线纹辨识与特征值提取算法研究
CSTR:
作者:
作者单位:

1. 合肥工业大学仪器科学与光电工程学院合肥230009; 2. 安徽省计量科学研究院合肥230051

作者简介:

通讯作者:

中图分类号:

TB921;TN707

基金项目:

安徽省计量科学研究院项目(W2014JSKF0454)、国家自然科学基金(51275149)资助项目


Study on algorithm for line scale identification andeigenvalue extraction
Author:
Affiliation:

1. School of Instrument Science and Optoelectronics Engineering, Hefei University of Technology,Hefei 230009, China; 2. Anhui Institute of Metrology, Hefei 230051, China

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:

    为解决线纹尺的线纹磨损、线纹划痕、摄像头成像失焦、曝光过弱或过强以及线纹尺照射光不均匀等对线纹辨识的影响,提出了一种线纹辨识及线纹中心线特征值提取算法:1)采用区域划分法确定每个线纹灰度极值的位置;2)通过极值加权法提取每个线纹中心线的特征值;3)对线纹中心线上的图像灰度值进行平滑处理后,再搜索线纹中心线端点坐标,并根据端点坐标值大小确定主线纹中心线的特征值。实验表明,所述的线纹辨识及线纹中心线特征值提取算法基本上可以消除上述问题对线纹辨识的影响,且线纹中心线的特征值提取正确。

    Abstract:

    To solve the problem of line scale identification affected by line scale wear, line scale scratch, camera imaging out of focus, weak or strong exposure and uneven illumination of the line ruler, an algorithm for line scale identification and central line eigenvalue extraction is proposed. Firstly, the location of each gray extremum of line scale is acquired by region division method.Then,the eigenvalue of each center line of line scale is determined by extremum weighted method. Finally, the gray value on each center line of image is firstly smoothed, and then the gray jump terminal coordinate value of each line scale is searched, and the central line eigenvalue of the main line scale is derived according to the coordinate value. The experimental results show that the algorithm for line scale identification and central line eigenvalue extraction can eliminate the influence of the problems above,and the eigenvalue extraction of each central line is correct.

    参考文献
    相似文献
    引证文献
引用本文

徐从裕,魏广智,丁晨,景加慧,杨雅茹.线纹辨识与特征值提取算法研究[J].电子测量与仪器学报,2017,31(12):1937-1942

复制
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:
  • 最后修改日期:
  • 录用日期:
  • 在线发布日期: 2018-01-24
  • 出版日期:
文章二维码