基于西夏古籍文字样本数据库设计与实现  

Desgination and Implementation of sample databases based on Tangut script

作　　者：孟一飞 杨文慧[1] 谢堂健 戴雪瑞 

机构地区：[1]宁夏大学物理与电子电气工程学院,宁夏回族自治区银川市750021 [2]北京交通大学电子信息工程学院,北京100044

出　　处：《电脑与信息技术》2017年第6期28-32,42,共6页Computer and Information Technology

基　　金：国家自然科学基金(项目编号:61162020)

摘　　要：随着计算机技术的发展和在文字处理方面的应用,逐渐开展西夏文数字化的研究,西夏文的图像分割和识别对开展文物研究以及文献翻译具有非常重要的价值,但制约西夏文识别的关键技术在于西夏文数据库的建立。文章针对西夏文字的特点,设计了针对字符提取和样本数据库建立的具体流程,并讨论了数据库检索的组织和方法。依据连通域标识算法和边缘检测的原理进行字符提取,再将所提取的字符信息储存在我们指定的文本中。最终将所提取到的西夏文字与汉字进行匹配,保存后即完成一个包括西夏文字、汉字、以及它们多对应的序号等信息的样本数据库的创建,数据库的建立对西夏文字的识别提供了测试标准。With the development of computer technology and the application of word processing, the digital research ofTangut script is gradually carried out. the image segmentation and recognition of Tangut script has very important value forcarrying out cultural relics research and literature translation. but the key technology that restricts the recognition of Tangutscript is the establishment of database. According to the feature of the Tangut script. designed a specific progress ofcharacter extraction and sample database establishment. the organization and method of retrieval of database were discussedin this paper. The Tangut script image based on the principle of connection domain tagging algorithm and edge detectionfor character extraction. After the character mark is completed, then let the information of extracted character stored in ourspecified text. Finally, the extracted Tangut script characters are matched to the Chinese characters, after the preservation ofthe completion of a Tangut script text, Chinese characters, as well as their multi-corresponding serial number and otherinformation on the establishment of the sample database. The establishment of the database provides a test standard for theidentification of the Tangut script.

关 键 词：西夏文字 古籍文献 图像分割 样本数据库 

分 类 号：TP391.1[自动化与计算机技术—计算机应用技术]