Determinación del punto de adquisición de datos para la clasificación de variedades de patata mediante tecnología NIRS

Resumo

The determination of the optimal area for the acquisition of NIR spectral data is a matter of discussion given that not all the food products have the same composition and homogeneity. Some products are very heterogeneous and the area of the data acquisition could interfere in the quantitative and qualitative determinations made by NIRS. Potato (Solanum tuberosum L.) has a relatively uniform internal matrix; however, it has been proved that the final results can vary depending on the area of the data acquisition. The purpose of the present work is to determine the optimal area for the data acquisition for an efficient classification of varieties. 672 samples of raw tubers corresponding to 50 different varieties were used in this study. Samples were scanned at 3 different areas, the first one located where the tuber joins the stolon, the second at the central axis and the third at the place opposite the first one. NIR spectral data were collected using a Luminar 5030 Miniature "Hand held" AOTF-NIR (Acousto-Optic Tunable Filter-Near Infrared) Analyser (Brimrose) in the reflectance mode. A discriminant analysis was performed to classify the samples using SPSS software (version 21). An independent analysis for each area of acquisition was carried out. Also, different pre-treatments of the data were performed in order to obtain more accurate results. The best classification rates were achieved for the spectral data obtained at the central axis of the samples.