Implementasi Deep Learning Menggunakan Cnn Untuk Klasifikasi Tingkat Kematangan Buah Jeruk Berbasis Android

Indonesia is a country that heavily relies on the agricultural sector, including various types of horticultural commodities, especially fruits. One example is oranges, which have many benefits and a sweet, refreshing taste. To obtain the best flavor and freshness, fully ripe oranges are the preferred choice. However, the process of recognizing the ripeness of oranges still faces many challenges. With advances in computer technology, particularly through smartphones, many human tasks can now be performed more efficiently and practically. One useful technology is computer vision, which can be used to automatically identify and determine the ripeness of oranges. This research aims to implement Convolutional Neural Networks (CNN) to measure the model's performance and ensure its capability in classifying the ripeness of oranges. The results of the research show that classification using CNN with the VGG-16 architecture achieved a training accuracy of 96% and a validation accuracy of 97%.