PERTANIKA JOURNAL OF SCIENCE AND TECHNOLOGY

The addition of pork to beef and mutton is a potentially fraudulent practice. This study aims to detect adulteration of pork in beef and mutton using visible near-infrared (Vis-NIR) spectroscopy. Classification models for Beef+Pork and Mutton+Pork were developed using principal component analysis (PCA) and linear discriminant analysis (LDA). PCA demonstrated that respiratory pigments and the Soret band influenced the classification of meat based on Vis-NIR spectra. The most effective model was achieved by combining PCA with linear discriminant analysis (PCA-LDA), utilizing the original Vis-NIR spectra. The PCA-LDA model achieved a calibration accuracy of 99% for Beef+Pork and 82.4% for Mutton+Pork, with prediction accuracies of 91.3% and 73.7%, respectively. These results demonstrate that Vis-NIR spectroscopy can be utilized to authenticate minced meat, providing a promising approach for on-site screening or halal verification.