基于双目视觉的引磁片定位与测量研究
DOI:
CSTR:
作者:
作者单位:

聊城大学 物理科学与信息工程学院

作者简介:

通讯作者:

中图分类号:

TN919.8

基金项目:

中央引导地方科技发展专项资金计划资助


Research on Magnetic Sheet Positioning and Measuring based on Binocular Vision
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:

    为了提高大批量生产中超薄磁性车载手机支架铁片的测量精度,提出采用双目视觉技术对引磁片定位,并基于鞍点法建立坐标系实现引磁片的高精度测量。首先通过HALCON的系统标定实现引磁片图像对的立体校正;利用金字塔算法与带有缩放的亚像素形状模板匹配实现引磁片特征点的提取。然后基于归一化互相关NCC(Normalized Cross-Correlation)的灰度匹配算法快速完成特征点的立体匹配;结合双目测距原理、坐标仿射转换对特征点重构获取以左相机为参考系下的3D坐标。最后通过空间曲线拟合公式和点积运算实现引磁片的高精度测量。实验结果表明,测量半径的误差小于0.1 mm、误差率小于1%;测量厚度误差小于0.1 mm、误差率小于6%,可有效的对超薄引磁片进行定位及尺寸测量。

    Abstract:

    To improve the measurement accuracy of the ultra-thin magnetic car phone holder iron piece in mass production, This paper presents using binocular vision technology to locate the target area, and introduce saddle point method to achieve high precision measurement of thickness. Firstly, HALCON-based stereo calibration method is proposed to realize stereo correction of the image of the magnetic sheet. Subpixel template matching with scaling and pyramid algorithm are presented, magnetic sheet feature points are extracted accurately. Secondly, the stereo matching of feature points can be completed quickly according to gray-value-based template matching using the NCC(Normalized Cross-Correlation). The binocular parallax principle and coordinate affine transformation are combined to complete reconstruction of feature points. Finally, the magnetic sheet is measured with high precision by the space curve fitting formula and the dot product operation. The experimental results show that the error of the measurement radius is less than 0.1mm, the error rate is less than 1%; the error of the thickness is less than 0.1mm, the error rate is less than 6%, which can effectively locate and measure ultra-thin magnetic sheets.

    参考文献
    相似文献
    引证文献
引用本文

朱荣华,葛广英,张广世,杨晓蕊,孙群.基于双目视觉的引磁片定位与测量研究计算机测量与控制[J].,2019,27(8):40-43.

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:2019-01-18
  • 最后修改日期:2019-02-27
  • 录用日期:2019-02-27
  • 在线发布日期: 2019-08-13
  • 出版日期:
文章二维码