identification and quantification of 16 inorganic ions in water by Gaussian curve fitting of near-infrared difference absorbance spectra

Gerrit W. Steen, Elmar C. Fuchs, Adam D. Wexler, and Herman L. Offerhaus
Appl. Opt. 54(19) 5937-5942 (2015)