• Laser & Optoelectronics Progress
  • Vol. 57, Issue 6, 061502 (2020)
Xixi Zhang1, Xiaogang Ji1、2、*, Haitao Hu1, Yuhao Luan1, and Jian'an Zhang1
Author Affiliations
  • 1School of Mechanical Engineering, Jiangnan University, Wuxi, Jiangsu 214122, China
  • 2Jiangsu Key Laboratory of Advanced Food Manufacturing Equipment and Technology, Jiangnan University, Wuxi, Jiangsu 214122, China
  • show less
    DOI: 10.3788/LOP57.061502 Cite this Article Set citation alerts
    Xixi Zhang, Xiaogang Ji, Haitao Hu, Yuhao Luan, Jian'an Zhang. Point Cloud Segmentation Method for Complex Micro-Surface Based on Feature Line Fitting[J]. Laser & Optoelectronics Progress, 2020, 57(6): 061502 Copy Citation Text show less
    Feature points of part with complex surface. (a) Front view of feature points; (b) isometric side view of feature points; (c) processed feature points
    Fig. 1. Feature points of part with complex surface. (a) Front view of feature points; (b) isometric side view of feature points; (c) processed feature points
    Element boundary points
    Fig. 2. Element boundary points
    Process of feature point reduction. (a) Elements after separation; (b) boundary points of elements; (c) lower boundary points of elements
    Fig. 3. Process of feature point reduction. (a) Elements after separation; (b) boundary points of elements; (c) lower boundary points of elements
    Model of banded feature points
    Fig. 4. Model of banded feature points
    Schematic of piecewise least square
    Fig. 5. Schematic of piecewise least square
    Separation process of lower boundary points. (a) Lower boundary points containing partial high points; (b) complete lower boundary points; (c) lower boundary points after filtering
    Fig. 6. Separation process of lower boundary points. (a) Lower boundary points containing partial high points; (b) complete lower boundary points; (c) lower boundary points after filtering
    Trend matching of elements. (a) Eight cases of element trend matching; (b) diagram of element recognition
    Fig. 7. Trend matching of elements. (a) Eight cases of element trend matching; (b) diagram of element recognition
    Segmentation process for point cloud data. (a) Interior points of surface identified by regional growth; (b) interior and exterior points of enlarged local surface; (c) segmented surface points by combining two methods
    Fig. 8. Segmentation process for point cloud data. (a) Interior points of surface identified by regional growth; (b) interior and exterior points of enlarged local surface; (c) segmented surface points by combining two methods
    Discrimination process for points near boundary. (a) Magnified local triangular mesh; (b) vector discrimination principle diagram; (c) diagram of distinguishing inner and outer points
    Fig. 9. Discrimination process for points near boundary. (a) Magnified local triangular mesh; (b) vector discrimination principle diagram; (c) diagram of distinguishing inner and outer points
    Reconstructed surface and comparison of surface accuracy. (a) Reconstructed surface; (b) comparison of surface accuracy
    Fig. 10. Reconstructed surface and comparison of surface accuracy. (a) Reconstructed surface; (b) comparison of surface accuracy
    Lower boundary extraction process of belt buckle. (a) Feature points of belt buckle part; (b) boundary points of feature points; (c) lower boundary points of feature points
    Fig. 11. Lower boundary extraction process of belt buckle. (a) Feature points of belt buckle part; (b) boundary points of feature points; (c) lower boundary points of feature points
    Comparison of segmentation effect of point cloud data. (a) Point cloud segmented by algorithm in Ref. [4]; (b) point cloud segmented by Geomagic Designx; (c) point cloud segmented by algorithm in this paper
    Fig. 12. Comparison of segmentation effect of point cloud data. (a) Point cloud segmented by algorithm in Ref. [4]; (b) point cloud segmented by Geomagic Designx; (c) point cloud segmented by algorithm in this paper
    Reconstructed surface and comparison of surface accuracy. (a) Reconstructed surface; (b) comparison of surface accuracy
    Fig. 13. Reconstructed surface and comparison of surface accuracy. (a) Reconstructed surface; (b) comparison of surface accuracy
    Parameter1234567891011121314
    x6.826.957.087.227.227.357.487.607.727.847.958.068.178.27
    y-6.77-6.62-6.48-6.30-6.33-6.18-6.02-5.87-5.71-5.55-5.38-5.22-5.05-4.88
    y20.440.440.440.440.440.440.440.440.440.440.440.440.440.44
    y30.230.270.310.350.350.390.430.460.500.530.570.600.630.66
    Table 1. Experimental data for judging concavity and convexity
    SegmentationmethodNumber of surfacesafter segmentationNumber of over-segmented surfacesNumber of insufficientsegmentation surfaces
    Method in Ref.[4]523
    Geomagic Designx1760
    Proposed method700
    Table 2. Comparison of number of surfaces segmented by different methods
    Xixi Zhang, Xiaogang Ji, Haitao Hu, Yuhao Luan, Jian'an Zhang. Point Cloud Segmentation Method for Complex Micro-Surface Based on Feature Line Fitting[J]. Laser & Optoelectronics Progress, 2020, 57(6): 061502
    Download Citation