Ultra sensor based on color and UV-excited fluorescence images for predicting quality attributes of Shine-Muscat grape bunches at different maturity stages.

Fruit maturity at harvest is a major factor in determining its quality. In this study, external skin color of grape has been utilized to predict their chemical content and, in turn, the maturity of the fruit. Measurements of the chemical content such as Brix and acidity were made on ten bunches of "Shine Muscat" grapes at three different harvest periods (immature, mature, and overmature). Using a machine vision system, color and UV-excited fluorescence images of grape berries and bunches were taken during the respective harvest stages. Acquired images were processed using ImageJ to obtain RGB values. Rratio and a*/b* are strongly related to Brix and sugar-to-acid ratio of grape fruit, with regression coefficients of 0.5626 and 0.5180, repectively. It was found that a* color index was the best predictor of grape bunch maturity. Furthermore, discriminant analysis has shown that color images of grape berries perform better than 365 nm fluorescence images.