CASIA OpenIR  > 毕业生  > 硕士学位论文
基于象素屏蔽的表格提取研究
宁屹
Subtype工学硕士
Thesis Advisor吴显礼
1999-06-01
Degree Grantor中国科学院自动化研究所
Place of Conferral中国科学院自动化研究所
Degree Discipline模式识别与智能系统
Abstract本论文的研究主要集中在表格处理的表格提取方面。 针对表格提取的具体任务与要求,文中系统地分析了基于象素屏 蔽的表格提取的处理过程和具体方法.。并且,在基本方法的基础上, 根据实际的情况,提出了改进的算法和一些独特的构想。 在表格提取的校倾阶段,文中提出了对于通用算法的改进---多重 过滤连通算法,利用该算法,能较干净地找到图象中的长直线。 在表格提取的坐标对应阶段,文中将坐标对应问题转化为一个最 优化问题,并且给出了满意解的算法。 在表格提取的字符修补阶段,文中提出了屏蔽前后图象对比找断 点的方法,并且对于数字,给出了修补算法。
Other AbstractThis thesis mainly focuses on Form Dropout, which is a branch of Form Processing. According to the purpose and realistic request of Form Dropout, this paper systematically analyzes the main procedures in Form Dropout system, based on the pixel masking method. Furthermore, the author improves some algorithms, and presents some new ideas. The main points of this paper are as following: First, in stage of skew correction, a Multi-filter Connected Component Algorithms is proposed, which aims at finding long straight lines in form image. Second, in stage of coordinate correspondence, this paper transforms the problem into an optimization problem, and give out the algorithms to find a satisfactory solution. Third, in stage of broken stroke reconstruction, this paper proposes a method, based on comparison of an image before and after masking, to find the position of the broken strokes. And give out the algorithms to reconstruct broken numerals.
shelfnumXWLW525
Other Identifier525
Language中文
Document Type学位论文
Identifierhttp://ir.ia.ac.cn/handle/173211/7273
Collection毕业生_硕士学位论文
Recommended Citation
GB/T 7714
宁屹. 基于象素屏蔽的表格提取研究[D]. 中国科学院自动化研究所. 中国科学院自动化研究所,1999.
Files in This Item:
There are no files associated with this item.
Related Services
Recommend this item
Bookmark
Usage statistics
Export to Endnote
Google Scholar
Similar articles in Google Scholar
[宁屹]'s Articles
Baidu academic
Similar articles in Baidu academic
[宁屹]'s Articles
Bing Scholar
Similar articles in Bing Scholar
[宁屹]'s Articles
Terms of Use
No data!
Social Bookmark/Share
All comments (0)
No comment.
 

Items in the repository are protected by copyright, with all rights reserved, unless otherwise indicated.