2026· IEEE Geoscience and Remote Sensing Letters· Vol 23, pp. 8003005-8003005· 0 citations· 14 references
Abstract
Remote sensing small-object detection (SOD) is challenged by limited pixels, dense distributions, complex backgrounds, and strict computational budgets for uncrewed aerial vehicle (UAV) and edge deployment. To address these coupled requirements, this letter proposes frequency–spatial lightweight adaptive YOLO (FSLA-YOLO), a YOLOv11-based detector following a three-stage design principle: frequency–spatial representation enhancement, lightweight cross-scale recovery, and scale-aware localization. In the representation stage, the high-frequency enhanced wavelet C3k2 (HFEWC3k2) module preserves fine details and structural semantics through frequency decomposition, while the decoupled multiscale spatial–channel attention (DMSCA) module strengthens object-region discrimination under complex backgrounds. In the fusion stage, the frequency-guided directional gated lightweight upsampling (FDG-LU) module efficiently recovers cross-scale details through directional filtering and lightweight channel interaction. In the optimization stage, the scale-aware focal complete intersection over union (SAF-CIoU) loss stabilizes small-box regression using scale-aware constraints and focal compensation. Experiments on RSOD and NWPU-VHR-10 achieve mAP50 values of 98.5% and 92.6%, respectively. Compared with YOLOv11s, FSLA-YOLO requires only 8.4 M parameters and 20.8 GFLOPs, corresponding to reductions of 10.5% and 3.70%, while reaching 145 and 78 FPS, with speed improvements of 16.0% and 11.4% on the two datasets.
Remote sensing small-object detection remains challenging: weak target details are attenuated by repeated downsampling, sparse object responses are difficult to separate from background clutter, and objects vary widely in scale and aspect ratio. We propose FD-YOLO, a compact detector with three complementary components...
Qi-Guang Zhu, Yan Zhu, Xing-Yong Lv et al.· Measurement science and tech...· 0 citations
Object detection in remote sensing imagery (RSI) underpins UAV monitoring, precision agriculture, and disaster assessment, yet is hindered by extreme scale variation, vast homogeneous backgrounds, and small target footprints. Existing lightweight YOLO-based detectors typically improve the backbone attention, downsampli...
Yi-Mo Peng, Xiang-Yu Ge· International Conference on...· 0 citations
Small object detection (SOD) remains a fundamental challenge in computer vision due to the limited spatial resolution of target regions and their susceptibility to background clutter. Three structural bottlenecks collectively constrain detection performance: systematic underrepresentation of fine-grained features in sh...
Yu-Hua Li, Li Zheng, Yun-Rui Wan et al.· IEEE Journal of Selected Top...· 0 citations
This work proposes RAD-YOLO, a YOLOv8s-based small-object detector, and develops an edge-deployable variant named RAD-YOLO-Slim, which provides a practical balance of accuracy, speed and energy efficiency on the RK3576 platform.
Shuai-Jie Nie, Jia-Jian Yang, Xin He et al.· Engineering Research Express· 0 citations
Small object detection (SOD) in optical remote sensing images (RSI) is essential for aerial perception yet remains challenged by the severe degradation of fine-grained spatial features during downsampling and the high sensitivity of Intersection over Union (IoU) metrics to tiny positional shifts. To address these limit...
Bin Xiao, Jun Yan· Applied Sciences· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.