首页> 外国专利> RECOGNITION OF HANDWRITING CHINESE CHARACTERS BASED ON DP MATCHING

RECOGNITION OF HANDWRITING CHINESE CHARACTERS BASED ON DP MATCHING

机译:基于DP匹配的手写汉字识别

摘要

This program allows the computer to recognize the letters written by the tablet mouse on the computer. More specifically, the pre-processing is performed by smoothing, normalizing, and normalizing the characters input by the writer. After extracting feature values such as absolute position value, relative position value, etc., find out the number of copies of Chinese characters, and modify the pattern recognition algorithm called dynamic pattern matching. Therefore, the purpose of this program is to find the number of copies in the kanji input by the writer, and to recognize the kanji faster and more efficiently by using stroke type values and various feature values. ; This program reduces the number of comparisons by setting the features with strokes without capturing the points when capturing the features. Dynamic pattern matching is also possible by obtaining stroke type values based on strokes. The dynamic pattern matching algorithm matches the next value of the relative feature value. However, Chinese characters are different in stroke number and stroke order, so those who are not familiar with Chinese characters can't use stroke number and stroke order. If the number of strokes and the stroke order are changed, it may not be recognized, so that the modified dynamic pattern matching can match the current feature value with the previous feature value of the relative stroke.If the writer writes a stroke, the overall stroke count is incorrect Because it loses, the stroke is attached based on the inflection point so that it can be recognized. Therefore, the number of strokes is reduced or the stroke order is changed so that it can be recognized. ; Chinese Character Recognition, Finding Copies, Strokes, Strokes
机译:该程序允许计算机识别平板电脑鼠标在计算机上写的字母。更具体地,预处理是通过平滑,归一化和归一化由书写者输入的字符来执行的。提取绝对位置值,相对位置值等特征值后,找出汉字的份数,并修改称为动态模式匹配的模式识别算法。因此,该程序的目的是在作者输入的汉字中查找份数,并通过使用笔划类型值和各种特征值更快,更有效地识别汉字。 ;此程序通过使用笔触设置要素而在捕获要素时不捕获点的情况下减少了比较次数。通过基于笔划获得笔划类型值,也可以进行动态模式匹配。动态模式匹配算法匹配相对特征值的下一个值。但是,汉字的笔划编号和笔划顺序不同,因此不熟悉汉字的人不能使用笔划编号和笔划顺序。如果笔划数和笔划顺序发生变化,则可能无法识别,因此修改后的动态模式匹配可以将当前特征值与相对笔划的先前特征值进行匹配。笔划数不正确因为笔划数丢失,所以会根据拐点附加笔划,以便可以识别它。因此,减少了笔划数或改变了笔划顺序,以便可以识别它。 ;汉字识别,查找副本,笔画,笔画

著录项

  • 公开/公告号KR20060104570A

    专利类型

  • 公开/公告日2006-10-09

    原文格式PDF

  • 申请/专利权人 KWON HEE YONG;

    申请/专利号KR20050026852

  • 发明设计人 KWON HEE YONG;JUN SANG YEOP;

    申请日2005-03-31

  • 分类号G06F3;

  • 国家 KR

  • 入库时间 2022-08-21 21:24:50

相似文献

  • 专利
  • 外文文献
  • 中文文献
获取专利

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号