Acta Optica Sinica, Volume. 41, Issue 23, 2312003(2021)

Volume Measurement of Irregular Objects Based on Improved Point Cloud Slicing Method

Jinjin Liu** and Haojun Li*
Author Affiliations
  • College of Surveying and Geo-Informatics, Tongji University, Shanghai 200092, China
  • show less
    Figures & Tables(18)
    Main processes of volume measurement based on improved point cloud slicing method
    Point cloud data acquisition of spatial objects. (a) Multi-view point cloud data; (b) point cloud model
    Diagram of point cloud data slicing. (a) Side view of Stanford Bunny and some slicing positions on it; (b) contour boundary points in slicing layers
    Diagram of segmentation based on euclidean clustering (z=-7.52 mm). (a) Contour boundary points and initial polygon in slicing layers; (b) classification results; (c) generation of contour polygons after classification; (d) side length frequency distribution histogram of this polygon; (e) a partial enlarged drawing of Fig. (d)
    Schematic diagram of some points (like point 5) being missed during the two-way nearest point search
    Diagram of polygon splitting and recombination method. (a) Split into multiple polylines; (b)recombination into individual contour boundary polygons
    Analysis of inclusion relationships between multi-contour boundary polygons in cross-section. (a) Position information of boundary polygons; (b) tree structure; (c) chain table; (d) cross-section region (shaded part)
    Test data. (a) Three-dimensional surface model of Stanford Bunny; (b)(c) physical objects, point cloud data, and surface models of Happy Buddha, and Lucy
    Comparison and analysis of multi-contour boundary segmentation results of different slicing positions (along z-axis) of Stanford Bunny. (a) Sorting results of the two-way nearest points search; (b) results of the SEC method; (c) results of the PSR method
    Comparison and analysis of multi-contour boundary segmentation results of different slicing positions (alonge y-axis) of Happy Buddha. (a) Sorting results of the two-way nearest points search; (b) results of the SEC method; (c) results of the PSR method
    Comparison and analysis of multi-contour boundary segmentation results of different slicing positions (along z-axis) of Lucy. (a) Sorting results of the two-way nearest points search; (b) results of the SEC method; (c) results of the PSR method
    Comparison and analysis of cross-sectional area results of different slicing layers at Stanford Bunny. (a) Area of each boundary polygon; (b) accuracy of cross-sectional area results
    Comparison and analysis of the cross-sectional area results of different slicing layers at Happy Buddha and Lucy. (a)(c) Area of each boundary polygon; (b)(d) accuracy of cross-sectional area results
    Comparison and analysis of volume measurement results of three datasets with different slicing intervals. (a)(c)(e) Accuracy of volume results; (b)(d)(f) calculation time
    • Table 1. Attribute information of test data

      View table

      Table 1. Attribute information of test data

      Test dataPoint cloud dataClosed 3D curved surface model
      Number of pointsPoint cloud density ρ /mmNumber of trianglesPoint cloud volume V /mm3
      Stanford Bunny10407520.1329822081496753955
      Happy Buddha100107220.04388710005248354206
      Lucy140278720.2644391020457281597353
    • Table 2. Various parameter settings in data processing

      View table

      Table 2. Various parameter settings in data processing

      Test dataParameter bProjection thickness δ /mmParameter kParameter q
      Stanford Bunny0.40.0531928
      Happy Buddha0.80.035109221.5
      Lucy0.60.1586634
    • Table 3. Comparison and analysis of the cross-section area results of different slicing layers

      View table

      Table 3. Comparison and analysis of the cross-section area results of different slicing layers

      MethodSlicing position /mmReference area /mm2Cross-sectional area /mm2Absolute error /mm2Relative error /%
      -19.525829.76755822.2247.54350.1294
      -13.028048.36028028.92619.43420.2415
      -8.029824.75979332.889491.87075.0064
      SEC-7.5210035.598310032.7092.88930.0288
      -2.0211767.035811743.30023.73580.2017
      38.488213.18216009.9222203.260326.8259
      42.985569.05525318.557250.49824.4980
      -19.525829.76755822.2027.56550.1298
      -13.028048.36028028.94619.41420.2412
      -8.029824.75979818.8325.92770.0603
      PSR-7.5210035.598310032.7342.86430.0285
      -2.0211767.035811761.3945.64180.0479
      38.488213.18218206.1407.04190.0857
      42.985569.05525571.9562.90030.0521
    • Table 4. Comparison and analysis of volume measurement results of three datasets (slicing interval: 1 mm)

      View table

      Table 4. Comparison and analysis of volume measurement results of three datasets (slicing interval: 1 mm)

      MethodTest dataVolume V /mm3Absolute error /mm3Relative error /%Number of layersCalculation time /s
      Stanford Bunny64393111002414.59291202.583
      SECHappy Buddha34441997872.763119839.311
      Lucy7724490143524525.33411597366.510
      Stanford Bunny7532766790.09011202.229
      PSRHappy Buddha3539982080.058719833.732
      Lucy81573737236160.02891597327.476
    Tools

    Get Citation

    Copy Citation Text

    Jinjin Liu, Haojun Li. Volume Measurement of Irregular Objects Based on Improved Point Cloud Slicing Method[J]. Acta Optica Sinica, 2021, 41(23): 2312003

    Download Citation

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

    Category: Instrumentation, Measurement and Metrology

    Received: May. 14, 2021

    Accepted: Jun. 17, 2021

    Published Online: Dec. 10, 2021

    The Author Email: Liu Jinjin (jinjin_liu@tongji.edu.cn), Li Haojun (lhjch@tongji.edu.cn)

    DOI:10.3788/AOS202141.2312003

    Topics