Non-invasive spectroscopic and imaging systems for prediction of beef quality in a meat processing pilot plant

Abstract

This study evaluated a range of diffuse reflectance spectroscopic (Vis-NIR spectrophotometers) and imaging (Hyperspectral imaging cameras) instruments for predicting pH, IMF and shear force values of beef in a meat processing pilot plant. A total of 364 beef striploin samples were evaluated and prediction models were developed using PLSR. Models for pH and IMF (except Vis snapshot camera) showed good fit with high Rcv2 (0.29–0.92) and low SECV values. Good prediction accuracy with high Rp2 (0.44–0.90), RPD and low SEP values was also observed. While low values of Rp2 for shear force was observed, the expected curvilinear relationship between predicted values of shear force and predicted values of pH was observed suggesting that spectroscopic measurements were able detect biophysical factors associated to these two attributes. Overall, it can be concluded that diffuse reflectance spectroscopy combined with chemometrics has a great potential to be used as an on/in-line quality monitoring system for the meat processing industry.