北京邮电大学学报

  • EI核心期刊

北京邮电大学学报 ›› 2020, Vol. 43 ›› Issue (3): 112-117.doi: 10.13190/j.jbupt.2019-165

• 研究报告 • 上一篇    下一篇

基于主成分分析与迭代最近点的三维膝关节配准

王小玉, 陈琳   

  1. 哈尔滨理工大学 计算机科学技术学院, 哈尔滨 150080
  • 收稿日期:2019-08-13 出版日期:2020-06-28 发布日期:2020-06-24
  • 作者简介:王小玉(1971-),女,教授,E-mail:wngxiaoyu@hrbust.edu.cn.
  • 基金资助:
    国家自然科学基金项目(60572153,60972127);黑龙江省教育厅科学技术项目(12541177)

Three-Dimensional Knee Joint Registration Based on Principal Component Analysis and Iterative Closest Point

WANG Xiao-yu, CHEN Lin   

  1. School of Computer Science and Technology, Harbin University of Science and Technology, Harbin 150080, China
  • Received:2019-08-13 Online:2020-06-28 Published:2020-06-24
  • Supported by:
     

摘要: 针对膝关节与假体空间坐标系不一致的问题,提出了一种基于主成分分析(PCA)与迭代最近点(ICP)算法相结合的三段式配准方法.采用PCA结合两次ICP算法的配准策略,对膝关节与关节假体的点云数据进行PCA配准处理,然后对得到的初始配准结果采用ICP算法进行调整,最后对调整好姿态的点云数据再次进行ICP配准,从而将其空间坐标轴调整到一致.实验结果表明,三段式配准方法相较于其他算法,可在保持较高配准精度的同时缩短配准时间.

关键词: 点云配准, 迭代最近点, 主成分分析, 膝关节

Abstract: In view of inconsistency between knee joint and prosthesis in spatial coordinate system, a three-stage registration method based on principal component analysis (PCA) and iterative closest point (ICP) algorithm is proposed, which adopts PCA and ICP twice. Firstly, the point cloud data of knee joint and joint prosthesis is registered by PCA. then ICP algorithm is used to adjust the initial registration results. Finally, ICP registration is carried out again for the adjusted point cloud data, so as to adjust its spatial coordinate axis to be consistent. Experiments show that compared with other algorithms, the three-stage registration method can keep high registration accuracy and shorten the registration time.

Key words: opint cloud registration, iterative closest point, principal component analysis, knee joint

中图分类号: