Photonics Research, Volume. 12, Issue 10, 2214(2024)
Enhanced ultrafine multimode fiber imaging based on mode modulation through singular value decomposition Editors' Pick
Fig. 1. Schematic overview of the mode modulation-enhanced MMF imaging system.
Fig. 2. Experimental setup for a mode modulation enhanced MMF imaging system. BE, beam expander; DMD, digital micromirror device; SLM, spatial light modulator; PBS, polarizing beam splitter; BS, beam splitter; OBJ, microscope objective lens; MMF, multimode fiber; M, mirror. (a) TM acquisition. (b) MMF imaging with mode modulation. Singular vector 2D map with orange box: a chosen singular vector 2D map displayed on the SLM for mode modulation.
Fig. 3. Speckle properties customization through mode modulation in the 40 μm core MMF. (a) Relationship between speckle grain size and singular vector number. (b) Relationship between speckle intensity and singular vector number. (c) Speckle patterns obtained with mode modulation by the two different singular vectors (No. 3, No. 279). (d) Fourier space distributions of the two speckle patterns. (e) Central cuts through the two Fourier space distributions.
Fig. 4. Image reconstruction results of (a) MNIST dataset, (b) QuickDraw dataset, (c) SIPaKMeD dataset, and (d) ImageNet dataset on the 40 μm core MMF system.
Fig. 5. Imaging analysis of the 40 μm core MMF system. (a) Image reconstruction accuracy at different speckle grain sizes and fitting curves of discrete data points. (b) Autocorrelation and FWHM of the speckle patterns obtained with mode modulation by the optimal singular vectors. The insets in the upper left corner and upper right corner are the corresponding speckle patterns and corresponding speckle intensities, respectively.
Fig. 6. Comparison of the MMF systems with two different core diameters. (a) Reconstruction results of both MMF systems without mode modulation and with mode modulation. (b) Output speckle patterns of the same input pattern captured by both MMF systems without mode modulation and with mode modulation.
Fig. 7. Speckle properties customization through mode modulation in the 105 μm core MMF. (a) Relationship between speckle grain size and singular vector number. (b) Relationship between speckle intensity and singular vector number. (c) Speckle patterns obtained with mode modulation by the two different singular vectors (No. 2, No. 225). (d) Fourier space distributions of the two speckle patterns. (e) Central cuts through the two Fourier space distributions.
Fig. 8. Image reconstruction results of (a) MNIST dataset, (b) QuickDraw dataset, (c) SIPaKMeD dataset, and (d) ImageNet dataset on the 105 μm core MMF system.
Fig. 9. Imaging analysis of the 105 μm core MMF system. (a) Image reconstruction accuracy at different speckle grain sizes and fitting curves of discrete data points. (b) Autocorrelation and FWHM of the speckle patterns obtained with mode modulation by the optimal singular vectors. The insets in the upper left corner and upper right corner are the corresponding speckle patterns and corresponding speckle intensities, respectively.
Get Citation
Copy Citation Text
Ning Zhan, Zhenming Yu, Liming Cheng, Jingyue Ma, Jiayu Di, Yueheng Lan, Kun Xu, "Enhanced ultrafine multimode fiber imaging based on mode modulation through singular value decomposition," Photonics Res. 12, 2214 (2024)
Category: Imaging Systems, Microscopy, and Displays
Received: May. 8, 2024
Accepted: Jul. 23, 2024
Published Online: Sep. 27, 2024
The Author Email: Zhenming Yu (yuzhenming@bupt.edu.cn)