Discrimination of Raw Pork, Chicken and Turkey Meat by Spectroscopy in the Visible, Near- and Mid-infrared Ranges

Abstract

Spectra of raw pork, turkey and chicken meat (n = 74 ) were recorded in the visible, near- and mid-infrared ranges. Discriminant models were initially developed separately in the mid-IR and the visible–near-IR; a number of discrete regions of the latter spectra were investigated. The best predictive model achieved using mid-IR spectra correctly classified 86.5% (32 of 37) of test samples; for visible–near-IR data, the optimum classification of 91.9% (34 of 37 samples) was achieved using data in the 400–1100 nm wavelength range. Using combined visible, near- and mid-IR data, the most accurate classification rate obtained was 94.6%.