Jurnal Sistem Informasi dan Informatika (SIMIKA)
Vol. 7 No. 2 (2024): Jurnal Sistem Informasi dan Informatika (Simika)

IDENTIFIKASI JENIS MANGGA BERDASARKAN CIRI DAUN MENGGUNAKAN METODE CNN

Sayyid Ahmad Wisak (Unknown)
Nindy Aulia Safirah (Unknown)
Yampi R Kaesmetan (Unknown)



Article Info

Publish Date
25 Jun 2024

Abstract

Mango, originating from India and bearing the scientific name Mangifera indica L, spread to Southeast Asia, including Malaysia and Indonesia. Rich in vitamins A and C, mango boosts immunity and exhibits a wide genetic diversity. From the genetic diversity and types of mango leaves, many people do not understand well about the types of mangoes based on mango leaves. Therefore it is necessary to identify the type of mango based on the leaves so that people can easily understand the type of a mango. The Convolutional Neural Network (CNN) method proves effective in identifying plants based on morphological features. CNN, a development from Multilayer Perceptron (MLP), is employed in testing using Teachable Machine with 60 mango leaf images, divided into 3 classes. Across 4 different classifications, the average confusion matrix shows CNN accuracy at 83.30%, precision at 94.43%, and recall at 88.28%. With CNN, the accuracy in identifying mango leaf characteristics improves.

Copyrights © 2024






Journal Info

Abbrev

jsii

Publisher

Subject

Computer Science & IT Control & Systems Engineering

Description

Jurnal Sistem Informasi dan Informatika aims to provide scientific literature specifically on studies of applied research in information systems (IS), information technology (IT) and public review of the development of theory, method, and applied sciences related to the ...