Jul 2026· Digital Signal and Computer Communications· Vol 14294, pp. 142940Y - 142940Y-6· 0 citations· 10 references
Engineering
TL;DR
A Multi-scale Local Contrast Module (MLCM) that utilizes dilated convolutions to mimic the Human Visual System, significantly enhancing rotor edge features for better fine-grained discrimination is designed.
Abstract
Infrared UAV detection and fine-grained recognition are pivotal for low-altitude security but suffer from two primary issues: the lack of texture in infrared imagery, which hinders fine-grained classification, and extreme long-tailed data distributions, which leads to poor performance on rare classes. We propose LC-YOLO, a real-time framework integrating local contrast enhancement and category balancing. To resolve texture deficiency, we design a Multi-scale Local Contrast Module (MLCM) that utilizes dilated convolutions to mimic the Human Visual System, significantly enhancing rotor edge features for better fine-grained discrimination. To mitigate data imbalance, we introduce a Category-Specific Mosaic (CS-Mosaic) strategy that enforces tail-class oversampling during data loading, preventing model overfitting to head classes at the source. Experiments on a multi-source heterogeneous dataset demonstrate that LC-YOLO substantially improves overall mAP and tail-class recall (e.g., single-rotors and fixed-wings) while maintaining real-time efficiency.
UAV infrared target detection in military reconnaissance and disaster response faces significant challenges, including low-contrast imagery with sparse textures, scale variations under dynamic flight perspectives, and occlusion-induced detection degradation. To address these issues, we propose RVGM-YOLO, a Hierarchical State Enhancement Network based on YOLOv11. The framework integrates three key innovations: a state-recursive spatial context module (SRSCM) enhances local features in shallow layers via recursive structures and strengthens global context in deep layers using state-space models; Haar wavelet downsampling (HWD) preserves small-target edge contours and texture details through multiresolution analysis; and a multiscale SEAM detection head (MultiSEAM) leverages multihead attention for robustness under occlusion. Experiments on HIT-UAV demonstrate RVGM-YOLO achieves 91.3% precision, 86.7% recall, 88.5% $F1$ -score, and 62.0% mAP@50:95, outperforming baselines while maintaining real-time efficiency. The code is available at: https://github.com/CarlWang-13/RVGM-YOLO
Yahan Wang, Xu Ren, Fei Ming et al.· IEEE Geoscience and Remote S...· 0 citations
Crop targets in UAV aerial images are typically characterized by small scale, dense distribution, severe mutual occlusion, and complex backgrounds, which often lead to low detection accuracy and large counting errors for existing deep learning models. To address these issues, this study proposes an improved YOLOv12-based crop detection and counting model, named YOLO-DC. By introducing an attention mechanism (LGCB-AM) and a multi-scale detection head (MS-DH), the proposed model effectively enhances local texture extraction, global modeling, foreground–background contrast, and boundary perception for dense small objects. Subsequently, a series of comparative experiments, ablation studies, and transfer experiments were conducted on the wheat and rice datasets. The results show that YOLO-DC achieves a favorable balance among detection accuracy, counting error, and model efficiency and overall outperforms the other comparison models. Ablation studies further verify the effectiveness of the proposed design, showing that LGCB-AM is the key contributor to the performance improvement, while the boundary branch and repulsion branch play critical roles in dense-target discrimination. In addition, an appropriate module insertion strategy can effectively balance high-level semantic enhancement and feature fusion stability. Transfer experiments demonstrate that pretraining on the wheat dataset and fine-tuning on the rice dataset significantly outperform training from scratch, indicating strong cross-crop transfer potential. Overall, the proposed YOLO-DC provides an effective solution for high-precision crop detection and counting in agricultural scenarios.
