Near Infrared Spectrometry and Pattern Recognition as Screening Methods for the Authentication of Virgin Olive Oils of Very Close Geographical Origins

Abstract

The authentication of foods requires the use of sophisticated and expensive analytical techniques. Thus, there is a need for new, fast and inexpensive analytical methodologies for use as effective screening methods. This paper proposes a NIR spectroscopy-based method for discriminating virgin olive oils of two very similar and geographically close denominations of origin, viz. “Siurana” and “Les Garrigues”, which are made from at least 90% of olives of the Arbequina variety. Two chemometric techniques, artificial neural networks (ANNs) and logistic regression (LR), were tested as classifying tools applied to NIR spectra. The results obtained were quite satisfactory in both cases, in spite of the similarity between the two denominations of origin. The proposed method is intended to fill a gap in the authentication of natural products and allow the discrimination of oil samples, and can be applied to the discrimination of other olive oils which belong to different denominations of origin.