International Journal of Scientific Engineering and Applied Science (IJSEAS) – Volume-3, Issue-1,Jan 2017 ISSN: 2395-3470 www.ijseas.com Recognizing ancient Sinhala Inscription Characters using Neural Network Technologies K.G.N.D. Karunarathne 1 , K.V. Liyanage 2 , D.A.S. Ruwanmini 3 , G.K.A. Dias 4 , S.T. Nandasara 5 1,2,3,4,5 University of Colombo School of Computing, Colombo, Sri Lanka Abstract Recognizing ancient Sinhala inscription characters enable archeologists to reveal historical events in ancient Sri Lanka. Currently, this is done by the archaeology experts with a huge effort. The inefficiency of this manual procedure will negatively impact on the future research in field of archaeology. This research involves in developing an application with Optical Character Recognition (OCR) functionality to recognize ancient Sinhala inscription. This paper focus on the OCR module of the application. OCR module comprises of the technologies of Artificial Neural Network (ANN) and Convolutional Neural Network (CNN). Experiments were carried out to evaluate the recognition rate of the two OCR technologies which performs on train data, test data (preprocessed) and test data (real images). After evaluating each OCR solution, CNN was selected as the best resulted OCR solution. Lack of data is the main limitation of this research and it will be highly impacted on the OCR accuracy. As a result, 9 characters were identified by the CNN OCR engine. Keywords: Epigraphy, Sinhala Inscriptions, Optical character recognition (OCR) 1. Introduction Many Inscriptions found in ancient cities of Anuradhapura and Polonnaruwa in Sri Lanka. Thonigala inscription, Mirror wall , Galpotha (Stone Book) inscription (see Fig. 1.1) are some of them. Figure 1.1: Galpotha (Stone Book) Inscription in Polonnaruwa These inscriptions are very important as they are the major sources of getting information about ancient Sri Lanka. These inscriptions also provide valuable information about the time, place and situation connected with the inscription and the evolution of the languages over centuries. [1][2] Revealing the content of these inscriptions will be highly valuable to investigate the history of ancient Sri Lanka. Currently the content of each of these inscriptions are translated to modern Sinhala language manually by an archaeology expert who has specialized knowledge to understand the ancient scripts. The Inscriptions letters are read through human eye with great difficulty and this manual procedure would be time consuming. Although the Inscriptions were used as a one of the information source to recognize expansion of Sinhala language, recognizing content of these inscription becomes a huge challenge due to various reasons. They may be damaged and/or partially erased. Lack of specialized knowledge and lack of available resources for inscription reading are also another major problem. Currently, the existing archaeology experts have to make considerable amount of effort to read inscriptions. Further they also have to recognize the characters of these inscriptions manually. The main aim of the research is to recognize each isolated character set of these inscriptions using a proper character recognition process and map them into modern Unicode characters. 1.1 Research Objectives The following are the objectives of our research. • Recognize the ancient characters and digitize them. • Map ancient character identified with modern Sinhala character • Obtain Sinhala interpretation of Sri Lankan inscriptions via the character recognition feature. 37