计算机科学

计算机科学 ›› 2023, Vol. 50 ›› Issue (11A): 230100028-5.doi: 10.11896/jsjkx.230100028

• 图像处理&多媒体技术 • 上一篇    下一篇

基于特征变换结合KD树改进ICP的快速点云配准方法

唐佳林, 林寿南, 周壮, 司炜, 王腾辉, 郑泽鑫   

  1. 北京理工大学珠海学院 广东 珠海 519088
  • 发布日期:2023-11-09
  • 通讯作者: 林寿南(2663560093@qq.com)
  • 作者简介:(01068@bitzh.edu.cn)
  • 基金资助:
    广东大学生科技创新培育专项资金(pdjh2021 a0625,pdjh2022 b0712)

Improved ICP Fast Point Cloud Registration Method Based on Feature Transformation Combined with KD Tree

TANG Jialin, LIN Shounan, ZHOU Zhuang, SI Wei, WANG Tenghui, ZHENG Zexin   

  1. Beijing Institute of Technology,Zhuhai,Zhuhai,Guangdong 519088,China
  • Published:2023-11-09
  • About author:TANG Jialin,born in 1982,Ph.D,associate professor,is a member of China Computer Federation.His main research interests include artificial intelligence,data science and computer vision.
    LIN Shounan,born in 2001,undergra-duate,bachelor degree.His main research interests is computer vision.
  • Supported by:
    Guangdong University Student Science and Technology Innovation Cultivation Special Fund Grant(pdjh2021 a0625,pdjh2022 b0712).

PDF (PC)

236

摘要： 点云配准是三维重建的关键技术。针对迭代最近点(ICP)算法存在收敛速度慢、配准效率低、配准时间长等难题,提出了一种基于特征变换结合 KD树改进ICP的快速点云配准方法。首先利用体素网格法进行初步降采样,在其差分高斯模型上获取三维尺度不变特征变换(SIFT)关键点;其次建立快速点特征直方图(FPFH);然后使用采样一致性初始配准(SAC-IA)算法,实现粗配准;最后根据得到的初始变换矩阵使用KD树改进的ICP算法,实现精配准。在斯坦福大学公开数据集上进行配准实验,结果表明,与ICP算法相比,所提改进算法具有较高的配准精确度和时间效率,且可为精确配准选择较优的初始位姿。文中在一定程度上避免了点云配准时存在的局部最优现象,为后续目标识别匹配和三维重建提供了一种高效的方法。

关键词: 特征变换, 采样一致性, 快速点特征直方图, 迭代最近点

Abstract: Point cloud registration is the key technology of 3D reconstruction.Aiming at the problems of slow convergence speed,low registration efficiency and long registration time in iterative closest point(ICP) algorithm,a fast point cloud registration method based on feature transformation combined with kdtree is proposed to improve ICP.First of all,the three-dimensional SIFT key points are obtained on the differential Gaussian model by down-sampling with voxel mesh method.Secondly,fast point feature histogram(FPFH) is established.Then sample consensus initial alignment(SAC-IA) algorithm is used to realize rough registration.Finally,according to the obtained initial transformation matrix and improved ICP algorithm based on KD tree,accurate registration is realized.Experimental results of Stanford data registration show that compared with ICP algorithm,the proposed algorithm has higher registration accuracy and time utilization,and can select a better initial pose for accurate registration.To some extent,this study avoids the local optimal phenomenon existing in point cloud collocation,and provides an efficient me-thod for subsequent target recognition and matching and 3D reconstruction.

Key words: Feature transform, Consistency of sampling, Fast point feature histograms, Iterative closest point

中图分类号: 

  • TP391
[1]GUO Y,WANG H,HU Q,et al.Deep Learning for 3D Point Clouds:A Survey[J].arXiv:1912.12033,2020.
[2]HUANG X,MEI G,ZHANG J,et al.A comprehensive survey on point cloud registration[J].arXiv:2103.02690,2021.
[3]BESL P J,MCKAY H D.A method for registration of 3-Dshapes[J].IEEE Transactions on Pattern Analysis & Machine Intelligence,1992,14(2):239-256.
[4]YAO Z,ZHAO Q,LI X,et al.Point cloud registration algorithm based on curvature feature similarity[J].Measurement,2021,177(11):109274.
[5]LIU Y Z,ZHANG Q,LIN S.Improved ICP Point Cloud Registration Algorithm Based on Fast Point Feature Histogram[J].Laser & Optoelectronics Progress,2021,58(6):611003.
[6]YUE X F,LIU Z Y,ZHU J,et al.Coarse-fine point cloud registration based on local point-pair features and the iterative closest point algorithm[J/OL].Applied Intelligence,2022:12569-12583.https://doi.org/10.1007/s10489-022-03201-3.
[7]LI Z,ZHUKUN L.Virtual Namesake Point Multi-Source Point Cloud Data Fusion Based on FPFH Feature Difference[J].Sensors,2021,21(16).
[8]CARVALHO L E,VON WANGENHEIM A.3D object recognition and classification:a systematic literature review[J].Pattern Analysis and Applications,2019:1-50.
[9]LU J,WANG W,SHAO H,et al.Point Cloud Registration Algorithm Fusing of Super 4PCS and ICP Based on the Key Points[C]//2019 Chinese Control Conference(CCC).2019.
[10]MATTHEW Y,CHRIS P,JOSH M,et al.Sparse point cloudregistration and aggregation with mesh-based generalized iterative closest point[J].Journal of Field Robotics,2021,38(8).
[11]LOW D G.Distinctive Image Features from Scale-Invariant Keypoints[J/OL].International Journal of Computer Vision 60,2004:91-110.https://doi.org/10.1023/B:VISI.0000029664.99615.94.
[12]RUSU R B,BLODOW N,BEETZ M.Fast Point Feature Histograms(FPFH) for 3D registration[C]//IEEE International Conference on Robotics & Automation.2009.
[13]ZHAO M,HUANG Z,SONG T,et al.Point cloud registration method based on sample consensus initial alignment and iterative closest point algorithm[J].Laser Journal,2019,40(10):45-50.
[14]SHI X J,LIU T,HAN X.Improved Iterative Closest Point(ICP) 3D point cloud registration algorithm based on point cloud filtering and adaptive fireworks for coarseregistration[J].International Journal of Remote Sensing,2020,41(8):3197-3220.
[15]SHI X J,LIV T,HAN X.Improved ICP Point Cloud Registration Algorithm Based on Fast Point Feature Histogram[J].Laser &Optoelectronics Progress,2021,58(6).
[16]DAMODARANS,SUDHEER A P,SUNIL KUMAR T K.An evaluation of spatial mapping of indoor environment based on point cloud registration using Kinect sensor[C]//2015 International Conference on Control Communication & Computing India(ICCC).Trivandrum,India,2015:548-552.
Viewed
Full text


Abstract

Cited
  Shared   
  Discussed   
No Suggested Reading articles found!
渝ICP备16013121号-4
地址：重庆市渝北区洪湖西路18号    邮编：401121  
电话：023-63500828（编辑部） 023-67039612（会议/专题） 023-67039623（财务/发票）
E-mail：jsjkx12@163.com
本系统由北京玛格泰克科技发展有限公司设计开发