• Laser & Optoelectronics Progress
  • Vol. 58, Issue 14, 1428001 (2021)
Chunhui Wang1、2、*, Aoyou Wang1、2, Wei Rong1、2, Yuliang Tao1、2, and Ruimin Fu1
Author Affiliations
  • 1Beijing Institute of Space Mechanics and Electricity, Beijing 100094, China
  • 2Key Laboratory for Space Laser Information Perception Technology, China Academy of Space Technology, Beijing 100094, China
  • show less
    DOI: 10.3788/LOP202158.1428001 Cite this Article Set citation alerts
    Chunhui Wang, Aoyou Wang, Wei Rong, Yuliang Tao, Ruimin Fu. Adaptive Denoising Algorithm for Photon-Counting LiDAR Point Clouds[J]. Laser & Optoelectronics Progress, 2021, 58(14): 1428001 Copy Citation Text show less
    Airport building photos and point cloud data. (a) Airport building; (b) point cloud data
    Fig. 1. Airport building photos and point cloud data. (a) Airport building; (b) point cloud data
    Change process of the search area shape
    Fig. 2. Change process of the search area shape
    Flow of the denoising algorithm
    Fig. 3. Flow of the denoising algorithm
    Statistical histogram of the neighborhood density
    Fig. 4. Statistical histogram of the neighborhood density
    Neighborhood density of noise points after fitting
    Fig. 5. Neighborhood density of noise points after fitting
    Denoising result of the point cloud. (a) Noise and signal points after processing; (b) partial enlarged view
    Fig. 6. Denoising result of the point cloud. (a) Noise and signal points after processing; (b) partial enlarged view
    Fitting result of the airport building contour
    Fig. 7. Fitting result of the airport building contour
    Processing results of the MABEL point cloud data. (a) No.6; (b) No.8; (c) No.3; (d) No.9
    Fig. 8. Processing results of the MABEL point cloud data. (a) No.6; (b) No.8; (c) No.3; (d) No.9
    Serial numberNumber of signal pointsStandard deviation /m
    1900.21
    21040.23
    36320.13
    4700.15
    5810.27
    61070.24
    7990.18
    8860.16
    9750.22
    Sum13440.18
    Table 1. Fitting error of the airport building contour
    θ/(°)h/ml/mμσFPFNTPR/%P/%
    30.224.26.52.311452292998.2696.25
    50.283.26.52.315218295799.3995.11
    100.402.36.52.31794298199.8794.34
    Table 2. Effect of θ on algorithm recognition rate and accuracy
    θ/(°)h/ml/mμσFPFNTPR/%P/%
    50.283.26.52.315218295799.3995.11
    50.354.09.42.916827294499.0994.60
    50.424.813.03.530741292198.6290.49
    Table 3. Effect of search area on algorithm recognition rate and accuracy
    Serial numberFlight timeScenarioNumber of points
    120120412T1644northwest Greenland25001
    220120412T1659central Greenland50000
    320120420T0954sea ice around Greenland40000
    420120420T1004edge of Greenland50001
    520120915T2300water15113
    620130919T1512vegetation, day30108
    720130920T2225vegetation, night33759
    820130927T1856land56201
    920140729T2106north pole40690
    Table 4. MABEL point cloud data
    Serial numberθ/(°)h/ml/mμσFPFNTPR/%P/%
    152.630.05.42.026733394899.1793.67
    252.630.06.62.31744180499.7891.20
    352.630.06.32.21382131699.8590.51
    452.630.06.62.31323130099.7790.78
    5513.1150.04.31.84351742699.9994.47
    653.945.04.71.960221564996.2498.95
    757.990.05.32.0142591914899.6999.26
    853.135.05.62.2352230374894.2291.41
    953.135.06.32.320130184298.4090.16
    Table 5. Processing parameters and results of MABEL point cloud data
    Chunhui Wang, Aoyou Wang, Wei Rong, Yuliang Tao, Ruimin Fu. Adaptive Denoising Algorithm for Photon-Counting LiDAR Point Clouds[J]. Laser & Optoelectronics Progress, 2021, 58(14): 1428001
    Download Citation