Journals Articles Conferences News About CLP
Submit a paper Email Alert
Search
Search
Articles
Journals
News
Advanced Search
Top Searches
2D MaterialsTransformation opticsQuantum PhotonicsSmall lasersSemiconductor UV PhotonicsSupersymmetric microring laser arrays

Laser & Optoelectronics Progress, Volume. 55, Issue 6, 061003(2018)

A Fast Global Registration Algorithm Based on Correcting Point Cloud Principal Component Coordinate System

Xu Chen and Bingwei He*
Author Affiliations
  • School of Mechanical Engineering and Automation, Fuzhou University, Fuzhou, Fujian 350108, China
    • show less
    AbstractGet PDF(in Chinese)
    Figures&Tables (12)
    Equations (0)
    References (26)
    Cited By (12)
    Tools
    Paper Information
    Topics
    Figures & Tables(12)
    PC coordinate system of Bunny point cloud model

    Fig. 1. PC coordinate system of Bunny point cloud model

    Download full size
    PC coordinate systems of two sets of similar point clouds

    Fig. 2. PC coordinate systems of two sets of similar point clouds

    Download full size
    (a) Position and pose of two sets of point clouds in initial PC coordinate system; (b) position and pose of two sets of point clouds in corrected PC coordinate system

    Fig. 3. (a) Position and pose of two sets of point clouds in initial PC coordinate system; (b) position and pose of two sets of point clouds in corrected PC coordinate system

    Download full size
    Test 1. (a) Initial position and pose; (b) proposed algorithm; (c) GO-ICP algorithm; (d) Super4PCS algorithm; (e) FDCP algorithm

    Fig. 4. Test 1. (a) Initial position and pose; (b) proposed algorithm; (c) GO-ICP algorithm; (d) Super4PCS algorithm; (e) FDCP algorithm

    Download full size
    Test 2. (a) Initial position and pose; (b) proposed algorithm; (c) GO-ICP algorithm; (d) Super4PCS algorithm; (e) FDCP algorithm

    Fig. 5. Test 2. (a) Initial position and pose; (b) proposed algorithm; (c) GO-ICP algorithm; (d) Super4PCS algorithm; (e) FDCP algorithm

    Download full size
    Test 3. (a) Initial position and pose; (b) proposed algorithm; (c) GO-ICP algorithm; (d) Super4PCS algorithm; (e) FDCP algorithm

    Fig. 6. Test 3. (a) Initial position and pose; (b) proposed algorithm; (c) GO-ICP algorithm; (d) Super4PCS algorithm; (e) FDCP algorithm

    Download full size
    Registration results after adding noise (σ=0.1). (a) Initial position and pose; (b) proposed algorithm; (c) GO-ICP algorithm; (d) Super4PCS algorithm; (e) FDCP algorithm

    Fig. 7. Registration results after adding noise (σ=0.1). (a) Initial position and pose; (b) proposed algorithm; (c) GO-ICP algorithm; (d) Super4PCS algorithm; (e) FDCP algorithm

    Download full size
    Registration results after adding noise (σ=0.2). (a) Initial position and pose; (b) proposed algorithm; (c) GO-ICP algorithm; (d) Super4PCS algorithm; (e) FDCP algorithm

    Fig. 8. Registration results after adding noise (σ=0.2). (a) Initial position and pose; (b) proposed algorithm; (c) GO-ICP algorithm; (d) Super4PCS algorithm; (e) FDCP algorithm

    Download full size
    Registration results after adding noise (σ=0.3). (a) Initial position and pose; (b) proposed algorithm; (c) GO-ICP algorithm; (d) Super4PCS algorithm; (e) FDCP algorithm

    Fig. 9. Registration results after adding noise (σ=0.3). (a) Initial position and pose; (b) proposed algorithm; (c) GO-ICP algorithm; (d) Super4PCS algorithm; (e) FDCP algorithm

    Download full size

    • Table 1. Eight different relationships between PC axes of two sets of point clouds

      View table

      Table 1. Eight different relationships between PC axes of two sets of point clouds

      RPCAGPCABPCA
      Q(1)111
      Q(2)11-1
      Q(3)1-11
      Q(4)1-1-1
      Q(5)-111
      Q(6)-11-1
      Q(7)-1-11
      Q(8)-1-1-1

    • Table 2. Results of the four algorithms

      View table

      Table 2. Results of the four algorithms

      AlgorithmTest 1Test 2Test 3
      Time /sRMSE /mmTime /sRMSE /mmTime /sRMSE /mm
      Proposed1.5720.0101.4070.0101.3780.010
      GO-ICP26.9910.01127.9460.01126.3940.011
      Super4PCS2.0540.0153.4710.0153.6880.015
      FDCP93.1730.05160.8910.01183.5620.011

    • Table 3. Results of the four algorithms after adding noise

      View table

      Table 3. Results of the four algorithms after adding noise

      Algorithmσ=0.1σ=0.2σ=0.3
      Time /sRMSE /mmTime /sRMSE /mmTime /sRMSE /mm
      Proposed2.3430.0132.6410.0143.5540.017
      GO-ICP25.1070.01125.4220.01725.4210.013
      Super4PCS15.5950.02224.5060.03030.0440.026
      FDCP23.4120.0305.8220.1546.0740.208
    Tools

    Get Citation

    Copy Citation Text

    Xu Chen, Bingwei He. A Fast Global Registration Algorithm Based on Correcting Point Cloud Principal Component Coordinate System[J]. Laser & Optoelectronics Progress, 2018, 55(6): 061003

    Download Citation

    EndNote(RIS)BibTexPlain Text
    Set citation alerts for article
    Save article for my favorites
    Paper Information

    Category: Image Processing

    Received: Oct. 27, 2017

    Accepted: --

    Published Online: Sep. 11, 2018

    The Author Email: Bingwei He (mebwhe@fzu.edu.cn)

    DOI:10.3788/LOP55.061003

    Topics

    laser devices and laser physics
    Lasers and Laser Optics
    Laser physics
    laser manufacturing
    Instrumentation, Measurement and Metrology