• Infrared and Laser Engineering
  • Vol. 54, Issue 5, 20240523 (2025)
Ying ZHANG1, Xing WANG1, Siwei ZHANG2, Songran DOU2..., Xin LIU2, Songyu JIN2, Ruchuan LI3, Liyi LUO1 and Xihai XU1,*|Show fewer author(s)
Author Affiliations
  • 1Key Laboratory of Precision Opto-Mechtronics Technology, Ministry of Education, Beihang University, Beijing 100191, China
  • 2China Shipbuilding Information Center, Beijing 100101, China
  • 3Aerospace Information Research Institute of Qilu, Jinan 250101, China
  • show less
    DOI: 10.3788/IRLA20240523 Cite this Article
    Ying ZHANG, Xing WANG, Siwei ZHANG, Songran DOU, Xin LIU, Songyu JIN, Ruchuan LI, Liyi LUO, Xihai XU. Multi-spectral fusion dehazing method based on polarization spectral images(invited)[J]. Infrared and Laser Engineering, 2025, 54(5): 20240523 Copy Citation Text show less
    Block diagram of dehazing method based on visible-NIR polarimetric multi-spectral image fusion
    Fig. 1. Block diagram of dehazing method based on visible-NIR polarimetric multi-spectral image fusion
    Relationship between detected light intensity and direction of polarization[13]
    Fig. 2. Relationship between detected light intensity and direction of polarization[13]
    Measurement and the model simulation result of skylight DOP in 530 nm. (a) Measurement result; (b) Model simulation result; (c) Relative error of two results
    Fig. 3. Measurement and the model simulation result of skylight DOP in 530 nm. (a) Measurement result; (b) Model simulation result; (c) Relative error of two results
    Flowchart diagram for the dehazing method using visible/near-infrared polarized spectral image fusion
    Fig. 4. Flowchart diagram for the dehazing method using visible/near-infrared polarized spectral image fusion
    Local region comparison of visible/NIR polarization spectral images
    Fig. 5. Local region comparison of visible/NIR polarization spectral images
    Four different original polarimetric images and estimation of maximum- and minimum-intensity image in 532 nm
    Fig. 6. Four different original polarimetric images and estimation of maximum- and minimum-intensity image in 532 nm
    Visible/NIR fusion dehazed image
    Fig. 7. Visible/NIR fusion dehazed image
    Dehazing results of several methods in different scenarios
    Fig. 8. Dehazing results of several methods in different scenarios
    Dehazing methodSSIMIEICPSNR
    SCHUAL0.8015.3527.99417.661
    HE0.4315.16794.9788.427
    FFA-Net0.7115.34442.41711.7847
    FREDERIKE0.5795.19437.6327.828
    Proposed method0.9085.35446.51020.1831
    Table 1. Comparison of different dehazing methods
    Ying ZHANG, Xing WANG, Siwei ZHANG, Songran DOU, Xin LIU, Songyu JIN, Ruchuan LI, Liyi LUO, Xihai XU. Multi-spectral fusion dehazing method based on polarization spectral images(invited)[J]. Infrared and Laser Engineering, 2025, 54(5): 20240523
    Download Citation