Open AccessDigitization of Batak Manuscripts Using Methods Learning Vector Quantization (LVQ)
Author(s) -
M A Muchtar,
Ivan Jaya,
Nandar Cholid Siregar,
Erna Budhiarti Nababan,
Syahril Effendi,
Opim Salim Sitompul,
Muhammad Zarlis,
U Andayani,
Tigor Hamonangan Nasution,
Ikhsan Siregar,
M F Syahputra
Publication year - 2020
Publication title -
iop conference series. materials science and engineering
Language(s) - English
Resource type - Journals
eISSN - 1757-899X
pISSN - 1757-8981
DOI - 10.1088/1757-899x/851/1/012066
Subject(s) - digitization , learning vector quantization , computer science , thresholding , artificial intelligence , information retrieval , pattern recognition (psychology) , vector quantization , computer vision , image (mathematics)
Digitalization of the Batak manuscript itself is still done manually. [2][4] Dr. Uli Kozok has made a Batak script font that can run on Windows, Linux, and Apple OSX but there is no system that can digitize Batak manuscripts properly. There is no system that can digitize the Batak manuscript itself automatically. Therefore, we need a system that can digitize Batak manuscripts. This system was built by applying the LVQ Method. Batak manuscript digitization in .jpg format is used as an image input file for the recognition process. Then image pre-processing including grayscale, contrast, thresholding, and segmentation is done in order to facilitate the recognition of Batak documents and recognition as a result of document recognition. The application of LVQ method can digitize the Batak manuscript well and in accordance with what is desired. This is supported by an accuracy rate of 97.9%. Based on system testing, the absence of training data in the input document, the letters are unclear, the position of the letters is not appropriate and the background in the document is still noise which greatly affects the success rate of the system in recognizing Batak script documents.