Skip to main navigation Skip to search Skip to main content

SAR-to-Optical Image Translation Based on Generative Artificial Intelligence

  • Hang Chen
  • , Jinyu Wang
  • , Haitao Yang
  • , Zhiliang Li
  • , Zhengjun Liu
  • , Lei Liu*
  • *Corresponding author for this work
  • Space Engineering University
  • Norinco Group Testing and Research Institute
  • School of Physics, Harbin Institute of Technology

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

Abstract

A cross-modal translation method from SAR images to optical images is provided in this chapter. Based on the latest generative artificial intelligence technologies, the interpretability of SAR images is enhanced. The basic principles and evaluation methods of generative adversarial networks and diffusion models are comprehensively elaborated. Two improved S2O algorithms are introduced, and their effectiveness is demonstrated through experiments. In terms of model design, two algorithm architectures, namely the dual-branch cycle-consistent and the diffusion model, are mainly introduced in this chapter.

Original languageEnglish
Title of host publicationStudies in Computational Intelligence
PublisherSpringer Science and Business Media Deutschland GmbH
Pages217-247
Number of pages31
DOIs
StatePublished - 2026
Externally publishedYes

Publication series

NameStudies in Computational Intelligence
Volume1246
ISSN (Print)1860-949X
ISSN (Electronic)1860-9503

Keywords

  • Diffusion model
  • Generative adversarial network
  • Image cross modal translation
  • SAR image augmentation

Fingerprint

Dive into the research topics of 'SAR-to-Optical Image Translation Based on Generative Artificial Intelligence'. Together they form a unique fingerprint.

Cite this