Transliteration Lampung Script with Method Intensity of Character

Nanang Himawan Fauzi (Universitas Sanata Dharma)
Evannoah Rolimarch Pratama
Reynaldi Setiawan
Yakobus Aris Arvanto
Yoga Dwi Prasetyo

Conference SENATIK STT Adisutjipto Yogyakarta
Vol 4 (2018): Transformasi Teknologi untuk Mendukung Ketahanan Nasional [ ISBN 978-602-52742-0-6 ]

The goal of writing this scientific work is to explain one of the ways to read Lampung Script using pattern recognition method.  There are three main steps in lampungnese image transliteration, i.e. step to preprocessing, feature extracting, and classifying. Each character is made into a binary image that is as thin as possible and the same size. The feature of every character is extracted using intensity of character. For the classification of this pattern recognition is using 1-Nearest Neighbor. By using 120 data of Lampung’s script in 6 different font, we got the accuracy of success 95%. From this accuracy we consider that the experiment is a success.