计算机工程与应用 ›› 2008, Vol. 44 ›› Issue (16): 18-22.

• 博士论坛 • 上一篇    下一篇

数学公式基线结构分析及识别算法研究

李永华1,王科俊1,上官伟1,唐立群2   

  1. 1.哈尔滨工程大学 自动化学院,哈尔滨 150001
    2.哈尔滨工程大学 计算机学院,哈尔滨 150001
  • 收稿日期:2007-12-24 修回日期:2008-04-07 出版日期:2008-06-01 发布日期:2008-06-01
  • 通讯作者: 李永华

Baseline structure analysis and recognition algorithm research of mathematical formula

LI Yong-hua1,WANG Ke-jun1,SHANGGUAN Wei1,TANG Li-qun2   

  1. 1.College of Automation,Harbin Engineering University,Harbin 150001,China
    2.College of Computer Science and Technology,Harbin Engineering University,Harbin 150001,China
  • Received:2007-12-24 Revised:2008-04-07 Online:2008-06-01 Published:2008-06-01
  • Contact: LI Yong-hua

摘要: 公式识别问题被分为字符分割和结构分析两部分内容。系统地研究了数学公式识别的全过程,使用自适应字符分割方法和基线结构分析算法成功地实现了一般数学公式的识别,识别率比较高,较好地完成了公式识别任务。从实验结果中可以看出,这种基于基线结构分析的数学公式识别方法能够满足大多数印刷体公式的识别,是一种较好的方法。

关键词: 数学公式, 公式识别, 字符分割, 基线结构分析

Abstract: The formula recognition problem was divided into character segmentation and structure analysis.In this paper,the whole recognition process was studied in detail,using character over-segmentation method and BST character structure analysis algorithm,the general formula could be recognized and remerged successfully,and the recognition ratio is very high.We can conclude from the experiment result that the mathematical formula recognition method based on baseline structure analysis can satisfy the need of most situations.

Key words: mathematical formula, formula recognition, character segmentation, baseline structure analysis