Improving optical fruit sorting by non-destructive determination of quality parameters affecting wine quality

摘要

The scope of this project is finding suitable parameters correlated with grape and wine composition to be implemented as a quality criteria to sort harvested grapes. Grape quality is not only defined by sugar concentration but by a complex equilibrium of parameters depending on the type of wine to be produced. Moreover, measuring berry composition is not always linked to the amount that will finally be extracted into wine. Sorting the berry according to their density and berry size showed that measuring berry size could be another good tool to segregate groups. It was possible to relate berry color with sugar accumulation and anthocyanin accumulation until berry coloration process (veraison). However, after veraison, berry color was not a suitable parameter anymore to describe differences in maturity. It was also possible to correlate berry color with aroma precursors in Riesling white grapes, showing that VIS measurements in white grapes could indeed be a possible way to sort the berries according to quality. A multiparametric optical sensor, Multiplex (Force A, France) showed that change in simple fluorescence ratio in green excitation was correlated to sugar concentration in white berries. Two ratios showed promising results: Firstly, the NBI_G indicating amino acid in berries and juice then secondly the ANTH_G correlating red color pigments (anthocyanin) concentration. This will have potential to determine quality parameters and in combination with user-friendly interfaces will improve grape sorting devices.