GRAMMATICAL EVOLUTION FOR FEATURE EXTRACTION IN LOCAL THRESHOLDING PROBLEM

The various lighting intensity in a document image causes diffculty to threshold the image. The conventional statistic approach is not robust to solve such a problem. There should be different threshold value for each part of the image. The threshold value of each image part can be looked as classif...

Full description

Bibliographic Details
Main Authors: Go Frendi Gunawan, Sonny Christiano Gosaria, Agus Zainal Arifin
Format: Article
Language:English
Published: Universitas Indonesia 2012-07-01
Series:Jurnal Ilmu Komputer dan Informasi
Online Access:http://jiki.cs.ui.ac.id/index.php/jiki/article/view/197
id doaj-cd048d21642442f1a04cb741a017ba8f
record_format Article
spelling doaj-cd048d21642442f1a04cb741a017ba8f2020-11-25T01:10:15ZengUniversitas IndonesiaJurnal Ilmu Komputer dan Informasi2088-70512502-92742012-07-015210611110.21609/jiki.v5i2.197162GRAMMATICAL EVOLUTION FOR FEATURE EXTRACTION IN LOCAL THRESHOLDING PROBLEMGo Frendi GunawanSonny Christiano GosariaAgus Zainal ArifinThe various lighting intensity in a document image causes diffculty to threshold the image. The conventional statistic approach is not robust to solve such a problem. There should be different threshold value for each part of the image. The threshold value of each image part can be looked as classifcation problem. In such a classifcation problem, it is needed to find the best features. This paper propose a new approach of how to use grammatical evolution to extract those features. In the proposed method, the goodness of each feature is calculated independently. The best features then used for classification task instead of original features. In our experiment, the usage of the new features produce a very good result, since there are only 5 miss-classification of 45 cases. Variasi intensitas pencahayaan pada citra dokumen akan menyebabkan kesulitan dalam menentukan nilai threshold dari citra tersebut. Pendekatan statistik konvensional tidak cukup baik dalam memecahkan masalah ini. Dalam hal ini, diperlukan nilai threshold yang berbeda-beda untuk setiap bagian citra. Nilai threshold dari setiap bagian citra dapat dipandang sebagai masalah klasifikasi. Dalam permasalahan klasifikasi semacam ini, dibutuhkan pencarian fitur-fitur terbaik. Di sini diusulkan sebuah pendekatan baru untuk mengekstrak fitur-fitur tersebut dengan menggunakan grammatical evolution. Nilai kebaikan dari masing-masing fitur akan dihitung secara saling lepas. Dalam percobaan yang dilakukan, tampak bahwa penggunaan fitur-fitur baru tersebut menghasilkan hasil yang sangat baik. Hanya ditemukan 5 kesalahan pengklasifikasian dalam 45 kasus.http://jiki.cs.ui.ac.id/index.php/jiki/article/view/197
collection DOAJ
language English
format Article
sources DOAJ
author Go Frendi Gunawan
Sonny Christiano Gosaria
Agus Zainal Arifin
spellingShingle Go Frendi Gunawan
Sonny Christiano Gosaria
Agus Zainal Arifin
GRAMMATICAL EVOLUTION FOR FEATURE EXTRACTION IN LOCAL THRESHOLDING PROBLEM
Jurnal Ilmu Komputer dan Informasi
author_facet Go Frendi Gunawan
Sonny Christiano Gosaria
Agus Zainal Arifin
author_sort Go Frendi Gunawan
title GRAMMATICAL EVOLUTION FOR FEATURE EXTRACTION IN LOCAL THRESHOLDING PROBLEM
title_short GRAMMATICAL EVOLUTION FOR FEATURE EXTRACTION IN LOCAL THRESHOLDING PROBLEM
title_full GRAMMATICAL EVOLUTION FOR FEATURE EXTRACTION IN LOCAL THRESHOLDING PROBLEM
title_fullStr GRAMMATICAL EVOLUTION FOR FEATURE EXTRACTION IN LOCAL THRESHOLDING PROBLEM
title_full_unstemmed GRAMMATICAL EVOLUTION FOR FEATURE EXTRACTION IN LOCAL THRESHOLDING PROBLEM
title_sort grammatical evolution for feature extraction in local thresholding problem
publisher Universitas Indonesia
series Jurnal Ilmu Komputer dan Informasi
issn 2088-7051
2502-9274
publishDate 2012-07-01
description The various lighting intensity in a document image causes diffculty to threshold the image. The conventional statistic approach is not robust to solve such a problem. There should be different threshold value for each part of the image. The threshold value of each image part can be looked as classifcation problem. In such a classifcation problem, it is needed to find the best features. This paper propose a new approach of how to use grammatical evolution to extract those features. In the proposed method, the goodness of each feature is calculated independently. The best features then used for classification task instead of original features. In our experiment, the usage of the new features produce a very good result, since there are only 5 miss-classification of 45 cases. Variasi intensitas pencahayaan pada citra dokumen akan menyebabkan kesulitan dalam menentukan nilai threshold dari citra tersebut. Pendekatan statistik konvensional tidak cukup baik dalam memecahkan masalah ini. Dalam hal ini, diperlukan nilai threshold yang berbeda-beda untuk setiap bagian citra. Nilai threshold dari setiap bagian citra dapat dipandang sebagai masalah klasifikasi. Dalam permasalahan klasifikasi semacam ini, dibutuhkan pencarian fitur-fitur terbaik. Di sini diusulkan sebuah pendekatan baru untuk mengekstrak fitur-fitur tersebut dengan menggunakan grammatical evolution. Nilai kebaikan dari masing-masing fitur akan dihitung secara saling lepas. Dalam percobaan yang dilakukan, tampak bahwa penggunaan fitur-fitur baru tersebut menghasilkan hasil yang sangat baik. Hanya ditemukan 5 kesalahan pengklasifikasian dalam 45 kasus.
url http://jiki.cs.ui.ac.id/index.php/jiki/article/view/197
work_keys_str_mv AT gofrendigunawan grammaticalevolutionforfeatureextractioninlocalthresholdingproblem
AT sonnychristianogosaria grammaticalevolutionforfeatureextractioninlocalthresholdingproblem
AT aguszainalarifin grammaticalevolutionforfeatureextractioninlocalthresholdingproblem
_version_ 1725175898792001536