Article Per Year (5 Year)
p-Index From 2020 - 2025
0.23
P-Index
This Author published in this journals
All Journal Jurnal Pengembangan Teknologi Informasi dan Ilmu Komputer
Kirana Sekar Ayu
Fakultas Ilmu Komputer, Universitas Brawijaya
Computer Science & IT Control & Systems Engineering Education Electrical & Electronics Engineering Engineering

Published : 1 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 1 Documents
Search
Journal : Jurnal Pengembangan Teknologi Informasi dan Ilmu Komputer

 1 
Rancang Bangun Sistem Tingkat Kemanisan Buah Sky Rocket Melon menggunakan Metode Gray Level Co-Occurrence Matrix dan Backpropagation Neural Network Kirana Sekar Ayu; Fitri Utaminingrum
Jurnal Pengembangan Teknologi Informasi dan Ilmu Komputer Vol 5 No 8 (2021): Agustus 2021
Publisher : Fakultas Ilmu Komputer (FILKOM), Universitas Brawijaya

Show Abstract | Download Original | Original Source | Check in Google Scholar

Abstract

One of the most popular types of melon in Indonesia is the Sky Rocket Melon. Melon fruit has a sweet flesh taste, yellowish green skin with a texture shaped like a net, and has a distinctive odor. The increasing interest in melons every year must be balanced with melon production. The postharvest sorting process is very important to determine the quality, sales, and prices in the market. One of the sorting processes is based on the level of sweetness of the melon. Measuring the size of the sweetness of melons can use a refractometer brix, but this is not practical because you have to split the melon first. Therefore, an innovation was made to determine the level of sweetness of melons by analyzing digital images based on the texture of the net found on the melon skin. This study uses 5 features of the Gray Level Co-Occurrence Matrix method, namely Dissimilarity, Homogeneity, Contrast, Correlation, and Energy with variations in the distance values ​​d=1,2,3 and θ = 0°, 45°, 90°, 135° and Backpropagation Neural Network to perform classification. This study uses a dataset of 375 data which will be divided into training data and test data with a ratio of 4:1. In testing the number of epochs and the learning rate, the highest training accuracy was obtained at 87% at 80,000 epochs and a learning rate of 0.1, at the values ​​of d = 2 and θ = 135°. The results of the epoch and learning rate tests are used to determine the values ​​of d and to be used in the integration test. The integration test obtained the highest accuracy of 82% on the bottom side of the melon with an average computation time of 2.2967 seconds.
Co-Authors Utaminingrum, Fitri