电子秤自动检定的示值数字识别方法
DOI:
CSTR:
作者:
作者单位:

作者简介:

通讯作者:

中图分类号:

基金项目:

陕西省自然科学基础研究计划资助项目(2024JC-YBMS-489);陕西省自然科学基础研究计划资助项目(2023-JC-YB-288);湖北省数字化纺织装备重点实验室开放课题(KDTL2020005);


Digital identification method for automatic calibration of electronic scales
Author:
Affiliation:

Fund Project:

  • 摘要
  • |
  • 图/表
  • |
  • 访问统计
  • |
  • 参考文献
  • |
  • 相似文献
  • |
  • 引证文献
  • |
  • 资源附件
  • |
  • 文章评论
    摘要:

    针对电子秤检定过程中人工读取记录示值数字等问题,提出一种基于模板匹配法的数码管字符识别算法;通过预处理提高电子秤示值图像的质量,提取了示值字符整体的规则布局的轮廓等特征;对整体示值图像采用投影法准确提取到单个示值图像,对于小数点,根据面积特征利用图像减法实现了精准提取;通过选取数码管数字构建模板使其与电子秤示值数字具有相同形状特征,采用相关系数法计算单个示值数字和模板图像的最大相关系数实现模板匹配识别示值数字;最后搭建实验平台进行砝码加载卸载实验,对动态变化的电子秤示值数字与穿线法进行对比识别验证,结果表明:算法均在50张少量实验图片时达到了100%的识别正确率,实现了精准的识别;超过50张实验图片进行大量实验验证直至500张实验图片,设计的模板匹配算法的识别效果优于穿线法,识别正确率达到了98%。

    Abstract:

    In order to solve the problems of manual reading and recording of the digital numerals in the process of electronic scale calibration, a digital tube character recognition algorithm based on the template matching method is proposed; the quality of the electronic scale numerals image is improved through pre-processing, and the overall regular layout of the numerals is extracted; the overall numerals image is accurately extracted by the projection method into a single numerals image, and the decimal point is accurately extracted by the image subtraction method based on the area characteristics. Accurately extracted; through the selection of digital tube numbers to build a template so that it has the same shape characteristics with the digital scale, using the correlation coefficient method to calculate the maximum correlation coefficient of a single digital scale and the template image to realize the template matching to identify the digital scale; finally, the experimental platform is built to carry out the experiment of loading and unloading the weights, and the dynamic changes of the digital scale and the threading method is compared with the identification and verification of the algorithms. The algorithm achieves 100% recognition correct rate in a small number of 50 experimental pictures, and realizes accurate recognition; more than 50 experimental pictures for a large number of experimental verification until 500 experimental pictures, the recognition effect of the designed template matching algorithm is better than that of the threading method, and the recognition correct rate reaches 98%.

    参考文献
    相似文献
    引证文献
引用本文

高伟,金守峰,昝杰,肖福礼,李毅,郭彩霞.电子秤自动检定的示值数字识别方法计算机测量与控制[J].,2025,33(7):210-218.

复制
相关视频

分享
文章指标
  • 点击次数:
  • 下载次数:
  • HTML阅读次数:
  • 引用次数:
历史
  • 收稿日期:2024-06-01
  • 最后修改日期:2024-07-11
  • 录用日期:2024-07-15
  • 在线发布日期: 2025-07-16
  • 出版日期:
文章二维码