Answers are grounded in this paper and cite their sources.
This research aims to build an avocado fruit ripeness identification system using digital image processing technology, focusing on the Convolutional Neural Network (CNN) method. Avocado, with its high-fat content and rich nutritional value, is a high-value commodity, but determining the correct level of ripeness remains a major challenge in avocado trading. Current ripeness identification methods often rely on visual observation, which is difficult to perform because avocados do not undergo color changes when ripe. Therefore, an automated system that can identify the ripeness level of avocados is needed. This study uses a dataset from Mendeley and applies CNN architecture to classify the ripeness levels of avocados into four categories: raw, semi-ripe, ripe, and overripe. However, the results of model training show challenges in overcoming overfitting and class imbalance. Recommendations for further research include adding regularization techniques such as dropout and expanding data variation with augmentation, as well as increasing samples in minority classes through oversampling.
No takes yet. Share an insight, caveat, or question.
Yosa Adytia Pratama (2024) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: