Pengembangan Perangkat Lunak Untuk Pengelompokkan Tumbuhan Berdasarkan Citra Digital Daun Menggunakan CNN

Abstract

This study aims to develop software to classify plants based on digital images of leaves, the data is used as training data is 222 data which is divided into 4 data groups. While the testing data is used in this study is 45 data. The method is used in this study is Convolutional Neural Network which is applied using Tensorflow Keras. From results of the study, algorithm produces an accuracy value of 20% on testing data and 96.29% on training data with 50 epochs.