- Article
- Open access
- Published:
Scientific Reports (2026) Cite this article
We are providing an unedited version of this manuscript to give early access to its findings. Before final publication, the manuscript will undergo further editing. Please note there may be errors present which affect the content, and all legal disclaimers apply.
Abstract
Small object detection in UAV aerial images is challenged by significant scale variation, weak edge information, and complex backgrounds. To address these issues, an efficient model termed BMDnet is proposed based on YOLOv11n. The proposed method introduces a collaborative optimization framework from three aspects: feature fusion, feature enhancement, and detection optimization. Specifically, a bidirectional multi-scale feature fusion network (BIMAFPN) is designed to enhance cross-layer feature interaction; a multi-scale edge information enhancement module (MEIE) is introduced to strengthen fine-grained structural representation; and a dynamic attention head (DAH) is employed to adaptively refine feature representations via scale-aware, spatial-aware, and task-aware mechanisms. Experimental results on the VisDrone2019 dataset demonstrate that BMDnet improves mAP50 and mAP50–95 by 3.5 and 2.2 percentage points, respectively, with only a moderate increase in computational cost. Additional experiments on the HIT-UAV dataset further verify the generalization capability of the proposed model. The results indicate that BMDnet achieves an effective balance between detection accuracy and computational efficiency.
Subjects
Funding
This work was supported by the University Natural Science Research Project of Anhui Province, the Grant ID is 2024AH050213.
Ethics declarations
Competing interests
The authors confirm that they have no financial or nonfinancial competing interests with any companies, organizations, or individuals while preparing and submitting the manuscript.
Additional information
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.
Reprints and permissions
About this article
Cite this article
Li, F., Wang, M., Yu, C. et al. BMDnet: an efficient UAV small object detection model via multi-scale feature enhancement and dynamic attention. Sci Rep (2026). https://doi.org/10.1038/s41598-026-65686-z
Download citation
Received:
Accepted:
Published:
DOI: https://doi.org/10.1038/s41598-026-65686-z