Skip to main navigation Skip to search Skip to main content

Hyperspectral Image Reconstruction Based on Spectral Super-resolution

  • School of Electronics and Information Engineering, Harbin Institute of Technology

Research output: Chapter in Book/Report/Conference proceedingChapterpeer-review

Abstract

To overcome the limitations of HSIs in achieving precise classification and the inability of MSIs to provide rich spectral information, a sub-pixel super-resolution model based on dictionary learning and an adaptive deep-learning-based mapping model are proposed in this part for the collaborative utilization of MSIs and HSIs, and fully leverage the high spectral resolution of HSIs and the large swath width and high spatial resolution of MSIs, ultimately achieve high-spatial-resolution HSIs in large-scale scenes.

Original languageEnglish
Title of host publicationMachine-Learning-Based Hyperspectral Image Processing
Publisherwiley
Pages87-127
Number of pages41
ISBN (Electronic)9781394267880
ISBN (Print)9781394267859
DOIs
StatePublished - 1 Jan 2026
Externally publishedYes

Keywords

  • Collaborative utilization
  • Deep learning
  • Dictionary learning
  • Hyperspectral images
  • Multispectral images
  • Super-resolution

Fingerprint

Dive into the research topics of 'Hyperspectral Image Reconstruction Based on Spectral Super-resolution'. Together they form a unique fingerprint.

Cite this