对粘连和缺损数字串分割的研究  

THE SEGMENTATION OF NUMERICAL CHARACTERS WITH DEFORMITY AND CONGLUTINATION

作　　者：高庆吉[1] 王晓华[1] 赵为平[1] 

机构地区：[1]东北电力学院自动控制系,吉林132012

出　　处：《模式识别与人工智能》2000年第1期99-102,共4页Pattern Recognition and Artificial Intelligence

摘　　要：把缺损和粘连的字符串分割成单个字符以便后续识别,是字符识别的关键问题之一。本文以货运列车车号为例,提出一种有效的分割方法。该方法以弧特征和变基准线区域分割为基础,引入了识别反馈分割策略。将其应用于实际运行的识别系统时获得了较高的正确分割率。It is a normal question to segment the characters in digital image with deformity and conglutination. Taking an example of characters on freight train, a method is presented in this paper. It based on recognition feedback. Also arc characteristic and variable base - line are used. High rate of correct recognition is obtained in practical system.

关 键 词：数字识别 数字串分割 字符识别 图像识别 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]