期刊文献+

非编码标志点自定位方法研究 被引量:1

RESEARCH ON SELF-LOCALIZATION METHOD OF NON-CODED MARK POINTS
下载PDF
导出
摘要 针对三维重建中非编码标志点和点云配准存在误匹配这一关键问题,设计了一种新的基于标志圆的自定位方法。本文构想出来了一种基于标记圆直径唯一性约束的新算法。首先,在校准的双目视觉系统上执行立体校正,使两个相机的成像平面共面并对齐行(列)。通过标记圆的最外缘两点与圆中心的距离等于圆直径的约束,在一定程度上得到了标记点的精确匹配。消除了匹配误差,计算了标记点的三维坐标。最后,根据标记圆间固定的空间拓扑关系,记录时间序列中的空间标记,同时又借助了LM优化算法实现局部坐标系和全局坐标系的转化。利用变换矩阵统一点云坐标系进而重建被测物体的三维形貌。实验结果表明,此种方法可以在复杂曲面的三维重建上得到较好的应用,并允许快速自动定位标记点。总测量误差小于0.125 mm,能够符合工业现场实际高精度测量要求。 In order to solve the mismatching key problem between non-coded marked points and point cloud registration during 3D reconstruction process,a new self-localization method based on marked circle was designed.Firstly,the image plane of binocular camera could be coplanar and row(column)alignment by means of rectifying binocular vision system.Traditional epipolar line constraint cannot proceed the robustness matching to artificial non-coded marks points.In this paper,we propose a new algorithm based on the uniqueness constraint of the circle diameter,using the distance of two points on the edge of a marked circle equals to the diameter of the circle,which can achieve accurate matching of marked point,eliminate mismatching points,and calculate 3D coordinates of the mark points.Finally,according to the space-invariant topological relationship between marked circles,registration of space marked circle of time serials can be achieved so that utilizes LM optimization algorithm fulfilling the transition from local coordinate system to global coordinate system.Thereby,based on transformation matrix unify point cloud coordinate of time serials,rebuild a 3D topography of the measured object.The experimental results show that this method is applicable to reconstruct the 3D topography of a complex surface,able to accomplish self-localization of marked point fast and automatically.The overall error in measurement is less than 0.125 mm,satisfies the requirement of high accuracy and real-time in industry scene.
作者 毛翠丽 马新波 唐沛 MAO Cui-li;MA Xin-bo;TANG Pei(School of Intelligent Manufacturing,Nanyang Institute of Technology,Nanyang 473000,China;Materials Department of Nanyang Public Works section,Zhengzhou Railway Bureau,Nanyang 473000,China)
出处 《南阳理工学院学报》 2021年第4期42-47,共6页 Journal of Nanyang Institute of Technology
基金 国家自然科学基金项目(51875164) 科技部支撑项目(2011BAK15B07) 河南省创新型科技人才队伍建设工程 河南省教育厅基金项目(21B535002)。
关键词 点云配准 立体校正 双目视觉系统 自定位 point cloud registration stereo rectification binocular vision system self-localization
  • 相关文献

参考文献8

二级参考文献75

共引文献174

同被引文献13

引证文献1

二级引证文献3

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部