Laser & Optoelectronics Progress, Volume. 56, Issue 14, 141007(2019)

Infrared and Visible Image Fusion via Rolling Guidance Filtering and Hybrid Multi-Scale Decomposition

Cheng Zhao1 and Yongdong Huang1,2、*
Author Affiliations
  • 1 Institute of Image Processing and Understanding, North Minzu University, Yinchuan, Ningxia 750021, China
  • 2 Center of Mathematics and Information Science, Dalian Minzu University, Dalian, Liaoning 116600, China
  • show less
    Figures & Tables(15)
    Basic framework of RGF
    Smooth results of RGF corresponding to different σr values
    Construction based on hybrid multi-scale decomposition
    Five sets of infrared and visible images used in test. (a) Un camp; (b) Octec; (c) Steamboat; (d) Kayak; (e) Infrared4-Visible4
    Fusion results obtained by different fusion methods applied in “Un camp” image
    Fusion results obtained by different fusion methods applied in “Octec” image
    Fusion results obtained by different fusion methods applied in “Steamboat” image
    Fusion results obtained by different fusion methods applied in “Kayak” image
    Fusion results obtained by different fusion methods applied in “Infrared4-Visible4” image
    Statistical polygonal images of different objective evaluation indicators
    Three-dimensional data images of “Octec”
    Three-dimensional data images of “Infrared4-Visible4”
    • Table 1. Objective evaluation indicators of visible image corresponding to different σr values

      View table

      Table 1. Objective evaluation indicators of visible image corresponding to different σr values

      σrSTDSFEI
      0.10.14700.03050.0936
      0.010.15140.04230.1290
      0.0010.15200.04430.1477
      0.00010.15210.04430.1481
      0.000010.15210.04430.1481
      0.20.14570.02640.0909
      0.020.15050.03990.1147
      0.0020.15200.04420.1466
      0.00020.15210.04430.1481
      0.000020.15210.04430.1481
    • Table 2. Objective evaluation indicators of different fusion methods

      View table

      Table 2. Objective evaluation indicators of different fusion methods

      ImageIndexDWTCVTDTCWTNSCT-DAPCNNNSDTSTLATLRRGTFMSVDours
      MI0.19160.18930.20110.20720.18020.24100.27930.21650.2841
      STD0.10290.09880.09740.07180.13320.10700.10510.08920.1152 (2)
      Un campQNCIE0.80290.80290.80300.80300.80290.80340.80410.80310.8039 (2)
      QCV425.1803461.3446426.053423.8996540.9333332.3001661.3364424.1480293.6759
      IE0.47090.46850.45600.45800.61320.62680.60260.45000.8334
      MI0.38760.37580.41270.42830.38380.44720.51760.46280.5200
      STD0.11670.11750.11360.11170.21320.11750.09590.10800.1573 (2)
      OctecQNCIE0.80660.80640.80700.80710.80730.80730.80890.80750.8094
      QCV648.2625408.1675629.2341626.6379391.1703274.64852264.01902.5192.0751
      IE0.96130.96450.96720.96600.96350.99340.83370.82981.0000
      MI0.25760.24180.26950.27380.29110.29080.62840.29650.4514 (2)
      STD0.05240.04820.04790.04490.1240.04980.13310.04150.0769 (3)
      SteamboatQNCIE0.80450.80430.80460.80460.80520.80470.81170.80470.8068 (2)
      QCV163.4563169.6701167.6022173.5677298.9255208.9697331.4260178.0732138.259
      IE0.39500.37420.37100.35670.61370.45270.57500.35160.7979
      MI0.44710.37860.44920.51840.49230.46960.54690.61980.5195(3)
      STD0.07310.07280.06960.06690.17070.06730.08640.06470.1444 (2)
      KayakQNCIE0.80590.80490.80580.80700.80700.80610.80720.80870.8087
      QCV876.5883622.103864.7246791.0015798.335855.55833060.6697.9614135.6981
      IE0.02790.02240.02070.01320.76510.02620.01910.00710.9980
      MI0.24850.24610.28660.27540.25520.29430.23900.28520.3233
      STD0.14650.13660.13640.12890.16970.14320.13300.12430.1707
      Infrared4-Visible4QNCIE0.80350.80340.80440.80380.80360.80400.80340.80390.8046
      QCV1133.91056.21502.31076.91845.6924.23831972.31045.83718.6361
      IE0.76430.75880.75110.74740.89730.87450.84500.73640.9504
    • Table 3. Running-time of eight other methods and proposed method

      View table

      Table 3. Running-time of eight other methods and proposed method

      ImagesDWTCVTDTCWTNSCT-DAPCNNNSDTSTLATLRRGTFMSVDours
      Un camp0.67072.16430.540841.7589101.5510227.67681.52590.164519.0259 (6)
      Octec0.29241.83120.4142139.7033326.4207218.30827.80890.508758.0978 (6)
      Steamboat0.28561.85110.4191118.7886267.4719251.77167.13310.435549.8169 (6)
      Kayak0.27771.85180.3950117.9741257.4367245.73366.29760.610449.4241 (6)
      Infrare4-Visible40.30541.88400.411527.812067.3537220.71371.12740.265311.6429 (6)
    Tools

    Get Citation

    Copy Citation Text

    Cheng Zhao, Yongdong Huang. Infrared and Visible Image Fusion via Rolling Guidance Filtering and Hybrid Multi-Scale Decomposition[J]. Laser & Optoelectronics Progress, 2019, 56(14): 141007

    Download Citation

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

    Category: Image Processing

    Received: Jan. 17, 2019

    Accepted: Feb. 20, 2019

    Published Online: Jul. 12, 2019

    The Author Email: Huang Yongdong (Huang_yongdong@163.com)

    DOI:10.3788/LOP56.141007

    Topics