Fusion deep learning with pre-post harvest quality management of grapes within the realm of supply chain management
Downloads
Published
DOI:
https://doi.org/10.58414/SCIENTIFICTEMPER.2024.15.1.26Keywords:
Pre-Post harvesting, Machine learning, CNN, Computer vision, Supply Chain Management, Deep LearningDimensions Badge
Issue
Section
License
Copyright (c) 2024 The Scientific Temper
This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.
It is becoming increasingly vital in supply chain management to use different algorithms, particularly when it comes to pre and post-harvesting of grapes. This is especially true in the wine industry. Grapes must be processed both before and after harvesting as part of the management process for supply chains in the food industry. The grape bunch identification in vineyards was performed using machine learning at various stages of growth, including early stages immediately after flowering and intermediate stages when the grape bunch reached intermediate developmental stages. The machine learning method can predict annual grape output and also identify grape harvesting. The impressive performance of the pre-trained model shows that architecture training using different algorithms differs in the performance of grape predictions. We achieved 100% accuracy in grape prediction using LR, DT, RF, NUSVC, Adaboost and gradient algorithms, while KNN and SVC lag behind with an accuracy of 83.33% each. Our model includes the color and size of the grapes to differ in grape quality using a variety of grape images as a reference. It is capable of predicting the maturity stage of grapes by predicting Brix, TA and pH values (ranging between 18.20–25.70, 5.67–9.83 and 2.93–3.77) according to the size and color of grapes.We compared different algorithms and their performances by evaluating grape quality prediction accuracy, processing time and memory consumption.Abstract
How to Cite
Downloads