• Acta Optica Sinica
  • Vol. 38, Issue 10, 1010005 (2018)
Min Zhao1、*, Qin Shu1、*, Wei Chen2, and Yunxiu Yang2
Author Affiliations
  • 1 College of Electrical Engineering and Information Technology, Sichuan University, Chengdu, Sichuan 610065, China
  • 2 Southwest Institute of Technical Physics, Chengdu, Sichuan 610041, China
  • show less
    DOI: 10.3788/AOS201838.1010005 Cite this Article Set citation alerts
    Min Zhao, Qin Shu, Wei Chen, Yunxiu Yang. Three-Dimensional Point Cloud Registration Algorithm Based on l p Spatial Mechanics Model [J]. Acta Optica Sinica, 2018, 38(10): 1010005 Copy Citation Text show less
    (a) Eigenvector set FP of target point cloud P'; (b) eigenvector set FQ of source point cloud Q'
    Fig. 1. (a) Eigenvector set FP of target point cloud P'; (b) eigenvector set FQ of source point cloud Q'
    Registration results of Bunny. (a) Original point cloud; (b) initial registration by algorithm in Ref. [26]; (c) precise registration by algorithm in Ref. [26]; (d) initial registration by proposed algorithm; (e) precise registration by proposed algorithm; (f) registration by classic ICP algorithm
    Fig. 2. Registration results of Bunny. (a) Original point cloud; (b) initial registration by algorithm in Ref. [26]; (c) precise registration by algorithm in Ref. [26]; (d) initial registration by proposed algorithm; (e) precise registration by proposed algorithm; (f) registration by classic ICP algorithm
    Registration results of Dragon. (a) Original point cloud; (b) initial registration by algorithm in Ref. [26]; (c) precise registration by algorithm in Ref. [26]; (d) initial registration by the proposed algorithm; (e) precise registration by proposed algorithm; (f) registration by classic ICP algorithm
    Fig. 3. Registration results of Dragon. (a) Original point cloud; (b) initial registration by algorithm in Ref. [26]; (c) precise registration by algorithm in Ref. [26]; (d) initial registration by the proposed algorithm; (e) precise registration by proposed algorithm; (f) registration by classic ICP algorithm
    Multi-view 3D point cloud data of part model
    Fig. 4. Multi-view 3D point cloud data of part model
    Registration results of part. (a) Original point cloud; (b) initial registration by algorithm in Ref. [26]; (c) precise registration by algorithm in Ref. [26]; (d) initial registration by proposed algorithm; (e) precise registration by proposed algorithm; (f) registration by classic ICP algorithm
    Fig. 5. Registration results of part. (a) Original point cloud; (b) initial registration by algorithm in Ref. [26]; (c) precise registration by algorithm in Ref. [26]; (d) initial registration by proposed algorithm; (e) precise registration by proposed algorithm; (f) registration by classic ICP algorithm
    Registration result of milk & cleanser. (a) Original point cloud; (b) initial registration by proposed algorithm; (c) precise registration by proposed algorithm
    Fig. 6. Registration result of milk & cleanser. (a) Original point cloud; (b) initial registration by proposed algorithm; (c) precise registration by proposed algorithm
    Algorithmt /sMSE /mm
    BunnyDragonBunnyDragon
    Classic ICP algorithm58.6573.220.273.83
    Algorithm in Ref. [26]41.1649.600.0820.15
    Proposed algorithm16.7113.350.0290.087
    Table 1. Point cloud registration data comparison
    AlgorithmPoint number of p, qt /sMSE /mm
    Classic ICP algorithm102.392.593
    Algorithm in Ref. [26]74537 7858172.570.452
    Proposed algorithm23.740.148
    Table 2. Point cloud registration data comparison
    Min Zhao, Qin Shu, Wei Chen, Yunxiu Yang. Three-Dimensional Point Cloud Registration Algorithm Based on l p Spatial Mechanics Model [J]. Acta Optica Sinica, 2018, 38(10): 1010005
    Download Citation