Open AccessIdentification of Lampung Script Using K-Neighbor, Manhattan Distance And Population Matrix Algorithm
Author(s) -
Gladys Ivana Augusta,
Lukman Hakim,
Anna Gustina Zainal,
Hendy Tannady
Publication year - 2021
Publication title -
journal of physics. conference series
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.21
H-Index - 85
eISSN - 1742-6596
pISSN - 1742-6588
DOI - 10.1088/1742-6596/1933/1/012064
Subject(s) - character (mathematics) , image (mathematics) , feature (linguistics) , population , mathematics , k nearest neighbors algorithm , pattern recognition (psychology) , artificial intelligence , computer science , linguistics , geometry , philosophy , demography , sociology
Language is a communication tool that is used as interaction with others. The use of indigenous languages is decreasing and erasing over time. Lampung script is a script that becomes the identity of the province of Lampung. However, only a few native Lampung residents who know the Lampung script and the younger generation often write Lampung in Latin letters because it is considered easier. This study uses Optical Character Recognition to recognize Lampung characters in image images using a smartphone. This study uses the Pixel Population Matrix feature extraction to extract the characteristics of the characters. The distance calculation for each test character and database uses Manhattan Distance and is classified using K-Nearest Neighbor with the value of k is 3. The results of the character recognition process can be translated into Indonesian. Testing this application is done by taking image samples with several conditions. The script image tested is in the form of a screenshot image of printed writing with random fonts, a handwritten photo image, a screenshot image with a combination of several words and a photo image with a smaller size and random slope. The test results according to these conditions obtained an average percentage of success of 91.49%.