• Chinese Optics Letters
  • Vol. 13, Issue 7, 071101 (2015)
Xuyang Xu, Enrong Li*, Xia Shen, and Shensheng Han
Author Affiliations
  • Key Laboratory for Quantum Optics and Center for Cold Atom Physics of the Chinese Academy of Science, Shanghai Institute of Optics and Fine Mechanics, Chinese Academy of Science, Shanghai 201800, China
  • show less
    DOI: 10.3788/COL201513.071101 Cite this Article Set citation alerts
    Xuyang Xu, Enrong Li, Xia Shen, Shensheng Han. Optimization of speckle patterns in ghost imaging via sparse constraints by mutual coherence minimization[J]. Chinese Optics Letters, 2015, 13(7): 071101 Copy Citation Text show less
    Experimental setup of GISC using DMD.
    Fig. 1. Experimental setup of GISC using DMD.
    Distributions of μij of the original matrix Dinit (blue line) and the optimized matrix Dopt (red line). Coherence μij of Dopt are concentrated in the region with smaller coherence values if compared with that of Dinit.
    Fig. 2. Distributions of μij of the original matrix Dinit (blue line) and the optimized matrix Dopt (red line). Coherence μij of Dopt are concentrated in the region with smaller coherence values if compared with that of Dinit.
    Simulation results of the test object (pattern of “tai chi”) with different numbers of samplings: (a) original object; reconstruction results with (b) K=500; (c) K=1000; (d) K=1500; (e) K=2000; (f) K=2500. Shown in the top line are the results obtained with Gaussian random matrix Ainit and in the bottom line the optimized matrix Aopt.
    Fig. 3. Simulation results of the test object (pattern of “tai chi”) with different numbers of samplings: (a) original object; reconstruction results with (b) K=500; (c) K=1000; (d) K=1500; (e) K=2000; (f) K=2500. Shown in the top line are the results obtained with Gaussian random matrix Ainit and in the bottom line the optimized matrix Aopt.
    RMSE and PSNR values of the reconstructed images in the simulation with different numbers of samplings using Ainit (blue line) and Aopt (red line).
    Fig. 4. RMSE and PSNR values of the reconstructed images in the simulation with different numbers of samplings using Ainit (blue line) and Aopt (red line).
    Experimental results with different numbers of samplings: (a) reference image; reconstructed results with (b) K=1000; (c) K=2000; (d) K=3000; (e) K=4000. Shown in the top line are the results obtained with Gaussian random matrix Ainit and in the bottom line the optimized matrix Aopt.
    Fig. 5. Experimental results with different numbers of samplings: (a) reference image; reconstructed results with (b) K=1000; (c) K=2000; (d) K=3000; (e) K=4000. Shown in the top line are the results obtained with Gaussian random matrix Ainit and in the bottom line the optimized matrix Aopt.
    RMSE and PSNR values of the reconstructed images with different numbers of samplings using Ainit (blue line) and Aopt (red line) in the experiment.
    Fig. 6. RMSE and PSNR values of the reconstructed images with different numbers of samplings using Ainit (blue line) and Aopt (red line) in the experiment.
    Xuyang Xu, Enrong Li, Xia Shen, Shensheng Han. Optimization of speckle patterns in ghost imaging via sparse constraints by mutual coherence minimization[J]. Chinese Optics Letters, 2015, 13(7): 071101
    Download Citation