Nurhasanah, Youllia Indrawaty and AMELIA DEWI, IRMA and Kurnia, Rizky Kharisma (2016) PENGENALAN KARAKTER BERDASARKAN TULISAN TANGAN MENGGUNAKAN EDGE DETECTION DAN SVM. In: Seminar Nasional Teknologi Informasi 2016.

Full text not available from this repository.

Abstract

Graphology is scientific method foridentification, evaluation, and to understandchararacter through handwriting pattern. Pattern recognition is basically a system to classifyobjects into categories or classes based on knowledge or statistical informationtaken from pattern. Margins and fontsize is a parameter that will be examined by handwriting. Human character recognition system usestwo methods that is edge detection and support vector machine (SVM). Input the handwriting image then do thegrayscale to fix the image to get a better quality, then resizingit to commesurate image size, then do theextraction process using edge detection with canny operator todeterminex axis and y axis, then thresholding for pixels separating process then thinning pixel to pixel decimation, next stage with svm classification to classify the weight on the training data, the weight of image who approaching the weight targetmeans the result of human character. The examination were conducted at eight samples with four categories that each categories had two samples. The conclusion gained from this research with character 1 percentage is 50% and 50%, character 2 is 50% and 100%, character 3 is 50% and 50%, character 4 is 100% and 100

Item Type:	Conference or Workshop Item (UNSPECIFIED)
Divisions:	Paper
Depositing User:	Asep Kamaludin
Date Deposited:	12 Nov 2018 00:51
Last Modified:	12 Nov 2018 00:51
URI:	http://eprints.itenas.ac.id/id/eprint/208

Actions (login required)

View Item