Chinese Optics, Volume. 17, Issue 5, 1098(2024)

Polarization spectral image fusion method for hybrid backgrounds of ground objects

Ying-chao LI1, Zhe-hao ZHAO2、*, Qi WANG1, Jia-nan LIU2, Hao-dong SHI1,2, Qiang FU1,2, and Hong-yu SUN1
Author Affiliations
  • 1Jilin Provincial Key Laboratory of Space Optoelectronics Technology, Changchun University of Science and Technology, Changchun 130022, China
  • 2School of Optoelectronic Engineering, Changchun University of Science and Technology, Changchun 130022, China
  • show less
    Figures & Tables(18)
    (a) Flowchart of NSCT decomposition and (b) ideal frequency domain segmentation for NSCT
    Flowchart of the fusion method discussed in this paper
    Comparison of low-frequency fusion results
    Comparison of high-frequency fusion results
    Final fusion result 最终融合结果
    Polarization spectral imager
    Objective evaluation curves of different parameters
    Comparison of the reference method and the proposed method
    Objective evaluation indexes of different fusion methods at four wavelengths
    Comparison of different fusion methods
    True color image comparison chart
    • Table 1. Objective evaluation metric results of low-frequency sub-band fusion

      View table
      View in Article

      Table 1. Objective evaluation metric results of low-frequency sub-band fusion

      评价指标融合规则
      区域方差匹配度取平均权重融合稀疏表示
      MI1.66851.45231.62381.9777
      Qw0.64080.63240.72410.7822
    • Table 2. Evaluation metric results of high-frequency contour information

      View table
      View in Article

      Table 2. Evaluation metric results of high-frequency contour information

      评价指标PCNN边缘探测绝对值取大引导滤波
      SSIM0.73540.68210.63780.7562
      Qab/f0.59160.44810.45070.5988
    • Table 3. Indicators of polarization spectral imager

      View table
      View in Article

      Table 3. Indicators of polarization spectral imager

      性能指标参数值
      波长范围/nm400~900
      光谱分辨率优于2 nm
      F/#3
      视场角/(°)±4.1
      分辨率2448×2048
      像元尺寸/μm3.45×3.45
      光学尺寸/inch2/3
      帧率36 fps@2448×2048
    • Table 4. The evaluation indexes of image fusion of each component for the reference method and the proposed method

      View table
      View in Article

      Table 4. The evaluation indexes of image fusion of each component for the reference method and the proposed method

      融合图像评价指标
      MISSIMQab/fQw
      参考方法I融合结果2.76230.71820.61470.8367
      本文I融合结果2.80310.73870.63930.8625
      参考方法DoLP融合结果1.97650.63760.57330.7615
      本文DoLP融合结果2.73420.70230.61420.8222
      参考方法AoP融合结果1.66810.56790.49350.6138
      本文AoP融合结果1.74370.67550.58830.7004
    • Table 5. The evaluation indexes of HSI pseudo-colored images for the reference method and the proposed method

      View table
      View in Article

      Table 5. The evaluation indexes of HSI pseudo-colored images for the reference method and the proposed method

      不同方法评价指标
      PSNRSSIM
      参考方法HSI融合结果22.81 dB0.6431
      本文方法HSI融合结果28.52 dB0.6947
    • Table 6. Evaluation results of different fusion methods

      View table
      View in Article

      Table 6. Evaluation results of different fusion methods

      融合方法评价指标
      MISSIMQab/fQw
      CNN2.23430.64340.56620.7787
      GFF1.76850.58620.51370.7507
      LP-SR2.11320.63610.53540.7427
      CVT-SR2.23560.60530.58330.7706
      NSCT-SR2.58350.59650.53500.7582
      本文方法2.73420.70230.61420.8222
    • Table 7. Evaluation indexes of original true color image and fused true color image

      View table
      View in Article

      Table 7. Evaluation indexes of original true color image and fused true color image

      真彩图像评价指标
      PSNRSSIM
      原始真彩图像33.72 dB0.8672
      融合后的真彩图像36.87 dB0.9104
    Tools

    Get Citation

    Copy Citation Text

    Ying-chao LI, Zhe-hao ZHAO, Qi WANG, Jia-nan LIU, Hao-dong SHI, Qiang FU, Hong-yu SUN. Polarization spectral image fusion method for hybrid backgrounds of ground objects[J]. Chinese Optics, 2024, 17(5): 1098

    Download Citation

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

    Category:

    Received: Oct. 8, 2023

    Accepted: --

    Published Online: Dec. 31, 2024

    The Author Email:

    DOI:10.37188/CO.2023-0185

    Topics