Comparative study on the use of anthocyanin profile, color image analysis and near-infrared hyperspectral imaging as tools to discriminate between four autochthonous red grape cultivars from La Rioja (Spain)

Three independent methodologies were investigated to achieve the differentiation of red grapes from different grape varieties (Garnacha, Graciano, Mazuelo and Tempranillo) collected from five vineyards located in the D.O.Ca. Rioja. Anthocyanin chromatographic analysis, color image analysis and near infrared hyperspectral imaging were carried out for the grapes. Then, a Stepwise Linear Discriminant Analysis (SLDA) was developed for each data set in order to discriminate grapes according to their grape variety. As a result, using anthocyanin profile, color image analysis and near infrared hyperspectral imaging respectively, 88%, 54% and 100% of the samples were correctly classified in the internal validation process and 86%, 52% and 86% were correctly classified in the leave-one-out cross-validation process. [Display omitted] •Comparison of three tools to discriminate between grape cultivars has been studied.•Chromatographic analysis achieves good results to discriminate between varieties.•Hyperspectral image presents a good potential to discriminate between grape varieties.•Image analysis does not improve the results obtained using hyperspectral image.