Implementation of hyperspectral image analysis for evaluating table grape quality on bunch and berry level

Typically, subjective, and visual methods are used by grape growers to assess harvest maturity. These methods may not accurately represent the maturity of an entire vineyard – especially if extensive and representative sampling was not used. New technologies have been investigated for improved harvest management decisions. Spectroscopy methods utilizing the near-infrared region of the light spectrum is one such technology investigated as an alternative to classic methods, particularly the application hyperspectral imaging (HSI) a spectroscopic technique that obtains hundreds of images at different wavelengths collecting spectra data for each pixel in the sample have gained attention in research. In this study HSI in the visible-near infrared (VNIR) and short-wave infrared (SWIR) regions was tested for determining total soluble solids (TSS), pH and total acidity (TA) of table grapes under laboratory conditions. Imaging of whole bunches as well as individual berries selected from top, middle and bottom positions (front and back) of bunches was done. The bunches and berries were collected over multiple samples sessions throughout the 2022-23 and 2023-24 seasons allowing for fruit of various maturity levels (unripe up to harvest maturity) to be included. Spectral data was extracted from the images using MATLAB codes developed for this purpose using different analysis approaches. The resulting data is spatial and numerical yielding mean reflectance per pixel across points, entire bunches, and entire berries. Models for determining grape chemical composition were developed using ground-truthing data collected from the samples fruit. Initial results indicate that this method could be suitable for determining grape ripeness on a bunch or berry level.