• Acta Optica Sinica
  • Vol. 39, Issue 3, 0304001 (2019)
Fusheng Sun* and Xie Han
Author Affiliations
  • School of Data Science and Technology, North University of China, Taiyuan, Shanxi 030051, China
  • show less
    DOI: 10.3788/AOS201939.0304001 Cite this Article Set citation alerts
    Fusheng Sun, Xie Han. Super-Resolution Algorithm Based on Precise Color Vector Constraint of Light Field Camera[J]. Acta Optica Sinica, 2019, 39(3): 0304001 Copy Citation Text show less
    Principle diagram of optical field imaging
    Fig. 1. Principle diagram of optical field imaging
    Measured results of PSF of LYTRO camera
    Fig. 2. Measured results of PSF of LYTRO camera
    Hexagonal RST coordinate system
    Fig. 3. Hexagonal RST coordinate system
    Distribution and location diagram of diffusion function points (a) and local magnification diagram (b)
    Fig. 4. Distribution and location diagram of diffusion function points (a) and local magnification diagram (b)
    Distribution of color filters and diffusion function points
    Fig. 5. Distribution of color filters and diffusion function points
    Pyramid model
    Fig. 6. Pyramid model
    Pyramid algorithm model of LYTRO camera
    Fig. 7. Pyramid algorithm model of LYTRO camera
    Algorithm flow chart
    Fig. 8. Algorithm flow chart
    Original image information collected by LYTRO camera. (a) Image 1; (b) image 2
    Fig. 9. Original image information collected by LYTRO camera. (a) Image 1; (b) image 2
    Experimental results and local enlargement of 4 algorithms
    Fig. 10. Experimental results and local enlargement of 4 algorithms
    Color recovery results of two sets of images. (a) Bilinear algorithm; (b) Lu's algorithm; (c) adaptive algorithm; (d) proposed algorithm
    Fig. 11. Color recovery results of two sets of images. (a) Bilinear algorithm; (b) Lu's algorithm; (c) adaptive algorithm; (d) proposed algorithm
    Original image and sub-aperture images. (a) Original image; (b) sub-aperture image; (c) 7×7 sub-aperture images
    Fig. 12. Original image and sub-aperture images. (a) Original image; (b) sub-aperture image; (c) 7×7 sub-aperture images
    Original image and local magnified images. (a) Original image; (b) local magnification image 1; (c) local magnification image 2; (d) local two times enlarged image 3
    Fig. 13. Original image and local magnified images. (a) Original image; (b) local magnification image 1; (c) local magnification image 2; (d) local two times enlarged image 3
    Experimental comparison of algorithms. (a) Toolbox W/O rectification+dual three-time on-sample; (b) toolbox W/O rectification+super-resolution; (c) document algorithm[25]; (d) proposed algorithm
    Fig. 14. Experimental comparison of algorithms. (a) Toolbox W/O rectification+dual three-time on-sample; (b) toolbox W/O rectification+super-resolution; (c) document algorithm[25]; (d) proposed algorithm
    Point1234
    a(0,11.671)(0,23.758)(0,35.239)(0,46.837)
    b(10.107,5.836)(20.575,11.879)(30.518,17.620)(40.562,23.419)
    c(10.107,-5.836)(20.575,-11.879)(30.518,-17.620)(40.562,-23.419)
    d(0,-11.671)(0,-23.758)(0,-35.239)(0,-46.837)
    e(-10.701,-5.836)(-20.575,-11.879)(-30.518,-17.620)(-40.562,-23.419)
    f(-10.701,5.836)(-20.575,11.879)(-30.518,17.620)(-40.562,23.419)
    Table 1. Diffusion function point coordinates of different layers
    Column No.Mode
    1RGrRGrRR
    2GbBGbBGbGb
    3RGrRGrRR
    4GbBGbBGbGb
    5RGrRGrRR
    NGbBGbBGbGb
    Table 2. Color filter distribution mode
    Image 1Mean square error ERMean square error EGMean square error EBRunning time /sSignal-to-noise ratio /dB
    Bilinear method326.3563120.8871335.10953.782325.6157
    Adaptive method36.225621.156530.36893.5177635.4862
    Lu's method23.369713.163920.3547260.695340.5687
    Proposed method6.415410.47818.69714.3192642.2369
    Table 3. Image 1 evaluation data
    Image 2Mean square error ERMean square error EGMean square error EBRunning time /sSignal-to-noise ratio /dB
    Bilinear method186.477183.9741173.01632.063123.5872
    Adaptive method51.358821.367423.48561.69630730.3547
    Lu's method30.768111.478316.1031109.587533.6971
    Proposed method19.387521.683314.31293.632736.3618
    Table 4. Image 2 evaluation data
    Fusheng Sun, Xie Han. Super-Resolution Algorithm Based on Precise Color Vector Constraint of Light Field Camera[J]. Acta Optica Sinica, 2019, 39(3): 0304001
    Download Citation