SUB-WORD STRUCTURE ANALYSIS OF UYGHUR LANGUAGE FOR UYGHUR CHARACTER RECOGNITION(International Workshop on Advanced Image Technology 2006)

概要

論文の詳細を見る
This paper presented a novel approach for the Uyghur character recognition based on structure analysis of a typical Uyghur text. Uyghur language is a Turkic dialect spoken by over 10 million Uyghur people in Central Asia including about 8.5 million in Northwestern China. Uyghur text is written from right to left and always cursive. It constructed by the 126 different shapes of 32 characters. The recognition of Uyghur character is particularly difficult due to the necessity of segmentation even for printed text. In order to understand the Uyghur word structure, it becomes necessary to analyze the structure of Uyghur text. The results of the analysis showed that the sub-word in Uyghur language is the basic block rather than the word. The approach suggested here is classify the sub-words according to the numbers of the characters they contain, and separate the text into three groups. Based on this result, we also discuss that design of online Uyghur character recognition systems are would be done based on the sub-word as the basic block of recognition, also it would be much applicable.
社団法人電子情報通信学会の論文
2006-01-02