首页> 外国专利> Process for real-time optical character recognition in an automation system and automation system

Process for real-time optical character recognition in an automation system and automation system

机译:自动化系统和自动化系统中实时光学字符识别的过程

摘要

The invention concerns a procedure (100) for real-time optical character recognition in an automation system (200), comprehensive:Detect the blocks of characters (400) in an input image (300) in a detection step (101);Determine a first time period as a time period needed to perform the detection step (101) in a first determination step (103);determine a second period as a difference between the predetermined period and the first period in a second stage of determination (105);Determine a detection period for each detected character block (400) in a third determination step (107);verify that a sum of the specified detection periods of the detected blocks (400) is greater than the second time period in a verification step (109);If the sum of the specific recognition periods of the detected blocks (400) is greater than the second time period, select from the detected blocks (400) of blocks for which a sum of the respective recognition periods is not greater than the second time period, in a selection step (111); andRecognize the character units (316) of the selected character blocks (400) within the second time period in a recognition step (113).The invention also concerns an automation system (200) to perform the process (100).
机译:本发明涉及一种用于自动化系统(200)的实时光学字符识别的过程(100),综合:在检测步骤(101)中检测输入图像(300)中的字符(400)的块;确定在第一确定步骤(103)中执行检测步骤(101)所需的时间段的第一时间段;确定第二个时段,作为预定时段与第二阶段的第一个时段之间的差异(105);确定第三确定步骤(107)中每个检测到的字符块(400)的检测周期;验证检测到的块(400)的指定检测周期之和大于验证步骤(109)中的第二时间段;如果检测到的块(400)的特定识别周期的总和大于第二时间段,则从检测到的块的检测块(400)选择相应识别周期的总和不大于第二时间段,在选择步骤(111)中;和识别识别步骤(113)中的第二时间段内所选字符块(400)的字符单元(316)。本发明还涉及一种自动化系统(200)来执行该过程(100)。

著录项

  • 公开/公告号DE102019134387A1

    专利类型

  • 公开/公告日2021-06-17

    原文格式PDF

  • 申请/专利权人 BECKHOFF AUTOMATION GMBH;

    申请/专利号DE201910134387

  • 发明设计人 TOBIAS KEMPE;MARKO TSCHEREPANOW;

    申请日2019-12-13

  • 分类号G06K9;

  • 国家 DE

  • 入库时间 2022-08-24 19:25:18

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