CASIA OpenIR  > 毕业生  > 硕士学位论文
Thesis Advisor刘迎建 ; 吴显礼 ; 刘昌平
Degree Grantor中国科学院研究生院
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
Keyword邮政 自动分拣
Abstract本文分析了邮政自动分拣系统的结构和其实现的过程。 该系统首先对信封图象进行预处理,包括对图象进行平滑、二值化、去噪声和 倾斜矫正等过程,并且介绍了求连通域的一些方法。然后对信封进行版面分析,包括 对信封图像版面的结构分析和位置分析,对信封图像基于规则的连通域合并算法把 版面分割一行行的块,然后基于版面信息选取出需要的邮政编码块和地址块,在这里 我们还分析了地址线剔除的方法. 在版面分析的基础上得到邮政编码块和地址块,我们对邮政编码块要剔除邮政 编码的方框,取出数字图像。对地址块的中文要进行单字分割,提出单个的汉字图 像信息。我们这里的合并过程,包括初步的合并和进一步处理的过程,我们把可能 的分割路径都予以保留。经过这些处理后将单个的数字图像和中文图像分别送入相 应的识别核心进。 最后根据中国地名字典对邮政编码和中文地址的识别结果进行基于知识规则 的理解,并从多条路径中选出最优路径,得到正确的结果。从而实现了邮政信封的自 动分拣。
Other AbstractIn this paper, our research mainly focuses on A Post Automatic Slection System, we analysis its structure and the process of how to realize this system. The first part of the system is the pretreatment of post image and document analysis, including image smoothing, filtering, figure background separation, boundary detection, image translation, document segmentation, run-length smearing, and Hough transform. At last we get post code image and address image from the whole. Then we get off the forms which are around numbers. And segment address image into single words image, which is called handwriting segmentation. In follow, the characters in every block are recognized. Then the result is further understood by using knowledge and Address Library, we get the best result and get out which is the best route of handwriting segmentation. After the process, the information of envelope is imported to computer automatically.
Other Identifier618
Document Type学位论文
Recommended Citation
GB/T 7714
白真龙. 邮政自动分拣系统[D]. 中国科学院自动化研究所. 中国科学院研究生院,2001.
Files in This Item:
There are no files associated with this item.
Related Services
Recommend this item
Usage statistics
Export to Endnote
Google Scholar
Similar articles in Google Scholar
[白真龙]'s Articles
Baidu academic
Similar articles in Baidu academic
[白真龙]'s Articles
Bing Scholar
Similar articles in Bing Scholar
[白真龙]'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.