• Laser & Optoelectronics Progress
  • Vol. 60, Issue 2, 0200001 (2023)
Liheng Bian1、2、**, Daoyu Li1、2, Xuyang Chang1、2, and Jinli Suo3、*
Author Affiliations
  • 1School of Information and Electronics, Beijing Institute of Technology, Beijing 100081, China
  • 2Advanced Research Institute of Multidisciplinary Sciences, Beijing Institute of Technology, Beijing 100081, China
  • 3Department of Automation, Tsinghua University, Beijing 100084, China
  • show less
    DOI: 10.3788/LOP221245 Cite this Article Set citation alerts
    Liheng Bian, Daoyu Li, Xuyang Chang, Jinli Suo. Theory and Approach of Large-Scale Computational Reconstruction[J]. Laser & Optoelectronics Progress, 2023, 60(2): 0200001 Copy Citation Text show less
    Flowchart of alternating projection algorithms. (a) GS algorithm; (b) input-output algorithm; (c) output-output algorithm; (d) hybrid input-output algorithm
    Fig. 1. Flowchart of alternating projection algorithms. (a) GS algorithm; (b) input-output algorithm; (c) output-output algorithm; (d) hybrid input-output algorithm
    FPM reconstruction using AP algorithm[25].(a) FPM system structure; (b) FPM imaging model; (c) strength constraints imposed by AP algorithm in solution space; (d) AP solving FPM process
    Fig. 2. FPM reconstruction using AP algorithm[25].(a) FPM system structure; (b) FPM imaging model; (c) strength constraints imposed by AP algorithm in solution space; (d) AP solving FPM process
    Applications of FPM technique[25]. (a) Quantitative phase imaging of blood smear; (b) phase images of live HeLa cell; (c) phase images of mitosis and apoptosis events of live HeLa cell captured by annular illumination FPM at a frame rate of 25 Hz; (d) reconstruction of three-dimensional refractive index of HeLa cells by FPM; (e) topographic map of a 3D surface via FP technology
    Fig. 3. Applications of FPM technique[25]. (a) Quantitative phase imaging of blood smear; (b) phase images of live HeLa cell; (c) phase images of mitosis and apoptosis events of live HeLa cell captured by annular illumination FPM at a frame rate of 25 Hz; (d) reconstruction of three-dimensional refractive index of HeLa cells by FPM; (e) topographic map of a 3D surface via FP technology
    CDI techniques and reconstruction algorithms[30]. (a) Plane-wave CDI; (b) Bragg CDI; (c) ptychographic CDI; (d) Fresnel CDI; (e) reflection CDI; (f) flowchart of alternating-projection-based CDI reconstruction algorithm
    Fig. 4. CDI techniques and reconstruction algorithms[30]. (a) Plane-wave CDI; (b) Bragg CDI; (c) ptychographic CDI; (d) Fresnel CDI; (e) reflection CDI; (f) flowchart of alternating-projection-based CDI reconstruction algorithm
    Applications of CDI technique[30]. (a) 3D mass density distribution of an unstained yeast spore cell; (b) 3D image of an unstained human chromosome; (c) reconstruction of an unstained herpesvirus virion; (d) quantitative 3D measurement of osteocyte; (e) representative diffraction pattern of a giant mimivirus particle; (f) 3D reconstruction of a mimivirus; (g) diffraction pattern of a nanocrystal; (h) electron density map of 2mFo-DFc
    Fig. 5. Applications of CDI technique[30]. (a) 3D mass density distribution of an unstained yeast spore cell; (b) 3D image of an unstained human chromosome; (c) reconstruction of an unstained herpesvirus virion; (d) quantitative 3D measurement of osteocyte; (e) representative diffraction pattern of a giant mimivirus particle; (f) 3D reconstruction of a mimivirus; (g) diffraction pattern of a nanocrystal; (h) electron density map of 2mFo-DFc
    Framework of DIP algorithm. (a) Principle of DIP; (b) common CNN structure; (c) UNet structure; (d) deep decoder structure
