首页 | 本学科首页   官方微博 | 高级检索  
     

支票图像中打印体字符的分割和识别
引用本文:娄震,陆建峰.支票图像中打印体字符的分割和识别[J].计算机工程,2003,29(9):134-136.
作者姓名:娄震  陆建峰
作者单位:南京理工大学计算机科学与技术系,南京,210094
基金项目:南京理工大学科研发展基金项目
摘    要:和通常文档中的打印体字符分割和识别问题相比,支票图像中的打印体字符的分割和识别有其一定的特殊性。针对这种特殊性,该文提出了一套简单实用的算法。对于分割问题,主要是采用投影的方法解决印章、边框和底纹的干扰问题,对于字符识别,主要是采用了基于模板和融合的方法,从而有效地解决了上述问题,取得了比较满意的分割效果和较高的识别率。

关 键 词:阈值  打印体字符识别  支票  分割
文章编号:1000-3428(2003)09-0134-03
修稿时间:2002年5月16日

Segmentation and Recognition of Printed Character in Check Image
LOU Zhen,LU Jianfeng.Segmentation and Recognition of Printed Character in Check Image[J].Computer Engineering,2003,29(9):134-136.
Authors:LOU Zhen  LU Jianfeng
Abstract:Compared with the problem of segmentation and recognition in common document, the problem of segmentation and recognition in check image has certain particularity. In the view of this particularity, this paper presents a set of simple and practical method. For the segmentation, it mainly adopts the projection method to overcome the interference of seal, frame and shading. For the recognition, the method based on pattern and fusion is used, which solves the above problems effectively. The satisfied segmentation result and high recognition rate is obtained.
Keywords:Threshold  Printed character recognition  Check  Segmentation
本文献已被 CNKI 维普 万方数据 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号