1887
PDF

Abstract

Summary

Image super-resolution is crucial in computer vision, particularly for enhancing micro-CT images which are essential for detailed scientific analysis. This technique involves reconstructing high-resolution images from their lower-resolution versions, focusing on improving image details and sharpness. Our study tested three deep learning models—PRIDNet, MW-CNN, and VDSR—on a dataset of 1400 uniformly sized images, downsampled by a factor of three for the experiment. We evaluated model performance using various metrics including SSIM, MS-SSIM, PSNR, and UIQ, which assess similarity, quality across scales, error, and image distortion respectively. A targeted experimental strategy was employed to optimize performance by exploring different combinations of loss functions. The best results were achieved by adapting loss functions specific to each model’s needs, with combinations like L1+MS-SSIM proving effective in enhancing perceptual quality by preserving structural information. This emphasizes the critical role of carefully selecting both the model and its corresponding loss function for superior super-resolution outcomes in specialized imaging applications like microscopy.

EAGE Publications BV © 2024 The Authors © European Association of Geoscientists and Engineers
Loading

Article metrics loading...

/content/papers/10.3997/2214-4609.2024636005
2024-09-16
2026-02-12

  • From This Site

    /content/papers/10.3997/2214-4609.2024636005
    dcterms_title,dcterms_subject,pub_keyword
    -contentType:Journal -contentType:Contributor -contentType:Concept -contentType:Institution
    10
    5
Loading full text...

Full text loading...

/deliver/fulltext/2214-4609/2024/eight-eage-high-performance-computing-workshop/5_-_Mohamed_Awadalla.html?itemId=/content/papers/10.3997/2214-4609.2024636005&mimeType=html&fmt=ahah

References

  1. Y.Zhao, Z.Jiang, A.Men and G.Ju, “Pyramid Real Image Denoising Network,” 2019 IEEE Visual Communications and Image Processing (VCIP), Sydney, NSW, Australia, 2019, pp. 1–4, doi: 10.1109/VCIP47243.2019.8965754.
     https://doi.org/10.1109/VCIP47243.2019.8965754 [Google Scholar]
  2. P.Liu, H.Zhang, W.Lian and W.Zuo, “Multi-Level Wavelet Convolutional Neural Networks,” in IEEE Access, vol. 7, pp. 74973–74985, 2019, doi: 10.1109/ACCESS.2019.2921451.
     https://doi.org/10.1109/ACCESS.2019.2921451 [Google Scholar]
  3. J.Kim, J. K.Lee and K. M.Lee, “Accurate Image Super-Resolution Using Very Deep Convolutional Networks,” 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Las Vegas, NV, USA, 2016, pp. 1646–1654, doi: 10.1109/CVPR.2016.182.
     https://doi.org/10.1109/CVPR.2016.182 [Google Scholar]
  4. Zhou Wang, A. C. Bovik, H. R. Sheikh and E. P.Simoncelli, “Image quality assessment: from error visibility to structural similarity,” in IEEE Transactions on Image Processing, vol. 13, no. 4, pp. 600–612, April 2004, doi: 10.1109/TIP.2003.819861.
     https://doi.org/10.1109/TIP.2003.819861 [Google Scholar]
  5. Z.Wang, E. P.Simoncelli and A. C.Bovik, “Multiscale structural similarity for image quality assessment,” The Thrity-Seventh Asilomar Conference on Signals, Systems & Computers, 2003, Pacific Grove, CA, USA, 2003, pp. 1398–1402 Vol.2, doi: 10.1109/ACSSC.2003.1292216.
     https://doi.org/10.1109/ACSSC.2003.1292216 [Google Scholar]
  6. ZhouWang and A. C. Bovik, “A universal image quality index,” in IEEE Signal Processing Letters, vol. 9, no. 3, pp. 81–84, March 2002, doi: 10.1109/97.9958.3.
     https://doi.org/10.1109/97.9958.3 [Google Scholar]
  7. H.Zhao, O.Gallo, I.Frosio and J.Kautz, “Loss Functions for Image Restoration With Neural Networks,” in IEEE Transactions on Computational Imaging, vol. 3, no. 1, pp. 47–57, March 2017, doi: 10.1109/TCI.2016.2644865.
     https://doi.org/10.1109/TCI.2016.2644865 [Google Scholar]
/content/papers/10.3997/2214-4609.2024636005
Loading
Comparative Analysis of Super Resolution Techniques in Micro-CT Imaging
Conference Proceedings 2024, 1 (2024); https://doi.org/10.3997/2214-4609.2024636005
/content/papers/10.3997/2214-4609.2024636005
/content/papers/10.3997/2214-4609.2024636005
Loading

Data & Media loading...

Most Cited This Month Most Cited RSS feed

This is a required field
Please enter a valid email address
Approval was a Success
Invalid data
An Error Occurred
Approval was partially successful, following selected items could not be processed due to error