Estimation of Sugar Content in Wine Grapes via In Situ VNIR–SWIR Point Spectroscopy Using Explainable Artificial Intelligence Techniques

Abstract

Spectroscopy is a widely used technique that can contribute to food quality assessment in a simple and inexpensive way. Especially in grape production, the visible and near infrared (VNIR) and the short-wave infrared (SWIR) regions are of great interest, and they may be utilized for both fruit monitoring and quality control at all stages of maturity. The aim of this work was the quantitative estimation of the wine grape ripeness, for four different grape varieties, by using a highly accurate contact probe spectrometer that covers the entire VNIR–SWIR spectrum (350–2500 nm). The four varieties under examination were Chardonnay, Malagouzia, Sauvignon-Blanc, and Syrah and all the samples were collected over the 2020 and 2021 harvest and pre-harvest phenological stages (corresponding to stages 81 through 89 of the BBCH scale) from the vineyard of Ktima Gerovassiliou located in Northern Greece. All measurements were performed in situ and a refractometer was used to measure the total soluble solids content (°Brix) of the grapes, providing the ground truth data. After the development of the grape spectra library, four different machine learning algorithms, namely Partial Least Squares regression (PLS), Random Forest regression, Support Vector Regression (SVR), and Convolutional Neural Networks (CNN), coupled with several pre-treatment methods were applied for the prediction of the °Brix content from the VNIR–SWIR hyperspectral data. The performance of the different models was evaluated using a cross-validation strategy with three metrics, namely the coefficient of the determination (R2), the root mean square error (RMSE), and the ratio of performance to interquartile distance (RPIQ). High accuracy was achieved for Malagouzia, Sauvignon-Blanc, and Syrah from the best models developed using the CNN learning algorithm (R2>0.8, RPIQ≥4), while a good fit was attained for the Chardonnay variety from SVR (R2=0.63, RMSE=2.10, RPIQ=2.24), proving that by using a portable spectrometer the in situ estimation of the wine grape maturity could be provided. The proposed methodology could be a valuable tool for wine producers making real-time decisions on harvest time and with a non-destructive way.

The Estimation of Chemical Properties of Pepper Treated with Natural Fertilizers Based on Image Texture Parameters

“…The sugar content of potatoes was predicted based on hyperspectral imaging data by Rady et al [39] with R reaching 0.97 for glucose. The sugar content was also estimated for apples with a high R parameter (0.8861) using multispectral imaging [40] and wine grapes with R 2 > 0.8 based on visible and near infrared and the short-wave infrared point spectroscopic data [41]. Furthermore, the prediction of soluble solid content of apples with R > 0.8 was possible based on features extracted from laser light backscattering images [42].…”

Estimation of Soybean Internal Quality Based on Improved Support Vector Regression Based on the Sparrow Search Algorithm Applying Hyperspectral Reflectance and Chemometric Calibrations

“…The support vector machine (SVM) regression algorithm has a strong generalization ability and robustness, and it is especially suitable for predicting regression problems with limited sample sizes. Eleni Kalopesa [25] used SVR (support vector machine regression) to estimate the sugar content in wine grapes. When using SVR, it is necessary to consider the optimization problem of the C and g parameters to control the risk of overfitting and the lack of generalization ability [26].…”

Research on an Improved Non-Destructive Detection Method for the Soluble Solids Content in Bunch-Harvested Grapes Based on Deep Learning and Hyperspectral Imaging

“…In research using line-scan spectroscopy [11][12][13][14], the reflectance spectrum of multiple grapes can be obtained in one image, but the methods require the manual selection of the Region of Interest (RoI) to obtain the average spectral reflectance, which cannot be automatically detected in batches. In research using point spectroscopy [15,16], the reflectance spectrum of one grape can be obtained each time; still, this one-by-one detection method has a low efficiency. Therefore, the applicability of current research is limited: for example, non-destructive detection of a small amount of grape berries in the laboratory.…”