EUVIMEDEUVIMED entdecken
Open Access · CC BY-NC-ND 4.0

YOLO-MRAF: A multi-scale reinforced attention fusion network for waterbody segmentation

Tao Zhou, Nan Lian, Zeming Yin, Ji Liu, Hui Meng, Chuan Zhang, Yao Zhang, Zhigang Li

Journal of King Saud University Computer and Information Sciences · 2026-09 · 38 (7)

DOI: 10.1007/s44443-026-01032-7 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Journal of King Saud University Computer and Information Sciences, online erschienen am 18.08.2026. Verfasst von Tao Zhou und 7 weiteren Autor:innen. Die Arbeit nennt 36 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC-ND 4.0). Gefördert von: Science and Technology Program of XPCC, Bing-tuan Science and Technology Public Relations Project, Project of Xinjiang Uygur Autonomous Region.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Accurate waterbody segmentation is important for environmental monitoring, water surface analysis, and water resource management. However, waterbody extraction from complex remote sensing images remains challenging due to irregular water morphology, shadows, illumination variations, reflection interference, and visually similar backgrounds. In addition to pixel-level mask extraction, practical waterbody monitoring also requires region-level localization and prediction confidence for separated water areas. To address these requirements, this paper proposes YOLO-MRAF, a YOLO-based instance-aware waterbody segmentation framework. The proposed model integrates three modules: a Multi-Scale Reinforced Feature (MSRF) module for preserving spatial details during hierarchical feature extraction, a Direction-Sensitive Attention Fusion (DSA-Fuse) module for enhancing horizontal and vertical spatial dependencies, and a Feature Recalibration Module (FRM) for strengthening feature interaction through additive, multiplicative, and transposed-convolution coupling. Experiments on the Water Detection & Segmentation Dataset under three random seeds show that YOLO-MRAF improves mask-level mAP50 and mAP50-95 from 0.594 and 0.398 to 0.644 and 0.448, respectively, compared with YOLOv11n-seg. The results also reveal a Precision–Recall trade-off, as YOLO-MRAF improves mask-level Precision but shows lower box-level Recall. Comparisons with semantic segmentation models and direct external evaluations on RIWA and LuFI-RiverSnap further demonstrate its competitive pixel-level performance and improved external mIoU and Dice. These results indicate that YOLO-MRAF is a practical YOLO-based solution for instance-aware waterbody segmentation.

Förderung

  • Science and Technology Program of XPCC
  • Bing-tuan Science and Technology Public Relations Project
  • Project of Xinjiang Uygur Autonomous Region

Zitieren

APA

Zhou, T., Lian, N., Yin, Z., Liu, J., Meng, H., Zhang, C., Zhang, Y., & Li, Z. (2026). YOLO-MRAF: A multi-scale reinforced attention fusion network for waterbody segmentation. Journal of King Saud University Computer and Information Sciences, 38(7). https://doi.org/10.1007/s44443-026-01032-7

Vancouver

Zhou T, Lian N, Yin Z, Liu J, Meng H, Zhang C, et al. YOLO-MRAF: A multi-scale reinforced attention fusion network for waterbody segmentation. Journal of King Saud University Computer and Information Sciences. 2026;38(7). doi:10.1007/s44443-026-01032-7

BibTeX
@article{Zhou2026,
  author = {Zhou, Tao and Lian, Nan and Yin, Zeming and Liu, Ji and Meng, Hui and Zhang, Chuan and Zhang, Yao and Li, Zhigang},
  title = {YOLO-MRAF: A multi-scale reinforced attention fusion network for waterbody segmentation},
  journal = {Journal of King Saud University Computer and Information Sciences},
  year = {2026},
  volume = {38},
  number = {7},
  doi = {10.1007/s44443-026-01032-7}
}
Kostenlos starten →