Klasifikasi Citra Pada Tingkat Kematangan Buah Pisang Menggunakan Algoritma Deep Learning

Abstract

This research aims to classify images to predict banana ripeness based on skin color features using a deep learning algorithm method. The research method used is a quantitative approach, where the population in this study is to collect data by downloading datasets from Mendeley & Kaggle, the dataset taken is 200 images of bananas, namely 100 images of ripe bananas & and 100 images of unripe bananas. The sample selection technique uses classified images through the pre-processing stage, the tomato fruit image will be divided into 2 methods for testing the Self-Organizing Maps Algorithm model, namely 80% training data and 20% testing data