Food fraud, particularly in staple commodities like rice, poses significant risks to trademark protection as well as economic challenges. In this study, we explored the efficacy of Raman spectroscopy in differentiating authentic ST25 rice—a premium variety of Vietnam from other rice types, amidst concerns of adulteration. By utilizing both backscattering and transmission Raman spectroscopy, a total of 125 rice samples consisting of both commercial rice and Vietnamese landrace rice varieties were analyzed. All samples were categorized into two main groups namely ST25 and Non-ST25 for the construction of the classification model. Through principal component analysis (PCA) and k-nearest neighbors (kNN) classification method, we achieved classification accuracies of 81.58% for backscattering data and up to 97.37% for transmission data at elevated temperatures. Our findings highlight the efficacy of Raman spectroscopy for rice authenticity verification, nevertheless, modification in spectral measurement schemes is necessary to obtain better method reproducibility as well as to maintain high discriminatory accuracy of the classification model.
Read full abstract