    Fig. 6. Framework of DIP algorithm. (a) Principle of DIP; (b) common CNN structure; (c) UNet structure; (d) deep decoder structure
    Results of DIP algorithm and comparisons with other algorithms in each task. (a) Inpainting[31]; (b) diffraction imaging[51]; (c) phase unwarpping[63]
    Fig. 7. Results of DIP algorithm and comparisons with other algorithms in each task. (a) Inpainting[31]; (b) diffraction imaging[51]; (c) phase unwarpping[63]
    Framework and applications of the PnP-GAP optimization. (a) Diagram of plug-and-play optimization framework based on GAP; (b) (c) comparison of large-scale snapshot compressive imaging and Fourier ptychographic microscopy between plug-and-play optimization and other methods, respectively[6,90]
    Fig. 8. Framework and applications of the PnP-GAP optimization. (a) Diagram of plug-and-play optimization framework based on GAP; (b) (c) comparison of large-scale snapshot compressive imaging and Fourier ptychographic microscopy between plug-and-play optimization and other methods, respectively[6,90]
    Fusion process of living glioblastoma observed by using plug-and-play optimization framework based on GAP[6]
    Fig. 9. Fusion process of living glioblastoma observed by using plug-and-play optimization framework based on GAP[6]
    AlgorithmNetworkInputRegularizationApplication
    Algorithm in Ref.[31UNetDegraded image/Denoising & Inpainting & SR
    Algorithm in Ref.[50CNNPSF/FPM
    Algorithm in Ref.[47CNNMeasurement/PR
    Algorithm in Ref.[52UNetMeasurement/CT
    Algorithm in Ref.[56UNetDegraded imageTVDenoising & Deblurring
    Algorithm in Ref.[51UNetMeasurement/PR
    Algorithm in Ref.[53DDNoise/MRI
    Algorithm in Ref.[612×CNNNoise/MR correction
    Algorithm in Ref.[603×DDNoiseDark channel priorRetinal image enhancement
    Algorithm in Ref.[57UNetInitial reconstructionTVCT
    Algorithm in Ref.[58UNetNoiseNonlocalOCT
    Algorithm in Ref.[59UNetReference image/MRI
    Algorithm in Ref.[412×CNN+VAEHaze imageHSV loss & Smooth loss on air-light mapDehaze
    Algorithm in Ref.[613×UNetNoiseTVDynamic PET
    Algorithm in Ref.[63UNetMeasurement/Phase unwarpping
    Algorithm in Ref.[64DNNTo-be-updated image/SCI
    Table 1. DIP algorithms and their applications
    ModelFrameworkFidelity termRegularization term
    y=fxHQSxk+1=argminxy-fx22+μz-x22zk+1=argminzRz+μz-x22
    FISTAxk=proxLzkzk+1=xk+tk-1tk+1xk-xk-1
    ADMMxk+1=argminxLρx,zk,λkzk+1=argminzLρxk+1,zk,λk
    y=AxGAPxk+1=zk+ATAAT-1y-Azkzk+1=argminxRxk+1
    AMPzk=y-Axk+1δzk-1R'xk-1+A*zk-1xk+1=Rxk+A*zk
    REDxk+1=xk-μ1σ2ATAxk-y+λxk-xk+1xk+1=Rxk
    Table 2. Plug-and-play optimization framework
    AlgorithmFrameworkDenoiserApplication
    Algorithm in Ref.[75ADMMBM3DTomography
    Algorithm in Ref.[71FISTA(FASTA)DnCNNCDI & CDP
    Algorithm in Ref.[74SGD & FISTATV & BM3DFPM
    Algorithm in Ref.[77ADMMDnCNNTomography
    Algorithm in Ref.[73ADMM & FISTADnCNN & MemNet & Residual UnetMRI & CDP
    Algorithm in Ref.[113GDBM3D & FFDNETCDP
    Algorithm in Ref.[114REDNLM & BM3DDenoising & SR & Deblurring
    Algorithm in Ref.[68HQSSRResNetSR
    Algorithm in Ref.[90GAPFFDNETSCI
    Algorithm in Ref.[115ADMMDnCNNCDI
    Algorithm in Ref.[77GDDnCNNTomography
    Algorithm in Ref.[88ISTANLMInpainting & Deblurring
    Algorithm in Ref.[116ADMMBM3DHyperspectral PR
    Algorithm in Ref.[6117GAPFFDNETCDI & CDP & FPM & Pixel SR
    Algorithm in Ref.[118GECModified DnCNNMRI
    Table 3. Plug-and-play optimization frameworks and their applications
    Liheng Bian, Daoyu Li, Xuyang Chang, Jinli Suo. Theory and Approach of Large-Scale Computational Reconstruction[J]. Laser & Optoelectronics Progress, 2023, 60(2): 0200001
    Download Citation