|
|||||
|
|
| 基于知识的满文识别后处理 | |
| 引用本文： | 张广渊,李晶皎,王爱侠. 基于知识的满文识别后处理[J]. 计算机辅助工程, 2006, 15(3): 69-71 |
| 作者姓名： | 张广渊 李晶皎 王爱侠 |
| 作者单位： | 沈阳大学,信息工程学院,辽宁,沈阳,110044;东北大学,信息科学与工程学院,辽宁,沈阳,110004;东北大学,信息科学与工程学院,辽宁,沈阳,110004 |
| 摘 要： | 为提高手写满文的整体文本识别率，基于规则和统计的方法提出以构建语料为主、规则为辅的满文语言知识库，并将其他应用于满文手写体字符识别结果中候选字的后验概率统计中．小范围测试样本表明该方法的有效性较高． |
| 关 键 词： | 满文 字符识别 后处理 知识库 语料 |
| 文章编号： | 1006-0871（2006）03-0069-03 |
| 收稿时间： | 2006-06-06 |
| 修稿时间： | 2006-06-06 |
Manchu handwritten character recognition post-processing based on knowledge base |
|
| ZHANG Guangyuan,LI Jingjiao,WANG Aixia. Manchu handwritten character recognition post-processing based on knowledge base[J]. Computer Aided Engineering, 2006, 15(3): 69-71 | |
| Authors: | ZHANG Guangyuan LI Jingjiao WANG Aixia |
| Abstract: | To optimize the recognition accuracy rate of handwritten Manchu character, the Manchu language knowledge base based on rule and statistics is presented, which is mainly constructed by corpus with the rules aided. The knowledge base is applied in the candidate item selection of handwritten Manchu character recognition post-processing. The Experimental results indicate that the method has more efficiency. |
| Keywords: | Manchu language character recognition post-processing knowledge base corpus |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ | |