• Laser & Optoelectronics Progress
  • Vol. 58, Issue 20, 2010001 (2021)
Haoyu Han**, Yuan Zhang*, and Xie Han
Author Affiliations
  • College of Big Data, North University of China, Taiyuan, Shanxi 030051, China
  • show less
    DOI: 10.3788/LOP202158.2010001 Cite this Article Set citation alerts
    Haoyu Han, Yuan Zhang, Xie Han. Improved Laser Point Cloud Filtering Algorithm[J]. Laser & Optoelectronics Progress, 2021, 58(20): 2010001 Copy Citation Text show less
    Results of statistical filtering algorithm before and after processing. (a) Initial scanning point cloud; (b) point cloud after processing
    Fig. 1. Results of statistical filtering algorithm before and after processing. (a) Initial scanning point cloud; (b) point cloud after processing
    Results of different tensor voting methods. (a) Normal tensor voting method; (b) point tensor voting method
    Fig. 2. Results of different tensor voting methods. (a) Normal tensor voting method; (b) point tensor voting method
    Estimated results at different K values. (a) K=15; (b) K=5
    Fig. 3. Estimated results at different K values. (a) K=15; (b) K=5
    Filtering results of noise plane
    Fig. 4. Filtering results of noise plane
    Point clouds of bird bottle model under different algorithms. (a) Initial point cloud; (b) proposed algorithm; (c) Ref. [4]; (d) Ref. [5]
    Fig. 5. Point clouds of bird bottle model under different algorithms. (a) Initial point cloud; (b) proposed algorithm; (c) Ref. [4]; (d) Ref. [5]
    Point clouds of hand model under different algorithms. (a) Initial point cloud; (b) proposed algorithm; (c) Ref. [4]; (d) Ref. [5]
    Fig. 6. Point clouds of hand model under different algorithms. (a) Initial point cloud; (b) proposed algorithm; (c) Ref. [4]; (d) Ref. [5]
    Point clouds of buddha model under different algorithms. (a) Initial point cloud; (b) proposed algorithm; (c) Ref. [4]; (d) Ref. [5]
    Fig. 7. Point clouds of buddha model under different algorithms. (a) Initial point cloud; (b) proposed algorithm; (c) Ref. [4]; (d) Ref. [5]
    Point clouds of stone model under different algorithms. (a) Initial point cloud; (b) proposed algorithm; (c) Ref. [4]; (d) Ref. [5]
    Fig. 8. Point clouds of stone model under different algorithms. (a) Initial point cloud; (b) proposed algorithm; (c) Ref. [4]; (d) Ref. [5]
    Point clouds of tortoise model under different algorithms.(a) Initial point cloud; (b) proposed algorithm; (c) Ref. [4]; (d) Ref. [5]
    Fig. 9. Point clouds of tortoise model under different algorithms.(a) Initial point cloud; (b) proposed algorithm; (c) Ref. [4]; (d) Ref. [5]
    ModelProposed algorithmRef. [4]Ref. [5]
    Bird bottle97.9085.0985.59
    Hand99.7697.5898.07
    Buddha98.9992.8096.01
    Stone99.8793.7295.87
    Tortoise99.2389.9093.31
    Table 1. Ratio of number of points between filter model and noiseless model unit: %
    ModelPointsRunning time of proposed algorithm/sRunning time of Ref. [4] /sRunning time of Ref. [5] /s
    Bird bottle3777831.316.629.3
    Hand5075638.818.329.6
    Buddha2932526.715.423.8
    Stone1540620.313.218.9
    Tortoise2293922.414.721.2
    Table 2. Points and processing time of different models
    Haoyu Han, Yuan Zhang, Xie Han. Improved Laser Point Cloud Filtering Algorithm[J]. Laser & Optoelectronics Progress, 2021, 58(20): 2010001
    Download Citation