Development of a Pre-Calibrated Multispectral Sensor Chipset for Beef Cuts Classification Using PCA-DA and PLS-DA Based Approach

Abstract

Beef cuts refer to different sections or portions which are typically obtained from specific muscles and are characterized by their tenderness, flavor, and cooking methods. A spectrometer instrument with an AS7265x spectroscopy sensor which has a 410nm-940nm spectral range was developed to classify the beef cuts. This instrument offers a prompt detection mechanism to safeguard consumers' rights and mitigate food fraud. This research used slice and mince beef cuts (brisket, shank, sirloin, tenderloin) with multivariate analyses and integrated with PCA, PCA-DA and PLS-DA techniques. The calibration accuracy and reliability of the model were >70%. The result shows that the preliminary research using a pre-calibrated multispectral sensor chipset has a good potential to develop as a portable rapid detector.