BGD-Net: An Object–Background Decoupling Network for Oriented Object Detection in Complex Remote Sensing Scenes
Sensors · 17.08.2026 · 26 (16) · 5193
DOI: 10.3390/s26165193 ↗Kurz gefasst
Fachartikel in der Zeitschrift Sensors, online erschienen am 17.08.2026. Verfasst von Jiaxin Xu und 3 weiteren Autor:innen. Beteiligte Einrichtung: College of Information Engineering and Artificial Intelligence, Henan University of Science and Technology, Luoyang 471000, China. Die Arbeit nennt 53 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: National Natural Science Foundation of China, the Major Science and Technology Program of Henan Province, the Central Government Guiding Local Science and Technology Development Fund Program of Henan Province.
Abstract (Originalsprache)
Oriented object detection in high-resolution remote sensing images remains vulnerable to object–background confusion in complex scenes, where visually similar background structures can produce high-confidence false positives. To address this problem, we propose BGD-Net, an object–background decoupling framework that suppresses confusing background responses from the feature, sample, and optimization levels. First, a Background-Decoupled Feature Module (BDFM) separately models object responses and background activations and performs residual feature purification before proposal generation. Second, Hard Background Mining (HBM) identifies high-confidence, low-overlap background RoIs that are most likely to cause false-positive detections and assigns them greater training emphasis. Third, an Object–Background Contrastive Suppression Loss (OBCS Loss) uses the mined hard backgrounds as targeted negative samples to enlarge the representation gap between true objects and confusing backgrounds in the RoI embedding space. Experiments on DOTA-v1.0 show that BGD-Net achieves 82.94% mAP@0.5, improving the Oriented R-CNN baseline by 2.07 percentage points while reducing Hard-FP by 38.48%. The method also achieves 98.47% mAP@0.5 on HRSC2016 and 68.35% mAP@0.5 on DIOR-R, demonstrating effective performance across different remote sensing scenes. Ablation, multi-seed, and threshold-sensitivity experiments further verify the complementary contributions and reproducibility of the proposed components. BGD-Net introduces only a modest increase in model complexity, indicating a favorable balance between detection accuracy and practical efficiency.
Schlagwörter
Stichwörter der Autor:innen
Autor:innen und Institutionen
- College of Information Engineering and Artificial Intelligence, Henan University of Science and Technology, Luoyang 471000, China.
Förderung
- National Natural Science Foundation of China
- the Major Science and Technology Program of Henan Province
- the Central Government Guiding Local Science and Technology Development Fund Program of Henan Province
- the Provincial Science and Technology Research and Development Plan Joint Fund Program of Henan Province
- Major Science and Technology Program of Henan Province
- Central Government Guiding Local Science and Technology Development Fund Program of Henan Province
- Provincial Science and Technology Research and Development Plan Joint Fund Program of Henan Province
Zitieren
APA
Xu, J., Huo, H., Mei, A., & Zhang, C. (2026). BGD-Net: An Object–Background Decoupling Network for Oriented Object Detection in Complex Remote Sensing Scenes. Sensors, 26(16), 5193. https://doi.org/10.3390/s26165193
Vancouver
Xu J, Huo H, Mei A, Zhang C. BGD-Net: An Object–Background Decoupling Network for Oriented Object Detection in Complex Remote Sensing Scenes. Sensors. 2026;26(16):5193. doi:10.3390/s26165193
BibTeX
@article{Xu2026,
author = {Xu, Jiaxin and Huo, Hua and Mei, Aokun and Zhang, Chen},
title = {BGD-Net: An Object–Background Decoupling Network for Oriented Object Detection in Complex Remote Sensing Scenes},
journal = {Sensors},
year = {2026},
volume = {26},
number = {16},
pages = {5193},
doi = {10.3390/s26165193}
}
EUVIMED