[1]李晓 袁保社 陈卿 任宏宇 张建华[].基于像素积分投影的印刷体维文字母切分方法[J].计算机技术与发展,2012,(04):41-44.
　LI Xiao,YUAN Bao-she,CHEN Qing,et al.A Segmentation Method of Printed Uyghur Character Based on Projection Histogram of Pixels[J].,2012,(04):41-44.
点击复制

基于像素积分投影的印刷体维文字母切分方法()
分享到：

《计算机技术与发展》[ISSN:1006-6977/CN:61-1281/TN]

卷:
期数:
2012年04期
页码:
41-44
栏目:
智能、算法、系统工程
出版日期:
1900-01-01

文章信息/Info

Title:
A Segmentation Method of Printed Uyghur Character Based on Projection Histogram of Pixels
文章编号:
1673-629X（2012）04-0041-04
作者:
李晓1 袁保社1 陈卿1 任宏宇2 张建华[3]
[1]新疆大学信息科学与工程学院[2]94537部队气象台[3]新疆公众信息产业股份有限公司
Author(s):
LI Xiao; YUAN Bao-she; CHEN Qing; REN Hong-yu; ZHANG Jian-hua
[1]College of Information Science and Engineering,Xinjiang University[2]Meteorological Observatory,94537 Troops[3]Public Information Industry Co.,Ltd.of Xinjiang
关键词:
维吾尔文; 印刷体; 切分; 像素投影积分; 光学字符识别
Keywords:
Uyghur;  printed text;  segmentation;  pixels projection histogram;  optical character recognition（OCR）
分类号:
TP391.43
文献标志码:
A
摘要:
维吾尔文字属于左向连写文字,字母之间的连笔与变形使得切分字母很困难,印刷体维吾尔文字母的准确切分是识别的关键。文中试验了一种基于像素积分投影的印刷体维吾尔文字母切分方法,包括使用行水平投影切出文字行与文字基线,通过垂直投影切出单词及单词中不粘连的字母,结合水平投影与垂直投影数据,外加相邻投影谷距、字母宽度与基线像素值等信息,设置了细化的连体段字母切分规则。实验结果表明,该方法能够较为准确的将印刷体维吾尔文字母切分开,为OCR系统的准确识别提供了基础
Abstract:
The Uyghur language is a kind of the right to left and concatenate writing language,the connection and the deformation between the characters make the segmentation difficult,the accuracy of the segmentation of the characters is crucial to the recognition.Experiment a Uyghur characters segmentation method which based on the projection histogram of the pixels,including the use of the horizontal projection to separate the line of the text,separate the word and the character without adhesion in the word through the vertical projection,detailed a set of segmentation rules of the conjoined characters according to the information of the horizontal and the vertical projection,in addition the distance of the adjacent valleys of the projection,the width of the characters and the value of the pixel on the baseline,etc.Experimental results show that the method can segment the printed Uyghur text into characters effectively which provide a basis for the recognition of the OCR system

相似文献/References:

[1]沈洁 卡米力·木依丁 张祖莲.维吾尔文笔迹鉴别预处理及边缘提取方法研究[J].计算机技术与发展,2012,(04):65.
　SHEN Jie,KAMIL·Moydi,ZHANG Zu-lian.Pre-Processing Andedge Extraction Research on Uyghur Writer Identification[J].,2012,(04):65.
[2]陈卿 袁保社 李晓 任宏宇 张建华[].基于模板匹配的印刷维吾尔文字符识别研究[J].计算机技术与发展,2012,(04):119.
　CHEN Qing,YUAN Bao-she,LI Xiao,et al.Printed Uyghur Character Recognition Based on Template Matching[J].,2012,(04):119.

备注/Memo

备注/Memo:
工信部2009年度电子信息产业发展基金项目（工信部财[2009]453）李晓（1986-），男，湖北襄阳人，硕士研究生，研究方向力中文信息处理；袁保社，教授，研究方向为多语种系统平台与嵌入式技术
更新日期/Last Update: 1900-01-01