Mongolian characters are not only the main body language in the autonomous region of Inner Mongolia, but also have influence in the world, Computer information processing about Mongolian characters stem from the eighties of the 20th century. A variety of current characters processing system are almost using the keyboard-encoding to enter data, which cost much time and manpower to arrange a great deal of documents and materials. The recognition of printed Mongolian characters provides a new method to input Mongolian characters, causes the information processing about Mongolian characters entering into intelligent period,. Under such circumstances, we provide to design the recognition system of the Multi-Font Printed Mongolian Characters.Much more effort has been done in Character Recognition area domestically and abroad. In our country, many companies and research organizations focus on the research of Chinese Characters Recognition. Many systems adopt the syntactic pattern recognition based on character stroke. Hot spot focuses on feature selection and description, fast method to extract features of the stroke, associating Chinese characters recognition with Chinese syntax and semantics and so on.Based on completely digesting the method and technology of Chinese and Western character recognition, project team study the feature selection, extraction and matching about Mongolian characters. Under the help of National Nature Scientific Foundation, project team spent 3 years on the research on multi-font Mongolian characters and obtained following achievements:(1).Feature selection and description about Mongolian characters(2)Feature extraction algorithm about Mongolian characters(3)Matching algorithm about Mongolian characters(4)Converting from character element to character coding(5)Using Mongolian corpus to post-processing Besides completing the above basic work, Using Visual C++ 6.0 project team developed a Multi-Font Printed Mongolian characters recognition system which could be run on Windows, combining the recognition component and the editing component into a whole applied system..
从文字识别的角度研究蒙文字特征的选择,蒙文字特征的提取及匹配问题。在此基础上提供一个多字体印刷蒙文字识别系统,为蒙古文的计算机输入增加一种新的方法,提高蒙文信息处理的效率和水平,对更好地繁荣和发展少数民族文化有着十分重要的意义,对内蒙古自治区的社会进步和经济发展有着一定的推动作用。
{{i.achievement_title}}
数据更新时间:2023-05-31
居住环境多维剥夺的地理识别及类型划分——以郑州主城区为例
基于细粒度词表示的命名实体识别研究
基于协同表示的图嵌入鉴别分析在人脸识别中的应用
适用于带中段并联电抗器的电缆线路的参数识别纵联保护新原理
基于Pickering 乳液的分子印迹技术
联机手写蒙文字识别的研究
蒙文TrueType字型字体技术的研究
制作蒙文字幕机系统的研究
多字体印刷汉字识别性能提高和实用化问题研究