Opto-Electronic Engineering, Volume. 50, Issue 6, 220341(2023)

A denoising algorithm based on neural network for side-scatter lidar signal

Yuzhao Ma1,*... Yanfeng Zhang1 and Shuai Feng2 |Show fewer author(s)
Author Affiliations
  • 1Tianjin Key Laboratory for Advanced Signal Processing, Civil Aviation University of China, Tianjin 300300, China
  • 2Engineering Techniques Training Center, Civil Aviation University of China, Tianjin 300300, China
  • show less
    Figures & Tables(15)
    Diagram of the side-scatter lidar
    The side-scatter lidar signal
    The structure of the proposed model
    Residual network structure
    CBAM structure
    Comparison of ReLU and SELU
    Simulation images. (a) Original image; (b) Noised image corresponding to noise model 1; (c) Noised image corresponding to noise model 2; (d) Noised image corresponding to noise model 3
    De-noising images. (a) Wavelet threshold (Soft); (b) Wavelet threshold (Hard); (c) Wiener filtering; (d) VGG16; (e) DnCNN; (f) DnCNN+
    Change of the Neural network loss
    Difference distribution of the signal photon number
    Distribution of the relative error
    • Table 1. Comparison of the PSNR images

      View table
      View in Article

      Table 1. Comparison of the PSNR images

      Denoising methodPSNR/dB
      NoneWavelet (Soft)19.9924.73
      Wavelet (Hard)24.88
      Wiener filteringVGG16DnCNNDnCNN+23.5425.1925.2625.87
    • Table 2. Comparison of the SSIM images

      View table
      View in Article

      Table 2. Comparison of the SSIM images

      Denoising methodSSIM
      NoneWavelet (Soft)0.050.22
      Wavelet (Hard)0.27
      Wiener filteringVGG16DnCNNDnCNN+0.160.240.270.29
    • Table 3. Comparison of the PSNR imageses at different noise intensities

      View table
      View in Article

      Table 3. Comparison of the PSNR imageses at different noise intensities

      Noise intensity (variance)Denoising method0.01PSNR/dB0.02PSNR/dB0.03PSNR/dB
      NoneWavelet (Soft)22.9827.7919.9624.5818.2719.78
      Wavelet (Hard)27.5224.8223.27
      Wiener filteringVGG16DnCNNDnCNN+26.6927.8428.2028.6223.5825.1725.2325.7521.7923.0923.2623.64
    • Table 4. Comparison of average deviation of the signal photon number

      View table
      View in Article

      Table 4. Comparison of average deviation of the signal photon number

      Denoising methodS
      Wavelet (Soft)0.35
      Wavelet (Hard)0.21
      Wiener filteringDnCNN+0.460.18
    Tools

    Get Citation

    Copy Citation Text

    Yuzhao Ma, Yanfeng Zhang, Shuai Feng. A denoising algorithm based on neural network for side-scatter lidar signal[J]. Opto-Electronic Engineering, 2023, 50(6): 220341

    Download Citation

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

    Category: Article

    Received: Dec. 15, 2022

    Accepted: Mar. 24, 2023

    Published Online: Aug. 9, 2023

    The Author Email:

    DOI:10.12086/oee.2023.220341

    Topics