首页> 外国专利> RECOGNITION SYSTEM OF HAND WRITING CHARACTER OF ONNLINE TYPE

RECOGNITION SYSTEM OF HAND WRITING CHARACTER OF ONNLINE TYPE

机译:在线型笔迹特征识别系统

摘要

PURPOSE:To increase the recognition function of ON-line hand writing characters, by using a stroke similarity calculator for the recognition of the characters many in the number of strokes and preventing the degradation of the recognizing capacity of characters due to miswriting order. CONSTITUTION:The elements of the similarity matrix between strokes between all the strokes of the character information X to be recognized and the standard pattern to be collated with the character information are detected at a similarity calculator between strokes 1, and a standard pattern storage unit 2, and the maximum element in all the columns or rows of the similarity matrix is detected at a row element maximum value detection circuit 4. Whether the result of detection of this circuit 4 is only one or a plurality is judged at a discrimination circuit 5. If the result of discrimination is affirmative, the maximum matrix element is added with respect to all the rows, and the result of addition is stored in a storage unit 3 of similarity between strokes. Further, if the result of discrimination is negative, as to the part of the result (minor matrix), the minor matrix elements selected from all the rows or columns of the minor matrix, are fed to a combining unit 3 with a minor matrix element combined unit 9.
机译:目的:通过使用笔划相似度计算器来识别许多笔画数目的字符并防止由于误写顺序而导致的字符识别能力下降,来提高在线手写字符的识别功能。组成:要识别的字符信息X的所有笔划之间的笔划之间的相似度矩阵和要与字符信息进行对照的标准图案,在笔划1和标准图案存储单元2之间的相似度计算器中进行检测然后,在行元素最大值检测电路4中检测相似矩阵的所有列或行中的最大元素。在判别电路5中判断该电路4的检测结果是一个还是多个。如果判别结果是肯定的,则对所有行相加最大矩阵元素,并且相加结果被存储在笔划之间相似度的存储单元3中。此外,如果判别结果为负,则关于结果的一部分(次要矩阵),将从次要矩阵的所有行或列中选择的次要矩阵元素与次要矩阵元素一起馈送到组合单元3。组合单元9。

著录项

  • 公开/公告号JPS55154669A

    专利类型

  • 公开/公告日1980-12-02

    原文格式PDF

  • 申请/专利权人 NIPPON TELEGRAPH & TELEPHONE;

    申请/专利号JP19790061146

  • 发明设计人 ODAKA KAZUMI;MASUDA ISAO;

    申请日1979-05-18

  • 分类号G06K9/62;G06K9/80;

  • 国家 JP

  • 入库时间 2022-08-22 16:46:14

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