• Laser & Optoelectronics Progress
  • Vol. 57, Issue 4, 041019 (2020)
Guoqing Yan, Fengbao Yang*, Xiaoxia Wang, and Yong Tao
Author Affiliations
  • College of Information and Communication Engineering, North University of China, Taiyuan, Shanxi 030051, China
  • show less
    DOI: 10.3788/LOP57.041019 Cite this Article Set citation alerts
    Guoqing Yan, Fengbao Yang, Xiaoxia Wang, Yong Tao. Computational Ghost Imaging Based on Orthogonal Sinusoidal Speckle[J]. Laser & Optoelectronics Progress, 2020, 57(4): 041019 Copy Citation Text show less
    Schematic of CGI scheme
    Fig. 1. Schematic of CGI scheme
    Reconstruction results. (a) Original image; (b) reconstruction result of SGI; (c) reconstruction result using only sinusoidal speckle; (d) reconstruction result using only cosine speckle
    Fig. 2. Reconstruction results. (a) Original image; (b) reconstruction result of SGI; (c) reconstruction result using only sinusoidal speckle; (d) reconstruction result using only cosine speckle
    Principle of image generation by using orthogonal sinusoidal speckle. (a) 45° sinusoidal speckle; (b) 135° sinusoidal speckle; (c) speckle after superposition
    Fig. 3. Principle of image generation by using orthogonal sinusoidal speckle. (a) 45° sinusoidal speckle; (b) 135° sinusoidal speckle; (c) speckle after superposition
    Reconstruction results of different gray targets obtained by proposed algorithm under different upper frequency limits. (a) Objects to be imaged; (b) 20 rad/rotation (1849 speckles); (c) 22 rad/rotation (2209 speckles); (d) 24 rad/rotation (2601 speckles); (e) 26 rad/rotation (3025 speckles); (f) 28 rad/rotation (3481 speckles); (g) 30 rad/rotation (4096 speckles)
    Fig. 4. Reconstruction results of different gray targets obtained by proposed algorithm under different upper frequency limits. (a) Objects to be imaged; (b) 20 rad/rotation (1849 speckles); (c) 22 rad/rotation (2209 speckles); (d) 24 rad/rotation (2601 speckles); (e) 26 rad/rotation (3025 speckles); (f) 28 rad/rotation (3481 speckles); (g) 30 rad/rotation (4096 speckles)
    Performance parameter curves of different gray-scale target reconstruction results under different upper frequency limits. (a) PSNR; (b) SSIM
    Fig. 5. Performance parameter curves of different gray-scale target reconstruction results under different upper frequency limits. (a) PSNR; (b) SSIM
    Multi-slit diagonal stripe patterns with different widths between black and white grids and their reconstructed results. (a) 2 pixel; (b) 4 pixel; (c) 6 pixel; (d) 8 pixel; (e) 10 pixel; (f) 12 pixel
    Fig. 6. Multi-slit diagonal stripe patterns with different widths between black and white grids and their reconstructed results. (a) 2 pixel; (b) 4 pixel; (c) 6 pixel; (d) 8 pixel; (e) 10 pixel; (f) 12 pixel
    PSNR curve of reconstructed image with different widths of multi-slit diagonal stripes
    Fig. 7. PSNR curve of reconstructed image with different widths of multi-slit diagonal stripes
    Imaging results of various reconstruction methods. (a) Objects to be imaged; (b) TGI; (c) DGI; (d) HCGI; (e) proposed method
    Fig. 8. Imaging results of various reconstruction methods. (a) Objects to be imaged; (b) TGI; (c) DGI; (d) HCGI; (e) proposed method
    Imaging results of SGI and proposed method
    Fig. 9. Imaging results of SGI and proposed method
    EvaluationindexTGI(Gaussian)DGI(Gaussian)HCGI(Hadamard)Proposedalgorithm
    PSNR63.037565.693962.988567.4625
    SSIM0.13910.31920.63990.8287
    Table 1. Performance comparison of four imaging methods for “rice”
    EvaluationindexTGI(Gaussian)DGI(Gaussian)HCGI(Hadamard)Proposedalgorithm
    PSNR58.413459.137062.230365.2461
    SSIM0.18570.26280.46050.7088
    Table 2. Performance comparison of four imaging methods for “camera-man”
    EvaluationindexTGI(Gaussian)DGI(Gaussian)HCGI(Hadamard)Proposed
    PSNR57.647158.636763.160964.4823
    SSIM0.05690.08080.49210.6750
    Table 3. Performance comparison of four imaging methods for “moon”
    Guoqing Yan, Fengbao Yang, Xiaoxia Wang, Yong Tao. Computational Ghost Imaging Based on Orthogonal Sinusoidal Speckle[J]. Laser & Optoelectronics Progress, 2020, 57(4): 041019
    Download Citation