A Mixed-Model Approach to Ripeness Classification using Convolutional Neural Networks and Support Vector Machines in the Mango Industry

Abstract

This study proposes a hybrid method that uses a convolutional neural network (CNN) and a support vector machine (SVM) to determine when mangoes are ripe. A crucial agricultural activity that boosts production productivity and decreases overages during storage is sorting mangoes according to maturity. Current approaches for mango ripeness classification might be made more efficient and accurate with the help of the proposed hybrid model. A dataset consisting of around one thousand photographs of ripe, unripe, and overripe mangoes was used to train and evaluate the hybrid CNN-SVM model. The suggested hybrid approach combines SVM classification precision with CNN's capacity to extract features from visual input. The hybrid model outperforms both deep learning and conventional machine learning in trials, with an impressive accuracy rate of 98.53%. These findings show that hybrid models may be used to determine when mangoes are ripe, which might help farmers make better decisions. Hybrid model, convolutional neural networks, support vector machines, agricultural image analysis, mango ripeness categorization