A Baybayin word recognition system

Abstract

Baybayin is a pre-Hispanic Philippine writing system used in Luzon island. With the effort in reintroducing the script, in 2018, the Committee on Basic Education and Culture of the Philippine Congress approved House Bill 1022 or the ”National Writing System Act,” which declares the Baybayin script as the Philippines’ national writing system. Since then, Baybayin OCR has become a field of research interest. Numerous works have proposed different techniques in recognizing Baybayin scripts. However, all those studies anchored on the classification and recognition at the character level. In this work, we propose an algorithm that provides the Latin transliteration of a Baybayin word in an image. The proposed system relies on a Baybayin character classifier generated using the Support Vector Machine (SVM). The method involves isolation of each Baybayin character, then classifying each character according to its equivalent syllable in Latin script, and finally concatenate each result to form the transliterated word. The system was tested using a novel dataset of Baybayin word images and achieved a competitive 97.9% recognition accuracy. Based on our review of the literature, this is the first work that recognizes Baybayin scripts at the word level. The proposed system can be used in automated transliterations of Baybayin texts transcribed in old books, tattoos, signage, graphic designs, and documents, among others.

Block-level Optical Character Recognition System for Automatic Transliterations of Baybayin Texts Using Support Vector Machine

“…In the study of Pino et al (2021a), the Baybayin OCR at the word level assumes that the input of the system is already in Baybayin. The system identifies the Latin equivalent of the already-assumed-Baybayin input word.…”

“…In this work, we present an algorithm that discriminates all the Baybayin words from Latin words on a text block. After identifying all the Baybayin words on the page, we use the scheme proposed in the study of Pino et al (2021a) to find the Latin transliterations of each Baybayin word. The system identifies all the words in the text image.…”

“…We used MATLAB (vR2020a) to code and implement the recognition system. Our proposed method uses the Baybayin OCR for word recognition proposed in the study of Pino et al (2021a), which relies heavily on the OCR for Baybayin characters presented in the study of Pino et al (2021b). Hence, each Baybayin character on the input image must satisfy all the assumptions set in the study of Pino et al (2021a, b), and are presented as follows:…”

“…All 𝑊(𝑝)�'s will be fed into the Baybayin Word Recognition System (𝐵𝑊𝑅𝑆) presented in the study of Pino et al (2021a). 𝐵𝑊𝑅𝑆 returns all the equivalent Latin forms of the input Baybayin word 𝑊(𝑝)�.…”

“…The use of the inception network improved the average validation accuracy to 96.2% (Nogra 2020). Mobile e-Learning applications have been developed based on these models (Nogra et al 2019a;Nogra 2020) (Pino et al 2021a). This is the first OCR system that can classify Baybayin at the word level, with a recognition accuracy of 97.9%.…”