Journal of Infrared and Millimeter Waves, Volume. 43, Issue 2, 242(2024)

Real time denoising method for spaceborne photon counting laser ranging radar

Chong-Tao TAN1, Wen-Bo YU1, Yu-Yan XIANG1, Shao-Hui LI2, Jing YU2, Qian-Ying WANG2, and Song LI1,3、*
Author Affiliations
  • 1School of Electronic Information,Wuhan University,Wuhan 430072,China
  • 2China Academy of Space Technology,Beijing 100098,China
  • 3Wuhan Institute of Quantum Technology,Wuhan 430010,China
  • show less
    Figures & Tables(17)
    Schematic diagram of the detection process of spaceborne photon counting LiDAR
    Schematic diagram of denoising algorithm flow
    Theoretical relationship between ATLAS noise rate and solar zenith angle and surface reflectance[24]
    β distribution curve:(a) the probability density function curve of β distribution;(b) the cumulative distribution function curve of β distribution
    Process flow diagram of Monte Carlo Simulation
    Land simulation results:(a) distribution map of the original signal and noise photon event time data;(b) distribution curve of the photon event time data after denoising,the diagram is illustrated in the box,the horizontal line represents the mean of the photon event time data after denoising,the green circle represents the signal,and the red circle represents the noise
    Ocean simulation results:(a) distribution map of the original signal and noise photon event time data;(b) distribution curve of the photon event time data after denoising,the diagram is illustrated in the box,the horizontal line represents the mean of the photon event time data after denoising,the green circle represents the signal,and the red circle represents the noise
    Land ice simulation results:(a) distribution map of the original signal and noise photon event time data;(b) distribution curve of the photon event time data after denoising,the diagram is illustrated in the box,the horizontal line represents the mean of the photon event time data after denoising,the green circle represents the signal,and the red circle represents the noise
    Ocean ice simulation results:(a) distribution map of the original signal and noise photon event time data;(b) distribution curve of the photon event time data after denoising,the diagram is illustrated in the box,the horizontal line represents the mean of the photon event time data after denoising,the green circle represents the signal,and the red circle represents the noise
    ICESat-2 data denoising results,(a),(b),(c) and (d) are strong beam data; (e),(f),(g) and (h) are weak beam data; (a),(b),(e) and (f) are sea ice data; (c),(d),(g) and (h) are ocean data; the noise rate of (a),(c),(e) and (g) is 1 MHz; the noise rate of (b),(d),(f) and (h) is 5 MHz,blue represents noise photons,green represents signal photons,and red represents denoised signal photons
    The denoising results of ICESat-2 data grid method,(a),(b),(c) and (d) are strong beam data; (e),(f),(g) and (h) are weak beam data; (a),(b),(e) and (f) are sea ice data; (c),(d),(g) and (h) are ocean data; the noise rate of (a),(c),(e) and (g) is 1 MHz; the noise rate of (b),(d),(f) and (h) is 5 MHz,blue represents noise photons,green represents signal photons,and red represents denoised signal photons
    • Table 1. Coarse and fine denoising algorithm

      View table
      View in Article

      Table 1. Coarse and fine denoising algorithm

      粗精去噪算法

      输入:采样到的待去噪的光子事件时间数据的集合ti

      输出:去噪后的光子事件时间数据集合tsf和探测目标的距离值Z

      粗去噪

      1: 确定单次去噪对应的发射脉冲数量NN发脉冲接收到的光子事件数量Nc

      2: ti=Sortti, i=1,2,,Nct(i)ti排序后的数据

      3: 在ti中每次连续选取n个数据tni进行计算

      tmax=max tn(i)

      tmin=min tn(i)

      t=tmax-tmin

      tsc={tnidi=tn-1<Tp,Tp=6σstsc为可能的信号光子事件时间数据集合, di为光子事件时间数据密度,Tp为发射激光脉冲脉宽

      4: 采用重叠选取的方式选择下一次的n个数据,直至遍历完ti中的所有数据,连续两次选取数据重叠数量为n-1。为保证查全率,判定为噪声光子事件的数据可在其他样本中改判为信号光子事件,反之,判定为信号光子事件的数据不可改判为噪声光子事件。

