Research on the character recognition of Tibetan ancient document based on CNN
Yuehui Han, Yusheng Hao, Weilan Wang · Journal of Physics Conference Series · 2019
Tibetan ancient documents are treasures among the World Cultural Treasure, which is a valuable historical material for studying Tibetan politics, economy, culture and religion. At present, the research on Tibetan character recognition mainly focuses on printed and handwritten Tibetan, while the research on Tibetan ancient document character recognition lags behind. This paper presents a method of recognizing 562 common characters in Tibetan ancient document based on CNN, and the experimental results show that the recognition rate of this method is 99.85%, which proves the effectiveness of this method.