Academia.eduAcademia.edu

Figure 2 – uploaded by Đan Đan

See full PDFdownloadDownload figure
Fig 2. Scheme of the DMD-based LED-illumination optical sectioning SIM system. The binary fringe pattern on DMD is de-magnified and projected onto the specimen through a collimating lens anda microscope objective lens. Higher orders of spatial frequencies of the binary fringe are naturally blocked by the optics, leading to a sinusoidal fringe illumination in the sample plane. Fluorescence light from the specimen is then imaged onto the sCMOS camera.

Figure 2 Scheme of the DMD-based LED-illumination optical sectioning SIM system. The binary fringe pattern on DMD is de-magnified and projected onto the specimen through a collimating lens anda microscope objective lens. Higher orders of spatial frequencies of the binary fringe are naturally blocked by the optics, leading to a sinusoidal fringe illumination in the sample plane. Fluorescence light from the specimen is then imaged onto the sCMOS camera.

Related Figures (4)

Fig 1. Flowcharts of the FABMED-HS algorithm (left side) and the SHT algorithm (right side).
Fig 1. Flowcharts of the FABMED-HS algorithm (left side) and the SHT algorithm (right side).
Fig 3. Raw image frames (a) (b) under structured illumination of the pollen grain with phase-shift by 2771/3, and the input image (c) obtained by subtracting (a) and (b). The scale bar is 5um. The tested specimen is a mixed pollen grain specimen purchased from Carolina Biological Sup- ply Company (Burlington, USA), which exhibits strong auto-fluorescence under the excitation of 450 nm LED light. In order to obtain the best signal-to-noise ratio of image [18], all of the raw images were captured by overwriting the 65536 gray scales of the dynamic range of 16bits gray depth of the camera in exposure time of 10 ms. Fig. 3(A) and 3(B) present the captured two raw images of the pollen grain with a phase-shift of 27/3. As described above, the subtract- ing operation between Fig. 3(A) and 3(B) generates an input image Is that contains the desir- able in-focus information shown in Fig. 3(C). Using the same input image Is, we reconstruct the sectioned image by using our proposed SHT algorithm (Fig. 4(A)) and the FABEMD-HS algorithm (Fig. 4(B)), respectively. In the FABEMD-HS algorithm, we decompose the Is into nine BIMFs and selectively reconstruct the optimal band-pass filtered pattern to obtain the sec- tioned image according to Ref. [14]. A subset of the data is enlarged in Fig. 4(C) and 4(D). To
Fig 3. Raw image frames (a) (b) under structured illumination of the pollen grain with phase-shift by 2771/3, and the input image (c) obtained by subtracting (a) and (b). The scale bar is 5um. The tested specimen is a mixed pollen grain specimen purchased from Carolina Biological Sup- ply Company (Burlington, USA), which exhibits strong auto-fluorescence under the excitation of 450 nm LED light. In order to obtain the best signal-to-noise ratio of image [18], all of the raw images were captured by overwriting the 65536 gray scales of the dynamic range of 16bits gray depth of the camera in exposure time of 10 ms. Fig. 3(A) and 3(B) present the captured two raw images of the pollen grain with a phase-shift of 27/3. As described above, the subtract- ing operation between Fig. 3(A) and 3(B) generates an input image Is that contains the desir- able in-focus information shown in Fig. 3(C). Using the same input image Is, we reconstruct the sectioned image by using our proposed SHT algorithm (Fig. 4(A)) and the FABEMD-HS algorithm (Fig. 4(B)), respectively. In the FABEMD-HS algorithm, we decompose the Is into nine BIMFs and selectively reconstruct the optimal band-pass filtered pattern to obtain the sec- tioned image according to Ref. [14]. A subset of the data is enlarged in Fig. 4(C) and 4(D). To
Fig 4. The reconstructed sectioned images of the pollen grain by the proposed SHT algorithm (a), and the FABEMD-HS algorithm (b). Magnifications of the boxed regions of (a) and (b) are shown in (c) and (d), respectively. Intensity profiles along the dashed lines indicated in (c) and (d) are plotted in (e). The scale bar is 5um.
Fig 4. The reconstructed sectioned images of the pollen grain by the proposed SHT algorithm (a), and the FABEMD-HS algorithm (b). Magnifications of the boxed regions of (a) and (b) are shown in (c) and (d), respectively. Intensity profiles along the dashed lines indicated in (c) and (d) are plotted in (e). The scale bar is 5um.
Fig 5. 3-D reconstructed images of the mixed pollen grains by using the RMS algorithm (Left) and the SHT algorithm (Right). The scale bar is 30 um. (S1 Video)
Fig 5. 3-D reconstructed images of the mixed pollen grains by using the RMS algorithm (Left) and the SHT algorithm (Right). The scale bar is 30 um. (S1 Video)
Related topics:
Computer ScienceAlgorithmsArtificial IntelligenceImage ProcessingMedicineOptical SectioningThree Dimensional ImagingCell Imaging
580 California St., Suite 400
San Francisco, CA, 94104
© 2025 Academia. All rights reserved