Please use this identifier to cite or link to this item:
https://repository.uksw.edu//handle/123456789/25452
Title: | Pengklasifikasian Aksara Jawa Metode Convolutional Neural Network |
Authors: | Hanindria, Ivan Sukma |
Keywords: | aksara jawa;Convolutional Neural Network;klasifikasi;citra |
Issue Date: | 30-May-2022 |
Abstract: | Secara garis besar bentuk Aksara Jawa terbagi menjadi 3 set aksara yaitu Dasar,Pasangan dan Sandhangan. beberapa karakter dalam Aksara Jawa memiliki wujud yang mirip sehingga dapat menambah tingkat kerumitan proses pengenalan. Metode Convolutional Neural Network (CNN) merupakan salah satu algoritma klasifikasi gambar dengan Langkah menggabungkan beberapa lapisan data untuk melakukan pengolahan,Tujuan penelitian kali ini mengukur seberapa efektif algoritma Convolutional Neural Network dalam pengklasifikasian menggunakan Akasara Jawa Dasar.Percobaan menggunakan 20 kelas data aksara jawa yang masing masing terdapat untuk tiap folder berisi 108 citra.Pada penelitian ini klasifikasi aksara jawa dengan metode Convolutional Neural Network (CNN) bisa melakukan pengklasifikasian dengan tingkat persentase akurasi 85%. Dari hasil terbukti dapat mengelompokkan aksara jawa “Ka” dan aksara jawa “Nya”. In General, the form of the Javanese script is divided into 3 sets of scripts, namely Basic, Pasangan and Sandhangan. Some characters in Javanese script have similar shape definitions so that it can be the hardest part of the recognition process. The Convolutional Neural Network (CNN) method is one of the image classification algorithms with convolution operations that combines several processing layers. The purpose of this study is to measure how effective the Convolutional Neural Network algorithm is in classifying using Basic Javanese script. The experiment uses 20 classes of Javanese script data, each there are for each folder containing 108 images. In this study,the classification of Javanese characters with the Convolutional Neural Network (CNN) can perform classification with accuracy percentage 85%. From the results it is proven that it can classify the Javanese script "Ka" and the Javanese script "Nya". |
URI: | https://repository.uksw.edu/handle/123456789/25452 |
Appears in Collections: | T1 - Informatics Engineering |
Files in This Item:
File | Description | Size | Format | |
---|---|---|---|---|
T1_672018048_Judul.pdf | 1.08 MB | Adobe PDF | View/Open | |
T1_672018048_Daftar Pustaka.pdf | 239.74 kB | Adobe PDF | View/Open | |
T1_672018048_Isi.pdf Until 2999-01-01 | 552.58 kB | Adobe PDF | View/Open |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.