Abstract

Instrumental techniques such a near-infrared spectroscopy (NIRS) are used in industry to monitor and establish product composition and quality. As occurs with other food industries, the Chilean flour industry needs simple, rapid techniques to objectively assess the origin of different products, which is often related to their quality. In this sense, NIRS has been used in combination with chemometric methods to predict the geographic origin of wheat grain and flour samples produced in different regions of Chile. Here, the spectral data obtained with NIRS were analysed using a supervised pattern recognition method, Discriminat Partial Least Squares (DPLS). The method correctly classified 76% of the wheat grain samples and between 90% and 96% of the flour samples according to their geographic origin. The results show that NIRS, together with chemometric methods, provides a rapid tool for the classification of wheat grain and flour samples according to their geographic origin. (C) 2013 Elsevier Ltd. All rights reserved.