Haotian Bai, Lei Liu, Haocheng Kong et al.· Remote Sensing· 0 citations
Infrared (IR) video target detection is important for long-range target perception in anti-UAV systems under complex lighting and background conditions. However, long-range tiny unmanned aerial vehicles (UAVs) usually occupy only a few pixels, exhibit weak thermal contrast, and are easily disturbed by cluttered backgrounds such as clouds, buildings, vegetation, feature edges, and thermal noise. Although YOLO-style detectors provide high real-time performance, their feature flow is prone to weak target response attenuation during downsampling, contextual ambiguity in deep feature representation, and background clutter propagation during cross-scale fusion. To address these degradation problems, this paper proposes FDR-YOLO, a feature-degradation-aware feature-flow reconstruction network based on YOLOv26. Specifically, LAE-based response-preserving downsampling (LAE-RPD) preserves weak but discriminative target responses during spatial compression; lightweight U-shaped dilated context aggregation (UCDC-Lite) enhances deep contextual discrimination between tiny UAV targets and cluttered backgrounds; and high-frequency prior-guided semantic injection fusion (HPG-SIF) uses shallow high-frequency priors to constrain the injection of deep semantic features. Experiments on multiple datasets show that FDR-YOLO improves detection accuracy while retaining lightweight and low-latency characteristics. On the Anti-UAV dataset, FDR-YOLO improves mAP50 and mAP50–95 by 3.0 and 3.3 percentage points, respectively, over YOLOv26s. Additional experiments on InfraredUAV and the RGB-based UAVSwarm dataset demonstrate the applicability of the proposed design to another infrared benchmark and to visible-light UAV detection under dataset-specific training.
Unmanned Aerial Vehicle (UAV) aerial photography is extensively utilized in security, traffic monitoring, and disaster rescue. However, UAV-captured images present significant challenges, including small target scales, dense distribution, and complex backgrounds. While conventional object detection algorithms like the YOLO series have made progress, they often struggle to balance accuracy and real-time performance in these resource-constrained environments. To address these issues, we propose SSM-YOLO11s, a lightweight model optimized for small object detection in aerial imagery. Our approach first introduces the Sitou module, which employs a deep-channel compression and shallow feature retention strategy with a secondary fusion branch to reduce parameters by 50% while enhancing fine-grained feature utilization. Furthermore, the lightweight SNGSConvE module is designed by integrating SNI, GSConvE, and CSPOmniKernel to mitigate feature misalignment and strengthen capture capabilities. Finally, a Multi-Scale Edge Enhancement (MSEE) module is constructed to fuse edge details across multiple scales, improving target discriminability. Experimental results on the VisDrone2019 dataset demonstrate that SSM-YOLO11s achieves a superior balance between precision and efficiency compared to state-of-the-art models.
Junfu Chen, Xi Zhao· International Conference on...· 0 citations
Unmanned Aerial Vehicle (UAV)-based small object detection is a challenging computer vision task. It is constrained by two primary factors: UAV platforms have limited onboard computational resources, and high-altitude objects often have weak features that are easily overwhelmed by complex backgrounds. To address these challenges, we propose LUIM-YOLO. First, a Lightweight Multi-Scale Feature Enhancement (LMSFE) module integrates parallel multi-scale convolutions with attention to strengthen small and low-contrast object feature extraction. Second, an Adaptive Multi-Scale Bottleneck (AMSB) module enhances key semantic features of small objects and spatial correlation of medium-scale objects. Third, an Enhanced Cross-layer Compensation Feature Pyramid Network (ECC-FPN) constructs cross-level interaction pathways to improve small object position and scale perception. Experimental results on VisDrone2019 show that compared with YOLOv8n, LUIM-YOLO reduces parameters by 57% and improves mAP@50 by 12.9%. Additional full-validation-set PyTorch inference tests on NVIDIA Jetson Orin show that LUIM-YOLO achieves 88.19 ms/image in FP32, indicating a parameter-efficient accuracy-oriented design with edge deployment potential.
Junjie Li, Yisheng Wang, Bo Zhang· Applied Sciences· 1 citation
Single-modal object detection methods often suffer from low accuracy, false detections, and missed detections in complex scenes such as low-light and foggy environments. To address these problems, this paper proposes SCD-YOLO, a visible-infrared dual-modal object detection network based on YOLOv11n. First, a spatial-channel decoupled fusion module, named SCD-Fusion, is designed to enhance cross-modal feature interaction. This module consists of a cross-directional spatial enhancement module (CDSE) and a decoupled channel fusion module (BCDF). CDSE models long-range spatial dependencies along the height and width directions, while BCDF captures differential and common channel information between visible and infrared features. In addition, a cascaded adaptive spatial detection head, CAS-Head, is introduced to progressively fuse multi-scale features and adaptively weight four-scale features. Experimental results on the FLIR dataset show that SCD-YOLO achieves 75.23% mAP@0.5 and 40.89% mAP@0.5:0.95, which are 4.13% and 3.84% higher than those of the baseline YOLOv11n model, respectively.
Tao Wang, Xu Ma, Jiaben Liang et al.· 2026 IEEE 27th China Confere...· 0 citations