• Laser & Optoelectronics Progress
  • Vol. 57, Issue 22, 221017 (2020)
Doudou Xue1、*, Yinglei Cheng1, Xiaosong Shi1, Xianxiang Qin1, and Pei Wen1、2
Author Affiliations
  • 1Information and Navigation College, Air Force Engineering University, Xi'an, Shaanxi 710077, China
  • 2The 93575 Unit, Chengde, Hebei 067000, China
  • show less
    DOI: 10.3788/LOP57.221017 Cite this Article Set citation alerts
    Doudou Xue, Yinglei Cheng, Xiaosong Shi, Xianxiang Qin, Pei Wen. Point Clouds Classification Algorithm Based on Cloth Filtering Algorithm and Improved Random Forest[J]. Laser & Optoelectronics Progress, 2020, 57(22): 221017 Copy Citation Text show less
    Flow chart of ICSF filtering algorithm
    Fig. 1. Flow chart of ICSF filtering algorithm
    Feature grid diagram
    Fig. 2. Feature grid diagram
    Weighted weakly correlated random forest model
    Fig. 3. Weighted weakly correlated random forest model
    Point clouds elevation distribution. (a) Unnormalized point cloud data; (b) normalized point cloud data
    Fig. 4. Point clouds elevation distribution. (a) Unnormalized point cloud data; (b) normalized point cloud data
    Classification results of ground points. (a) Traditional CSF algorithm; (b) improved CSF algorithm
    Fig. 5. Classification results of ground points. (a) Traditional CSF algorithm; (b) improved CSF algorithm
    Classification results of different algorithms. (a) Correct classification results; (b) classification results of traditional RF algorithm; (c) classification results of SVM algorithm; (d) classification results of proposed algorithm
    Fig. 6. Classification results of different algorithms. (a) Correct classification results; (b) classification results of traditional RF algorithm; (c) classification results of SVM algorithm; (d) classification results of proposed algorithm
    ItemImproved RF
    GroundBuildingVegetationArtificialOAK
    Ground0.940.000.000.06
    Building0.000.890.110.000.880.85
    Vegetation0.000.110.860.03
    Artificial0.070.030.090.81
    ItemICSF+Improved RF
    GroundBuildingVegetationArtificialOAK
    Ground0.960.000.000.04
    Building0.000.910.090.000.910.90
    Vegetation0.000.080.90.02
    Artificial0.000.030.100.87
    Table 1. Comparison of experimental indicators of the two models
    ModelCAOAKTime /s
    GroundBuildingVegetationArtificial
    Traditional RF0.910.860.830.810.860.8227.75
    SVM0.940.890.880.820.890.8728.20
    ICSF + Improved RF0.960.900.910.870.910.9023.56
    Table 2. Classification accuracy of different feature combinations
    Doudou Xue, Yinglei Cheng, Xiaosong Shi, Xianxiang Qin, Pei Wen. Point Clouds Classification Algorithm Based on Cloth Filtering Algorithm and Improved Random Forest[J]. Laser & Optoelectronics Progress, 2020, 57(22): 221017
    Download Citation