JOMLAI: Journal of Machine Learning and Artificial Intelligence
Vol. 1 No. 4 (2022): December

Calligraphy Text Types Recognition Using Learning Vector Quantization

Mhd Furqan (Universitas Islam Negeri Sumatera Utara, Medan, Indonesia)
Abdul Halim Hasugian (Universitas Islam Negeri Sumatera Utara, Medan, Indonesia)
Ziqra Addilah (Universitas Islam Negeri Sumatera Utara, Medan, Indonesia)



Article Info

Publish Date
30 Dec 2022

Abstract

Calligraphy is the art of beautiful writing. The term calligraphy comes from simplified English (calligraphy) taken from the Latin word "kalios" which means beautiful, and "graph" which means writing or script. The art of writing Arabic letters is called the science of khat, known as the science of Arabic calligraphy or Islamic calligraphy. There are many types and varieties of Islamic calligraphy. Each has a different form and function. There are seven types of calligraphy that are popular and known by lovers of calligraphy art in Indonesia, such as, Khat Naskhi, Tsuluts, Farisi, Riq'ah, Diwani, Diwani Jali, and Kufi. The method commonly used to identify what type of calligraphy is made is by looking directly at the shape and characteristics of the calligraphy itself (calligraphy experts). Here the author tries to create a computerized calligraphy type recognition system using the Learning Vector Quantization method. Where this method is a method that works with each unit of output representing a class. So with this system, we can recognize the type of calligraphy text computerized. The accuracy value obtained in the results of calligraphy image recognition is 75%.

Copyrights © 2022






Journal Info

Abbrev

jomlai

Publisher

Subject

Computer Science & IT Engineering

Description

Focus and Scope JOMLAI: Journal of Machine Learning and Artificial Intelligence is a scientific journal related to machine learning and artificial intelligence that contains scientific writings on pure research and applied research in the field of machine learning and artificial intelligence as well ...