• Laser & Optoelectronics Progress
  • Vol. 59, Issue 4, 0409001 (2022)
Xiao Sun and Chao Han*
Author Affiliations
  • Key Laboratory of Advanced Perception and Intelligent Control of High-end Equipment, Ministry of Education, Anhui Polytechnic University, Wuhu , Anhui 241000, China
  • show less
    DOI: 10.3788/LOP202259.0409001 Cite this Article Set citation alerts
    Xiao Sun, Chao Han. Three-Dimensional Phase-Only Holographic Display Based on Deep Learning and Angular-Spectrum Layer-Oriented[J]. Laser & Optoelectronics Progress, 2022, 59(4): 0409001 Copy Citation Text show less
    Schematic diagram of improved algorithm
    Fig. 1. Schematic diagram of improved algorithm
    LeNet model structure
    Fig. 2. LeNet model structure
    Contrast simulation of two-dimensional Lena image. (a) Original image; (b) hologram obtained of the proposed algorithm; (c) hologram obtained by the algorithm in Ref.[22]; (d) reconstructed image of the proposed algorithm; (e) reconstructed image using the algorithm in Ref.[22]
    Fig. 3. Contrast simulation of two-dimensional Lena image. (a) Original image; (b) hologram obtained of the proposed algorithm; (c) hologram obtained by the algorithm in Ref.[22]; (d) reconstructed image of the proposed algorithm; (e) reconstructed image using the algorithm in Ref.[22]
    Small train model of modeling. (a) Depth map; (b) shading map
    Fig. 4. Small train model of modeling. (a) Depth map; (b) shading map
    Output image. (a) Phase-only hologram of three-dimensional small train; (b) reconstruction image of three-dimensional small train image
    Fig. 5. Output image. (a) Phase-only hologram of three-dimensional small train; (b) reconstruction image of three-dimensional small train image
    Contrast simulation images. (a) Reconstructed image of proposed algorithm at d1=210 mm ; (b) reconstructed image at d1=210 mm using the algorithm in Ref.[22]; (c) reconstructed image of proposed algorithm at d2=220 mm; (d) reconstructed image at d2=220 mm using algorithm in Ref.[22]; (e) reconstructed image of proposed algorithm at d3=230 mm; (f) reconstructed image at d3=230 mm using the algorithm in Ref.[22]
    Fig. 6. Contrast simulation images. (a) Reconstructed image of proposed algorithm at d1=210 mm ; (b) reconstructed image at d1=210 mm using the algorithm in Ref.[22]; (c) reconstructed image of proposed algorithm at d2=220 mm; (d) reconstructed image at d2=220 mm using algorithm in Ref.[22]; (e) reconstructed image of proposed algorithm at d3=230 mm; (f) reconstructed image at d3=230 mm using the algorithm in Ref.[22]
    Experimental light path diagram
    Fig. 7. Experimental light path diagram
    Experiment reconstruction images. (a) Reconstructed image of the proposed algorithm at d1=210 mm; (b) reconstructed image at d1=210 mm using the algorithm inRef.[22]; (c) reconstructed image at d1=210 mm using the algorithm in Ref.[23]; (d) reconstructed image of the proposed algorithm at d2=220 mm; (e) reconstructed image at d2=220 mm using the algorithm in Ref.[22]; (f) reconstructed image at d2=220 mm using the algorithm in Ref.[23]; (g) reconstructed imageof the proposed algorithm at d3=230 mm; (h) reconstructed image at d3=230 mm using the algorithm in Ref.[22]; (i) reconstructed image at d3=230 mm using the algorithm in Ref.[23]
    Fig. 8. Experiment reconstruction images. (a) Reconstructed image of the proposed algorithm at d1=210 mm; (b) reconstructed image at d1=210 mm using the algorithm inRef.[22]; (c) reconstructed image at d1=210 mm using the algorithm in Ref.[23]; (d) reconstructed image of the proposed algorithm at d2=220 mm; (e) reconstructed image at d2=220 mm using the algorithm in Ref.[22]; (f) reconstructed image at d2=220 mm using the algorithm in Ref.[23]; (g) reconstructed imageof the proposed algorithm at d3=230 mm; (h) reconstructed image at d3=230 mm using the algorithm in Ref.[22]; (i) reconstructed image at d3=230 mm using the algorithm in Ref.[23]
    ParameterFig. 3(d)Fig. 3(e)
    pPSNR /dB24.310015.5758
    sSSIM0.80320.5946
    Table 1. PSNR and SSIM of reconstructed image
    ParameterFig. 3(b)Fig. 3(c)
    sSSIM0.78250.4369
    Table 2. SSIM of pure phase-only hologram
    Image size512×5121024×10242048×20484096×4096
    Running time of algorithm in Ref. [22] /ms3.47.919.345.9
    Running time of improved algorithm /ms4.18.920.249.4
    Table 3. Comparison of operation time of different size pictures
    Xiao Sun, Chao Han. Three-Dimensional Phase-Only Holographic Display Based on Deep Learning and Angular-Spectrum Layer-Oriented[J]. Laser & Optoelectronics Progress, 2022, 59(4): 0409001
    Download Citation