• Laser & Optoelectronics Progress
  • Vol. 58, Issue 16, 1610016 (2021)
Shuai Li1、2 and Yuhong Du1、2、*
Author Affiliations
  • 1College of Mechanical Engineer, Tiangong University, Tianjin 300387, China
  • 2Tianjin Key Laboratory of Modern Mechanical and Electrical Equipment Technology, Tianjin 300387, China
  • show less
    DOI: 10.3788/LOP202158.1610016 Cite this Article Set citation alerts
    Shuai Li, Yuhong Du. Boundary Extraction of Scattered Point Cloud with Normal Estimation Based on Improved 3RDP Algorithm[J]. Laser & Optoelectronics Progress, 2021, 58(16): 1610016 Copy Citation Text show less
    Original point cloud “xbox”
    Fig. 1. Original point cloud “xbox”
    Effect diagrams for different values of K and R
    Fig. 2. Effect diagrams for different values of K and R
    Point set sorting process
    Fig. 3. Point set sorting process
    Three-dimensional Douglas-Puck algorithm diagram
    Fig. 4. Three-dimensional Douglas-Puck algorithm diagram
    Comparison of rejection rates of 10 targets
    Fig. 5. Comparison of rejection rates of 10 targets
    Contour extraction effect of Meleon40 data
    Fig. 6. Contour extraction effect of Meleon40 data
    Preprocessing of original point cloud. (a) Real environmental objects; (b) original point cloud collected; (c) point cloud after preprocessing
    Fig. 7. Preprocessing of original point cloud. (a) Real environmental objects; (b) original point cloud collected; (c) point cloud after preprocessing
    Contour extraction effect diagram
    Fig. 8. Contour extraction effect diagram
    Algorithm comparison analysis. (a) Original point cloud of Sink; (b) extraction effect of Sink using Gaussian clustering algorithm; (c) extraction effect of Sink using improved 3RDP algorithm; (d) original point cloud of Table; (e) extraction effect of Table using Gaussian clustering algorithm; (f) extraction effect of Table using improved 3RDP algorithm
    Fig. 9. Algorithm comparison analysis. (a) Original point cloud of Sink; (b) extraction effect of Sink using Gaussian clustering algorithm; (c) extraction effect of Sink using improved 3RDP algorithm; (d) original point cloud of Table; (e) extraction effect of Table using Gaussian clustering algorithm; (f) extraction effect of Table using improved 3RDP algorithm
    ObjectOriginal point cloudNormal vector algorithmImproved 3RDP algorithm
    Remaining point cloudRejection rate /%Remaining point cloudRejection rate /%
    Monitor10000536246.38229477.06
    Wardrobe10000462553.75163283.68
    Sink10000431256.88133186.69
    Table10000463253.68309469.06
    Dresser10000593140.69226377.37
    Curtain10000215578.45206479.36
    Cup10000235476.46103989.61
    Cone10000186581.3598890.12
    Stairs10000455154.49282771.73
    Stool10000305669.44154084.60
    Table 1. Comparison ofnormal line estimation algorithm and improved RDP algorithm
    Serial numberObjectsOriginal point cloudEdge pointRejection rate /%
    1Paper towel8607267868.89
    2Medicine bottle8421246170.78
    3Toner bottle 110814320870.33
    4Pill box14386357075.18
    5Yogurt9204286168.92
    6Toner bottle 28965205877.04
    7Staple box19811680365.66
    8Pen case14074202785.60
    9Ink cartridge13132308376.52
    10Soda water14197203085.70
    11Ink tank10643298571.95
    Table 2. Effect of point cloud extraction of 11 targets
    AlgorithmSinkTableRejection rate /%
    Gaussian clustering2710874642.72
    Improved 3RDP1331309477.87
    Table 3. Comparison of improved 3RDP algorithm and Gaussian clustering algorithm
    Shuai Li, Yuhong Du. Boundary Extraction of Scattered Point Cloud with Normal Estimation Based on Improved 3RDP Algorithm[J]. Laser & Optoelectronics Progress, 2021, 58(16): 1610016
    Download Citation