• Laser & Optoelectronics Progress
  • Vol. 60, Issue 16, 1610003 (2023)
Fang Zhang1,3, Wenheng Li2,3, Wen Wang1,3,*, and Rui Zhao2,3
Author Affiliations
  • 1School of Life Sciences, Tiangong University, Tianjin 300387, China
  • 2School of Electronics & Information Engineering, Tiangong University, Tianjin 300387, China
  • 3Tianjin Key Laboratory of Photoelectric Detection Technology and System, Tianjin 300387, China
  • show less
    DOI: 10.3788/LOP222277 Cite this Article Set citation alerts
    Fang Zhang, Wenheng Li, Wen Wang, Rui Zhao. Phase Recovery of Electronic Speckle Interferometric Fringe Pattern Using Deep Learning[J]. Laser & Optoelectronics Progress, 2023, 60(16): 1610003 Copy Citation Text show less
    U-Net network structure
    Fig. 1. U-Net network structure
    Schematic of sub-pixel convolution upsampling
    Fig. 2. Schematic of sub-pixel convolution upsampling
    DSSINet network structure
    Fig. 3. DSSINet network structure
    USS-Net structure
    Fig. 4. USS-Net structure
    SSIF dataset examples
    Fig. 5. SSIF dataset examples
    ESIF dataset examples
    Fig. 6. ESIF dataset examples
    Schematic of phase recovery effect of SSIF dataset. (a) Simulated fringe image; (b) network output diagram; (c) true package phase value; (d) unfolding phase of network output diagram; (e) true unwrapping phase value
    Fig. 7. Schematic of phase recovery effect of SSIF dataset. (a) Simulated fringe image; (b) network output diagram; (c) true package phase value; (d) unfolding phase of network output diagram; (e) true unwrapping phase value
    Schematic of phase recovery effect of ESIF dataset. (a) Experimental fringe image; (b) network output diagram; (c) wrapped phase value obtained by four-step phase shifting method; (d) unfolding phase of network output diagram; (e) true unwrapping phase value
    Fig. 8. Schematic of phase recovery effect of ESIF dataset. (a) Experimental fringe image; (b) network output diagram; (c) wrapped phase value obtained by four-step phase shifting method; (d) unfolding phase of network output diagram; (e) true unwrapping phase value
    Unfolded phase error analysis. (a)-(e) Phase error corresponding to the five experimental fringe patterns in Fig. 8 respectively
    Fig. 9. Unfolded phase error analysis. (a)-(e) Phase error corresponding to the five experimental fringe patterns in Fig. 8 respectively
    Subpixel convolution moduleStructured feature enhancement moduleRMSE
    ××4.5381
    ×4.3488
    ×4.0381
    4.0310
    Table 1. Ablation of USS-Net
    NetworkRMSENumber of parameters /106
    FCN100.2439129.2291
    ENet145.73630.3743
    DeepLabV389.167238.5404
    UperNet106.0545120.7908
    U-Net4.53817.7600
    USS-Net4.031031.5247
    Table 2. Comparison of several classical semantic segmentation networks
    Fang Zhang, Wenheng Li, Wen Wang, Rui Zhao. Phase Recovery of Electronic Speckle Interferometric Fringe Pattern Using Deep Learning[J]. Laser & Optoelectronics Progress, 2023, 60(16): 1610003
    Download Citation