| Title: | 大规模古藉电子化关键技术及实现 |
| Author: | 王晓波
|
| Issued Date: | 2000
|
| Major: | 计算机应用技术
|
| Degree Grantor: | 中国科学院软件研究所
|
| Place of Degree Grantor: | 中国科学院软件研究所
|
| Degree Level: | 博士
|
| Keyword: | 古籍电子化
; XML标准
; 软件环境
|
| Abstract: | 本文在总结作者亲自参与的《文渊阁四库全书》电子版的基础上，探讨了大规模古籍电子化的关键技术及实现。本文探讨了OCR相关的前后处理技术，使OCR真正实际应用于大规模古籍电子化的工程上。根据特征提取的概念，舍弃图像上的细枝末节，构造了一个自上而下的版面分析系统，并开发了OCR后处理的一系列校对工具。使《四库全书》电子化的错误率见到万分之一以下，达到了重点出版物的出版标准。为衡量OCR的准确度，文中提出了一个在工程中行之有效的独特方法来衡量OCR的可信度，这在工程及理论上都有意义。文中探讨了UNICODE在大规模古籍电子化的实现技术：显示、全文检索及跨平台等技术实现。为实现在Internet上享用古藉电子化，文中结合XML标准，构造了一个古籍电子化的软件环境。 |
| English Abstract: | Summarizing author's experiences in the engineering of digitization for Wenyuange Siku Quanshu, the thesis discusses the key technique of digitization for large scale Chinese ancient books and introduces its implementation. The thesis focuses on the processing technique before and after OCR, PreOCR and PostOCR, which enables OCR to be really utilized in the engineering of digitization of large scale books. Extracting the major features while ignoring the minor details of images, author constructs a Layout Analysis System using top-down method, and developed a several proof-reading tools to achieve the publication national standard with error rate less than 0.01%. Based on the statistics of OCR parameter DISTANCE, author establishes a unique method to evaluate the reliability of OCR output. The method has been proved useful in the engineering, and it seems meaningful in theory. The thesis introduces another major technique in the large scale digitization engineering: Unicode enabling for Chinese text display, full text retrieval as well as the single data/single binary for multilingual Windows. Finally, the thesis introduces how author made a software environment for digitizing such large scale ancient books with XML technology in order to put the e-publication onto Internet. |
| Language: | 中文
|
| Content Type: | 学位论文
|
| URI: | http://ir.iscas.ac.cn/handle/311060/5792
|
| Appears in Collections: | 中科院软件所
|
| File Name/ File Size |
Content Type |
Version |
Access |
License |
|
|---|
| LW002151.pdf（1697KB） | -- | -- | 限制开放 | -- | 联系获取全文 |
|
| Recommended Citation: | |
王晓波. 大规模古藉电子化关键技术及实现[D]. 中国科学院软件研究所. 中国科学院软件研究所. 2000-01-01.
|
|
|
