Word Search Method for Traditional Mongolian Script

The aim of this research is to develop character recognition techniques for a Mongolian script which can be used in the digitization process of contents written in traditional Mongolian script. In this paper, we introduce a word search system developed for Mongolian-specific documents, as an initial result of the research goal. To have quick and robust searching result, we develop a center-line-matching method which uses layout features of Mongolian script. Experimental results show that our proposed technique outperforms conventional methods in terms of accuracy and computational cost.