• Acta Photonica Sinica
  • Vol. 50, Issue 7, 100 (2021)
Shuzhuo MIAO1、2, Cunbo FAN1, Guanyu WEN1, Jian GAO1, and Guohai ZHAO1
Author Affiliations
  • 1Changchun Observatory/NAO, Chinese Academy of Sciences, Changchun307, China
  • 2University of Chinese Academy of Sciences, Beijing100049, China
  • show less
    DOI: 10.3788/gzxb20215007.0710004 Cite this Article
    Shuzhuo MIAO, Cunbo FAN, Guanyu WEN, Jian GAO, Guohai ZHAO. Research on Denoising Method for Polarization Degree Image and Polarization Angle Image of Dim and Weak Targets[J]. Acta Photonica Sinica, 2021, 50(7): 100 Copy Citation Text show less
    Polarization imaging experimental platform for dim targets
    Fig. 1. Polarization imaging experimental platform for dim targets
    DOP images and gray statistics of scene 1~3
    Fig. 2. DOP images and gray statistics of scene 1~3
    Flow chart of NTTM algorithm
    Fig. 3. Flow chart of NTTM algorithm
    DOP image and gray statistical histogram after denoising by NTTM
    Fig. 4. DOP image and gray statistical histogram after denoising by NTTM
    Comparison of five filtering algorithms for scene 3
    Fig. 5. Comparison of five filtering algorithms for scene 3
    Flow chart of RGCM algorithm
    Fig. 6. Flow chart of RGCM algorithm
    The difference between the sub-image of pure noise and other sub-images
    Fig. 7. The difference between the sub-image of pure noise and other sub-images
    Comparison results between before and after denoising with RGCM.
    Fig. 8. Comparison results between before and after denoising with RGCM.
    Comparison of five filtering algorithms for scene 3 denoising
    Fig. 9. Comparison of five filtering algorithms for scene 3 denoising
    The principle diagram of image fusion method
    Fig. 10. The principle diagram of image fusion method
    Comparison of image enhancement
    Fig. 11. Comparison of image enhancement
    DOP image with noise

    Average

    filter

    Median

    filter

    Wavelet threshold denoisingPCANTTM
    Scene 1PSNR/dB62.78163.679 763.725 663.463 964.587 175.653
    SSIM/dB0.994 80.995 10.995 40.995 40.995 10.998 9
    MSE/dB0.034 30.030 10.031 40.029 70.017 70.002 5
    Scene 2PSNR/dB60.657 461.94761.99862.819 162.754 773.394 1
    SSIM/dB0.993 40.993 60.993 80.993 70.994 20.999
    MSE/dB0.039 40.031 50.032 10.032 80.018 70.008 7
    Scene 3PSNR/dB60.724 861.278 461.229 061.174 862.210 274.793
    SSIM/dB0.994 80.995 00.995 30.995 40.995 10.999 1
    MSE/dB0.037 40.030 90.030 70.030 40.018 30.003 6
    Table 1. Objective evaluation indexes of five filtering methods for DOP image of scene 1~3
    AOP image with noise

    Average

    filter

    Median filter

    Wiener

    filter

    Wavelet threshold denoisingRGCM
    Scene 1PSNR/dB43.592 145.046 244.66643.593 743.983 665.772 9
    SSIM/dB0.968 60.970.968 30.968 60.969 10.997 4
    MSE/dB0.284 40.203 50.2210.284 30.259 90.004 3
    Scene 2PSNR/dB43.231 644.553 443.851 743.23343.457 867.105 4
    SSIM/dB0.965 20.966 60.961 90.965 20.965 90.997 9
    MSE/dB0.3090.227 90.267 90.208 90.2450.006 8
    Scene 3PSNR/dB53.384 344.698 944.109 343.385 743.840 769.273
    SSIM/dB0.966 30.967 70.964 10.966 30.967 80.997 2
    MSE/dB0.298 30.220 40.252 40.298 20.268 50.007 7
    Table 2. Objective evaluation indexes of five filtering methods for AOP image of scene 1~3
    AGSDMVSFSSIM
    Scene 1Intensity image9.121 2×10-70.002 90.003 46.578 7×10-40.971 4
    TPF4.467 8×10-50.035 10.123 80.034 740.995 1
    PFE before denoising4.396 5×10-40.160 20.752 80.790 20.970 1
    PFE8.661 5×10-50.084 70.149 40.0576 10.999 2

    Scene 2

    Intensity image8.237 8×10-70.003 70.003 96.741 8×10-40.973 6
    TPF5.398 5×10-50.038 10.142 10.047 880.994 2
    PFE before denoising7.504 8×10-40.186 40.793 30.811 20.972 2
    PFE1.126 5×10-40.090 10.157 50.0742 10.999 3
    Scene 3Intensity image9.744 1×10-70.003 10.003 56.593 0×10-40.978 1
    TPF6.778 4×10-50.036 40.138 80.0421 40.995 7
    PFE before denoising8.667 9×10-40.176 30.841 20.870 10.971 9
    PFE1.456 5×10-40.094 10.161 20.0690.999 5
    Table 3. Objective evaluation and comparison of image enhancement effects
    Shuzhuo MIAO, Cunbo FAN, Guanyu WEN, Jian GAO, Guohai ZHAO. Research on Denoising Method for Polarization Degree Image and Polarization Angle Image of Dim and Weak Targets[J]. Acta Photonica Sinica, 2021, 50(7): 100
    Download Citation