      精去噪

      5: tmax'=histogramtsctmax'为直方图统计峰值对应的时间值

      6: tsf=tsc|(tmax'-Tp)tsc(tmax'+Tp)tsf为最终去噪后保留下来的信号光子事件时间数据

      7: tmean=meantsftmean为筛选出的信号光子时间数据的平均值。

      8: Z=c2tmeanc为光速,Z为距离值。

    • Table 2. Experimental results of coarse and fine denoising algorithm

      View table
      View in Article

      Table 2. Experimental results of coarse and fine denoising algorithm

      序号目标TPFPTNFN查全率R查准率P调和平均值F数据压缩比
      1陆地300.2299.8010.993 377 50.996 677 710.927 152 32
      2海洋300.9499.1010.970 873 80.985 221 717.152 103 56
      3陆地冰300.4799.6010.986 842 10.993 377 527.302 631 58
      4海洋冰300.6999.4010.980 392 20.990 09933.660 130 72
    • Table 3. Experimental results of common histogram statistical denoising algorithms

      View table
      View in Article

      Table 3. Experimental results of common histogram statistical denoising algorithms

      序号目标TPFPTNFN查全率R查准率P调和平均值F数据压缩比
      1陆地300.2299.8010.993 377 50.996 677 710.927 152 32
      2海洋300.9499.1010.970 873 80.985 221 717.152 103 56
      3陆地冰300.5799.5010.983 606 60.991 735 627.213 114 75
      4海洋冰300.6999.4010.980 392 20.990 09933.660 130 72
    • Table 4. ICESat-2 data denoising results

      View table
      View in Article

      Table 4. ICESat-2 data denoising results

      目标噪声率(MHz)平均信号光子数(个)N查全率(%)查准率(%)调和平均值(%)
      海洋51.24597.6593.5695.56
      海洋58.81599.1699.3699.26
      海洋11.245099.9899.1199.54
      海洋18.815098.8699.8899.36
      海冰11.15096.4297.7397.07
      海冰14.425094.3899.5596.9
      海冰51.1594.8494.4494.64
      海冰54.42595.8297.9296.86
    • Table 5. Denoising results of ICESat-2 data grid method

      View table
      View in Article

      Table 5. Denoising results of ICESat-2 data grid method

      目标噪声率(MHz)平均信号光子数(个)查全率(%)查准率(%)调和平均值(%)
      海洋51.2498.7279.5688.11
      海洋58.8192.9090.0591.45
      海洋11.2498.7294.3196.47
      海洋18.8192.9092.1992.55
      海冰11.199.6296.0097.78
      海冰14.4299.2198.2698.73
      海冰51.199.4284.9791.63
      海冰54.4298.9794.6096.74
    • Table 6. Comparison of denoising calculation time

      View table
      View in Article

      Table 6. Comparison of denoising calculation time

      目标噪声率(MHz)平均信号光子数(个)本文算法计算时间(s)栅格法计算时间(s)
      海洋51.240.642 20.716 0
      海洋58.810.754 10.871 8
      海洋11.240.320 00.346 7
      海洋18.810.438 30.473 3
      海冰11.10.325 80.354 7
      海冰14.420.342 00.378 0
      海冰51.10.634 10.738 6
      海冰54.420.643 20.798 9
    Tools

    Get Citation

    Copy Citation Text

    Chong-Tao TAN, Wen-Bo YU, Yu-Yan XIANG, Shao-Hui LI, Jing YU, Qian-Ying WANG, Song LI. Real time denoising method for spaceborne photon counting laser ranging radar[J]. Journal of Infrared and Millimeter Waves, 2024, 43(2): 242

    Download Citation

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

    Category: Research Articles

    Received: Jun. 28, 2023

    Accepted: --

    Published Online: Apr. 29, 2024

    The Author Email: Song LI (ls@whu.edu.cn)

    DOI:10.11972/j.issn.1001-9014.2024.02.014

    Topics