首页> 外国专利> METHOD FOR PROCESSING UNDERLINE IN RECOGNITION OF HANDWRITTEN NUMERAL

METHOD FOR PROCESSING UNDERLINE IN RECOGNITION OF HANDWRITTEN NUMERAL

机译:手写数字识别中的下划线处理方法

摘要

PURPOSE:To surely separate an underline from a numeral even when the underline crosses with a character line by delivering data to a numeral recognition processing after reconfiguring the data by eliminating a branch comprising only the underline from thinning data so as to decide the branch comprising only the underline from an underline candidate branch. CONSTITUTION:The underline is left as part of the numeral in an enter area where the underline exists when the lower part of a handwritten numeral is brought into contact with the underline, and only the underline is judged and eliminated when the underline crosses with part of the numeral or no underline crosses entirely with the numeral. In other words, the branches BR1-BR13 are decided by an endpoint and branch points E1-E7 after the handwritten numeral is read and a thinning processing is applied, and also, the branch comprising only the underline W and the branch comprising the underline W and the numeral in common are discriminated and taken out from all branches and the branch comprising only the underline W is discriminated and taken out. After that, the branch comprising only the underline W is eliminated, and image data is regenerated newly, then, numeral recognition is performed.
机译:目的:即使在下划线与字符线交叉的情况下,也可以通过从稀疏数据中删除仅包含下划线的分支以决定仅包含下划线的分支,从而在重新配置数据后将数据传递给数字识别处理,从而确保将下划线与数字分开下划线候选分支中的下划线。组成:下划线在手写数字的下部与下划线接触时留在存在下划线的输入区域中,作为数字的一部分;当下划线与下划线的一部分相交时,仅判断并消除了下划线数字或没有下划线完全与数字交叉。换句话说,分支BR1-BR13由端点和分支点E1-E7确定,在读取手写数字并应用间隔剔除处理之后,并且,仅包括下划线W的分支和包括下划线W的分支从所有分支中区分和取出共同的数字,并且从仅下划线W的分支中区分和取出。之后,去除仅包括下划线W的分支,并且重新生成图像数据,然后执行数字识别。

著录项

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