eduzhai > Applied Sciences > Engineering >

Multi-Dimension Fusion Network for Light Field Spatial Super-Resolution using Dynamic Filters

  • king
  • (0) Download
  • 20210507
  • Save

... pages left unread,continue reading

Document pages: 16 pages

Abstract: Light field cameras have been proved to be powerful tools for 3Dreconstruction and virtual reality applications. However, the limitedresolution of light field images brings a lot of difficulties for furtherinformation display and extraction. In this paper, we introduce a novellearning-based framework to improve the spatial resolution of light fields.First, features from different dimensions are parallelly extracted and fusedtogether in our multi-dimension fusion architecture. These features are thenused to generate dynamic filters, which extract subpixel information frommicro-lens images and also implicitly consider the disparity information.Finally, more high-frequency details learned in the residual branch are addedto the upsampled images and the final super-resolved light fields are obtained.Experimental results show that the proposed method uses fewer parameters butachieves better performances than other state-of-the-art methods in variouskinds of datasets. Our reconstructed images also show sharp details anddistinct lines in both sub-aperture images and epipolar plane images.

Please select stars to rate!


0 comments Sign in to leave a comment.

    Data loading, please wait...