EUVIMEDEUVIMED entdecken
Crossref

DGFF-Net: a dual-branch gated frequency-aware network for infrared and visible image fusion

Sun Bin, Yuzhang Jiang

Applied Optics · 20.07.2026 · 65 (21) · 7114

DOI: 10.1364/ao.600078 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Applied Optics, online erschienen am 14.07.2026. Verfasst von Sun Bin und 1 weiteren Autor:innen. Die Arbeit nennt 27 Literaturangaben. Gefördert von: Yunnan Province Key Laboratory of Low Light Detection and Intelligent Visual Navigation.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Infrared and visible image fusion aims to combine the thermal saliency of infrared imaging with the rich structural and textural information of visible imaging, providing a more informative representation for optical scene perception and subsequent visual analysis. However, existing fusion methods often suffer from a trade-off between salient target preservation and structural detail retention, particularly in complex scenes where salient targets and background textures are highly intertwined, leading to weak target responses and degraded structural clarity during cross-modal feature aggregation. To address this issue, we propose DGFF-Net, a dual-branch gated frequency-aware network for infrared–visible image fusion. Specifically, a dual-branch encoder is introduced to extract modality-specific features from infrared and visible inputs. A frequency decoupling enhancement module is then designed to decompose fused features into low-frequency and high-frequency components, enabling differentiated preservation of thermal targets and fine textures. In addition, a region-guided gated fusion module is developed to adaptively regulate cross-modal contributions in spatially heterogeneous regions, so that salient targets and background details can be jointly preserved in a content-aware manner. An edge-structure refinement module is further incorporated in the reconstruction stage to enhance boundary clarity and suppress reconstruction artifacts. Extensive experiments on public benchmarks demonstrate that the proposed method achieves superior performance in both visual quality and quantitative metrics, producing fused images with enhanced target saliency, clearer structural details, and improved perceptual quality. These results indicate that DGFF-Net provides an effective and robust approach for high-quality infrared–visible image fusion in complex imaging scenes.

Förderung

  • Yunnan Province Key Laboratory of Low Light Detection and Intelligent Visual Navigation

Zitieren

APA

Bin, S., & Jiang, Y. (2026). DGFF-Net: a dual-branch gated frequency-aware network for infrared and visible image fusion. Applied Optics, 65(21), 7114. https://doi.org/10.1364/ao.600078

Vancouver

Bin S, Jiang Y. DGFF-Net: a dual-branch gated frequency-aware network for infrared and visible image fusion. Applied Optics. 2026;65(21):7114. doi:10.1364/ao.600078

BibTeX
@article{Bin2026,
  author = {Bin, Sun and Jiang, Yuzhang},
  title = {DGFF-Net: a dual-branch gated frequency-aware network for infrared and visible image fusion},
  journal = {Applied Optics},
  year = {2026},
  volume = {65},
  number = {21},
  pages = {7114},
  doi = {10.1364/ao.600078}
}
Kostenlos starten →