Extraction of Shape Characteristics in Kawi Letters

Abstract

The Javanese kawi script is basically an ancient script that appeared in the 18th to 16th centuries. Which the kawi script is also a derivative of the Pallawa script, considering the very importance of the Kawi language because previous texts mainly used the history of Hindu texts still using the Kawi language. (Surada, 2018). And to preserve the kawi language, it can be developed through a kawi identification system. For image identification, shape feature extraction is used for the identification system to find information from digital images (Herdiansah, 2022). In this study, an image identification system was created by extracting shape features to get the initial value of the shape of the Kawi letters and then going through a classification process so that they could recognize Kawi letters. There was some data collected from 2 data sources with a total of 6 data collected. In this manufacture using the matlab application (Prayoga, 2019) by conducting tests to process image data that has been obtained using shape feature extraction with metric and eccentricity parameters which are then processed using an artificial neural network method. From the results of this study it is hoped that the extraction value of shape features with these parameters can be used for the next step for classifying kawi letters.