• Acta Optica Sinica
  • Vol. 42, Issue 9, 0911001 (2022)
Moxuan Yang1、2、3、4, Yuanmeng Zhao1、2、3、4、*, Fengxia Zhu1、2、3、4, Hongfei Zhang1、2、3、4, Zhilong Li1、2、3、4, and Cunlin Zhang1、2、3、4
Author Affiliations
  • 1Key Laboratory of Terahertz Optoelectronics, Ministry of Education, Beijing 100048, China
  • 2Beijing Key Laboratory for Terahertz Spectroscopy and Imaging, Beijing 100048, China
  • 3Beijing Advanced Innovation Center for Imaging Theory and Technology, Beijing 100048, China
  • 4Department of Physics, Capital Normal University, Beijing 100048, China
  • show less
    DOI: 10.3788/AOS202242.0911001 Cite this Article Set citation alerts
    Moxuan Yang, Yuanmeng Zhao, Fengxia Zhu, Hongfei Zhang, Zhilong Li, Cunlin Zhang. Terahertz Light Field Imaging Occlusion Removing Algorithm Based on Multi-Perspective Synthetic Aperture[J]. Acta Optica Sinica, 2022, 42(9): 0911001 Copy Citation Text show less
    Comparison between large aperture camera and single camera
    Fig. 1. Comparison between large aperture camera and single camera
    Processing flowchart of proposed algorithm
    Fig. 2. Processing flowchart of proposed algorithm
    Reconstruction process of sub-image array
    Fig. 3. Reconstruction process of sub-image array
    Estimation result of minimum generalization error
    Fig. 4. Estimation result of minimum generalization error
    Digital refocus images with different depths. (a) n=100 mm; (b) n=110 mm; (c) n=120 mm; (d) n=130 mm; (e) n=140 mm; (f) n=150 mm
    Fig. 5. Digital refocus images with different depths. (a) n=100 mm; (b) n=110 mm; (c) n=120 mm; (d) n=130 mm; (e) n=140 mm; (f) n=150 mm
    Curve signals before and after filtering
    Fig. 6. Curve signals before and after filtering
    Reconstructed images processed by different algorithms. (a) Visible image; (b) original image; (c) digital refocusing algorithm; (d) proposed algorithm; (e) unocclusion
    Fig. 7. Reconstructed images processed by different algorithms. (a) Visible image; (b) original image; (c) digital refocusing algorithm; (d) proposed algorithm; (e) unocclusion
    Influence of occlusion with different thickness on reconstructed image. (a) n=5 cm; (b) n=6 cm; (c) n=7 cm
    Fig. 8. Influence of occlusion with different thickness on reconstructed image. (a) n=5 cm; (b) n=6 cm; (c) n=7 cm
    IndexOriginal imageRef. [16]Proposed algorithm
    MSE1720.87576435.4115353.320661
    MAE11.6940881.6041260.161418
    SNR70.76141987.62746797.906668
    PSNR15.77330832.63935642.918558
    Table 1. Objective evaluation of image quality with reference
    IndexOriginal imageRef. [16]Proposed algorithmNo shielding image
    Brenner gradient function1483.10011492.19061563.67971575.1642
    Roberts function20.59434020.76618021.21580021.982600
    SMD12.261412.124912.535913.0081
    EOG1294.69321331.12841338.51961349.5099
    Table 2. Objective evaluation of image quality without reference
    Moxuan Yang, Yuanmeng Zhao, Fengxia Zhu, Hongfei Zhang, Zhilong Li, Cunlin Zhang. Terahertz Light Field Imaging Occlusion Removing Algorithm Based on Multi-Perspective Synthetic Aperture[J]. Acta Optica Sinica, 2022, 42(9): 0911001
    Download Citation