• Infrared and Laser Engineering
  • Vol. 50, Issue 9, 20200431 (2021)
Qi Lu1, Tingting Lin1, Pengcheng Li2, Ronghua Li1、*, and Yanjun Ge1
Author Affiliations
  • 1Institute of Mechanical Engineering, Dalian Jiaotong University, Dalian 116028, China
  • 2SIASUN Robot & Automation CO., Ltd,Shenyang 110000, China
  • show less
    DOI: 10.3788/IRLA20200431 Cite this Article
    Qi Lu, Tingting Lin, Pengcheng Li, Ronghua Li, Yanjun Ge. Improved registration algorithm for spatial non-cooperative target point cloud clustering[J]. Infrared and Laser Engineering, 2021, 50(9): 20200431 Copy Citation Text show less
    Positional relationship of measurable points in radar coordinate system
    Fig. 1. Positional relationship of measurable points in radar coordinate system
    Spatial non-cooperative target simulation visual point cloud
    Fig. 2. Spatial non-cooperative target simulation visual point cloud
    Regional growth clustering results
    Fig. 3. Regional growth clustering results
    Color visual point cloud
    Fig. 4. Color visual point cloud
    Boundary extraction result
    Fig. 5. Boundary extraction result
    Clustering results of the proposed algorithm
    Fig. 6. Clustering results of the proposed algorithm
    Small-scale distinctive feature point cloud
    Fig. 7. Small-scale distinctive feature point cloud
    Pitch angle registration error
    Fig. 8. Pitch angle registration error
    Yaw angle registration error
    Fig. 9. Yaw angle registration error
    Roll angle registration error
    Fig. 10. Roll angle registration error
    Algorithm consumes time
    Fig. 11. Algorithm consumes time
    GroupsThree axis translation distance
    dx/mm dy/mm dz/mm
    21-20−0.4724−0.52631.0306
    22-21−1.22400.63750.4955
    23-22−0.1181−0.30580.7028
    ............
    48-470.5716−1.08550.3489
    49-48−0.97712.6975−0.3801
    50-49−0.19050.9025−0.4344
    GroupsThree axis rotation angle
    ax/(°) ay/(°) az/(°)
    21-20−0.00674.4969−0.3413
    22-21−0.25734.59630.8183
    23-22−0.26534.90131.0290
    ............
    48-47−0.01734.6220−0.2492
    49-480.21484.8584−0.0877
    50-49−0.12504.9179−0.0648
    Table 1. Registration results of proposed algorithm
    GroupsThree axis translation distance
    dx/mm dy/mm dz/mm
    21-200.3604−0.34640.4818
    22-210.1552−0.69670.5961
    23-220.1630−0.2454−0.1001
    ............
    48-470.0407−0.1771−0.4323
    49-480.0297−0.12390.2898
    50-490.21010.17760.1921
    GroupsThree axis rotation angle
    ax/(°) ay/(°) az/(°)
    21-20−0.20884.82900.6914
    22-21−0.07874.78671.3158
    23-220.15384.96921.0110
    ............
    48-470.32254.85530.2024
    49-480.35854.85230.1457
    50-490.45464.78520.2026
    Table 2. Registration results of ICP algorithm
    GroupsProposed algorithm consumes time/s ICP algorithm consumes time/s
    21-2012.929.5
    22-2115.933.3
    23-2217.437.5
    .........
    48-4713.149.8
    49-4813.044.1
    50-4912.040.7
    Table 3. Algorithm time-consuming comparison
    GroupsThree axis translation distance
    dx/mm dy/mm dz/mm
    Proposed algorithm0.71012.42220.3235
    ICP algorithm2.0442−0.71002.7629
    GroupsThree axis rotation angle
    ax/(°) ay/(°) az/(°)
    Proposed algorithm0.262410.45493.2799
    ICP algorithm2.24424.42457.8200
    Table 4. Registration cumulative error statistics
    Qi Lu, Tingting Lin, Pengcheng Li, Ronghua Li, Yanjun Ge. Improved registration algorithm for spatial non-cooperative target point cloud clustering[J]. Infrared and Laser Engineering, 2021, 50(9): 20200431
    Download Citation