Prediction of pH and color in pork meat using VIS-NIR near-infrared spectroscopy (NIRS)

Abstract

The potential of near-infrared spectroscopy (NIRS) to predict the physicochemical characteristics of the porcine longissimus dorsi (LD) muscle was evaluated in comparison to the standard methods of pH and color for meat quality analysis compared to the pH results with Colorimeter and pH meter. Spectral information from each sample (n = 77) was obtained as the average of 32 successive scans acquired over a spectral range from 400-2498 nm with a 2-nm gap for calibration and validation models. Partial least squares (PLS) regression was used for each individual model. An R2 and a residual predictive deviation (RPD) of 0.67/1.7, 0.86/2, and 0.76/1.9 were estimated for color parameters L*, a*, and b*, respectively. Final pH had an R2 of 0.67 and a RPD of 1.6. NIRS showed great potential to predict color parameter a* of porcine LD muscle. Further studies with larger samples should help improve model quality.