Garlic (Allium sativum L.) is a valuable medicinal plant that is widely used in pharmaceutical, food, cosmetic, and hygiene industries due to the various products it contains in the form of extracts and powders. Non-destructive and rapid methods for assessing food quality are of interest to industries and research. In the present study, garlic powder was determined by spectrometry in the range of 936 to 1660 nm and chemometric techniques. For this purpose, 120 spectra were recorded from the prepared samples of garlic powders during the storage period of 3, 90, and 360 days. The effects of SNV, MSC, D1+SG, and D2+SG preprocessing methods on the performance of ANN, SVM, and KNN classifiers were investigated. The principal component analysis (PCA) technique was used to reduce the spectral variables, and the first 4 principal components (PCs) were considered as the input of the models. The D1+SG preprocessing showed the greatest effect on the identification of the spectra of the garlic powder samples. Results showed that the highest classification accuracy of 96.88% was achieved in all the three classifiers. The results of the present study confirm the feasibility of using the near-infrared (NIR) technique and chemometrics for the rapid identification of the freshness of garlic powder.
