Laser & Optoelectronics Progress, Volume. 62, Issue 4, 0437007(2025)

High Dynamic Range Image Compression Based on a Multi-Scale Feature Network

Yabo Liu1、*, Xiaoquan Yang2, and Tao Jiang2
Author Affiliations
  • 1School of Biomedical Engineering, Hainan University, Haikou 570228, Hainan , China
  • 2Suzhou Brain Space Information Research Institute, Huazhong University of Science and Technology, Suzhou 215000, Jiangsu , China
  • show less
    Figures & Tables(8)
    Cellular architecture data and neural fiber data of the macaque brain. (a) Cellular architecture data; (b) gray matter background data of cortex; (c) white matter background data near the internal capsule; (d) signal data of cortex; (e) neural fiber data; (f) background data; (g) signal data of the soma and dendrite; (h) signal data of the axon terminal; (i) gray frequency histogram corresponding to (b)‒(d); (j) gray frequency histogram corresponding to (f); (k) gray value change curve corresponding to (g) and (h)
    Overall network structure of MCNet
    Comparison of the rate-distortion curves for various methods on the cellular architecture dataset. (a) MS-SSIM; (b) PSNR
    Comparison of reconstruction results for different methods at various compression ratios on the cellular architecture dataset. (a1)‒(a5) Reconstructed images of different methods at low compression ratios; (b1)‒(b5) reconstructed images of different methods at high compression ratios; (c) gray value variation curves corresponding to the dotted lines in (a1)‒(a5); (d) gray value variation curves corresponding to the dotted lines in (b1)‒(b5)
    Comparison of the rate-distortion curves for various methods on the nerve fiber dataset. (a) MS-SSIM; (b) PSNR
    Comparison of the reconstruction results for different methods at various compression ratios on the neural fiber dataset. (a1)‒(a4) Reconstructed images of terminal region data by different methods at low compression ratios; (b1)‒(b4) reconstructed images of terminal region data by different methods at high compression ratios; (c) reconstructed images of soma data by different methods at high compression ratios; (d) gray value variation curves corresponding to the dotted lines in (a1)‒(a4); (e) gray value variation curves corresponding to the dotted lines in (b1)‒(b4); (f) gray value variation curves corresponding to the dotted lines in (c1)‒(c4)
    Reconstruction images and comparison of resolution for nerve fiber data by different methods. (a) Reconstruction images; (b) comparison of resolution corresponding to the dotted lines in (a)
    Comparison of the rate-distortion curves fordifferent models
    Tools

    Get Citation

    Copy Citation Text

    Yabo Liu, Xiaoquan Yang, Tao Jiang. High Dynamic Range Image Compression Based on a Multi-Scale Feature Network[J]. Laser & Optoelectronics Progress, 2025, 62(4): 0437007

    Download Citation

    EndNote(RIS)BibTexPlain Text
    Save article for my favorites
    Paper Information

    Category: Digital Image Processing

    Received: Jun. 17, 2024

    Accepted: Jul. 29, 2024

    Published Online: Feb. 10, 2025

    The Author Email:

    DOI:10.3788/LOP241492

    CSTR:32186.14.LOP241492

    Topics