Abstract

Near-infrared, and more recently, mid-infrared diffuse reflection spectroscopy (more commonly and erroneously called reflectance spectroscopy) have come to be extensively used to determine the composition of products ranging from forages and drugs to soils. In these methods, spectra are generally collected as reflectance or <i>R</i> and transformed to log (1/reflectance). However, some near-infrared researchers do not transform the data, but use the data directly as reflectance. As it is generally held that procedures such as partial least squares regression do not work well with nonlinear data and the log (1/reflectance) transformation is held to be a best effort at linearization for near-infrared diffuse reflection spectral data, the question arises as to why then does not everyone transform the data? The objective of this work was to investigate this question using near-infrared and mid-infrared spectra in various formats. Calibrations were developed using spectral data from forages in several formats: reflectance, log (1/reflectance), non-background corrected single beam spectra, interferograms, and Kubelka–Munk transformed data. Calibrations were developed using both non-pretreated spectra and using data pretreatments such as derivatives. Results showed that calibrations using partial least squares regression did not require any specific data format. Accurate calibrations were developed for fiber, digestibility, and protein measures in forages using any of the aforementioned spectral formats including non-background-corrected single beam spectra and even interferograms. While calibrations could be developed using any of the formats, results indicated that those using Kubelka–Munk and especially interferograms did not perform as well as the others, although they were still quite good. In conclusion, results using forage spectra indicated that accurate and equivalent calibrations can be developed using diffuse reflectance data, with (reflectance) or without background correction (single beam spectra), or log (1/reflectance) at least when using partial least squares regression for calibration development.

Spectral data mining for rapid measurement of organic matter in unsieved moist compost

Somsubhra Chakraborty, David C. Weindorf, Md. Nasim Ali, Bin Li, Yufeng Ge, and Jeremy L. Darilek
Appl. Opt. 52(4) B82-B92 (2013)

Diffuse reflectance spectroscopy: a comparison of the theories

E. L. Simmons
Appl. Opt. 14(6) 1380-1386 (1975)

Extraction of information from laser-induced breakdown spectroscopy spectral data by multivariate analysis

Nicole Labbé, Isabel Maya Swamidoss, Nicolas André, Madhavi Z. Martin, Timothy M. Young, and Timothy G. Rials
Appl. Opt. 47(31) G158-G165 (2008)

Visible spectral dependence of the scattering and absorption coefficients of pigmented coatings from inversion of diffuse reflectance spectra

Fernando Curiel, William E. Vargas, and Rubén G. Barrera
Appl. Opt. 41(28) 5969-5978 (2002)

Optical diffuse reflectance accessory for measurements of skin tissue by near-infrared spectroscopy

R. Marbach and H. M. Heise
Appl. Opt. 34(4) 610-621 (1995)