EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Bi-Level Meta-Learning for Reliable Remote Sensing Image Registration

Lin Shi, Renzhen Wang, Xiaofeng Zhu, Cong An, Kai Zhao, Jun Shu, Dongfang Yang, Deyu Meng

Remote Sensing · 16.06.2026 · 18 (12) · 2007

DOI: 10.3390/rs18122007 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Remote Sensing, online erschienen am 16.06.2026. Verfasst von Lin Shi und 7 weiteren Autor:innen. Die Arbeit wurde bisher 1-mal zitiert (Crossref) und nennt 40 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: the National Key R&D Program of China, other research grants.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Unmanned aerial vehicle (UAV) visual navigation relies critically on robust image matching between UAV-acquired aerial imagery and pre-existing satellite reference maps. However, extreme cross-domain heterogeneity—encompassing temporal, radiometric, viewpoint, and sensor variations—causes severe performance degradation in existing deep learning-based matchers trained on conventional benchmarks. Furthermore, manual annotation of ground-truth correspondences is prohibitively expensive. This paper proposes a semi-supervised saliency-aware image matching framework with bi-level meta-learning. Our approach comprises two synergistic stages: (1) automated dense correspondence generation via parameterized geometric synthesis, which constructs a large-scale coarse dataset Dc (approximately 50,000 pairs) without dense manual point annotation, serving as the primary training corpus for the feature matching network; (2) expert-validated meta-data curation producing a high-quality meta-dataset Dm (500 pairs) that supervises the training of a Saliency Judgment Network through bi-level meta-optimization, enabling the network to identify and prioritize geometrically reliable correspondences. Experimental results on the proposed RS-Hetero-50K benchmark and cross-domain FuJian-Mountain dataset demonstrate substantial improvements over representative sparse and detector-free matchers, including LoFTR, SuperGlue, and LightGlue. The complete CNN-attention and saliency-aware framework achieves 95.4% matching precision, which is consistent with the best result reported in the experimental section. The plug-and-play experiments further confirm that the proposed saliency module consistently improves representative sparse and detector-free matchers, indicating that the performance gain stems from both stronger feature representation and saliency-guided correspondence selection. The largest terrain-specific gain is observed in gobi scenes, where the AUC@5 px improves by 16.8% relative to the LoFTR baseline, demonstrating improved robustness in weakly textured remote sensing environments.

Förderung

  • the National Key R&D Program of China
  • other research grants

Zitieren

APA

Shi, L., Wang, R., Zhu, X., An, C., Zhao, K., Shu, J., Yang, D., & Meng, D. (2026). Bi-Level Meta-Learning for Reliable Remote Sensing Image Registration. Remote Sensing, 18(12), 2007. https://doi.org/10.3390/rs18122007

Vancouver

Shi L, Wang R, Zhu X, An C, Zhao K, Shu J, et al. Bi-Level Meta-Learning for Reliable Remote Sensing Image Registration. Remote Sensing. 2026;18(12):2007. doi:10.3390/rs18122007

BibTeX
@article{Shi2026,
  author = {Shi, Lin and Wang, Renzhen and Zhu, Xiaofeng and An, Cong and Zhao, Kai and Shu, Jun and Yang, Dongfang and Meng, Deyu},
  title = {Bi-Level Meta-Learning for Reliable Remote Sensing Image Registration},
  journal = {Remote Sensing},
  year = {2026},
  volume = {18},
  number = {12},
  pages = {2007},
  doi = {10.3390/rs18122007}
}
Kostenlos starten →