EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

FEDM: Feature-Encoding Diffusion Model for Large-Scale SAR Image Inpainting

Junyu Yang, Wenzheng Wang, Chenwei Deng

Remote Sensing · 01.07.2026 · 18 (13) · 2116

DOI: 10.3390/rs18132116 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Remote Sensing, online erschienen am 01.07.2026. Verfasst von Junyu Yang und 2 weiteren Autor:innen. Die Arbeit nennt 22 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: National Natural Science Foundation of China.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

With the wide application of generative models in the field of SAR image inpainting, inadequate reconstruction quality of scattering characteristics and insufficient global coherence of semantic logic remain the core challenges of such tasks. To address these issues, this paper proposes a Feature-Encoding Diffusion Model (FEDM). Guided by local valid regions, the proposed model accurately learns the microwave backscattering distribution law of ground features through a SAR-specific Variational Auto-Encoder (SAR-VAE), thus improving the reconstruction accuracy of backscattering statistics. Meanwhile, it integrates semantic embedding and cross-attention mechanism to strengthen the semantic constraints of SAR scenes, ensuring the logical rationality of the ground feature layout. With progressive diffusion generation and sliding window strategy, the model achieves high-quality reconstruction with coherent semantics and consistent global spatial structure for large-scale missing regions. Experiments on public datasets including OSdataset, SEN1-2, SRSDD-v1.0 and MRSSC show that the proposed method achieves excellent performance in terms of scattering characteristic reconstruction quality and globally coherent generation of semantic logic, and realizes high-quality SAR image inpainting.

Förderung

  • National Natural Science Foundation of China

Zitieren

APA

Yang, J., Wang, W., & Deng, C. (2026). FEDM: Feature-Encoding Diffusion Model for Large-Scale SAR Image Inpainting. Remote Sensing, 18(13), 2116. https://doi.org/10.3390/rs18132116

Vancouver

Yang J, Wang W, Deng C. FEDM: Feature-Encoding Diffusion Model for Large-Scale SAR Image Inpainting. Remote Sensing. 2026;18(13):2116. doi:10.3390/rs18132116

BibTeX
@article{Yang2026,
  author = {Yang, Junyu and Wang, Wenzheng and Deng, Chenwei},
  title = {FEDM: Feature-Encoding Diffusion Model for Large-Scale SAR Image Inpainting},
  journal = {Remote Sensing},
  year = {2026},
  volume = {18},
  number = {13},
  pages = {2116},
  doi = {10.3390/rs18132116}
}
Kostenlos starten →