Search Results (325)

The precise identification of minute, high-velocity entities within unconstrained visual fields represents a significant hurdle in computational perception. This difficulty primarily arises from the geometric degradation stemming from scale volatility, motion-induced asymmetry, and heterogeneous background clutter. To mitigate the critical deficit of high-fidelity benchmarks for dynamic micro-targets, we present Soccer-Wild. This comprehensive dataset is characterized by the extreme visual complexity of microscopic objects in diverse ecological settings. Built upon this empirical foundation, we introduce GOAL (Global Object Alignment for Localization). This novel computational paradigm is designed to enhance the weak features of tiny targets by integrating frequency-domain filtering, dynamic feature routing, and entropy-guided probabilistic modeling. The GOAL framework rigorously preserves spatial-structural equilibrium and information fidelity through three synergetic mechanisms: (1) Spectral Purification: We implement a Frequency-aware Spectral Gating approach that operates in the Fourier manifold, suppressing stochastic noise to accentuate the spectral signatures of the targets; (2) Geometric Adaptation: A Multi-Granularity Mixture of Experts (MG-MoE) is formulated with heterogeneous receptive fields to dynamically rectify anisotropic distortions caused by kinetic blurring. This adaptive routing ensures cross-state representation consistency; (3) Information Recovery: We propose Information-Guided Gaussian Distribution Estimation (IGDE), which utilizes information entropy to conceptualize target coordinates as radially symmetric probability densities. This facilitates the implicit recovery of latent signals typically discarded by rigid deterministic regression. Empirical validations on the Soccer-Wild and VisDrone2019 benchmarks reveal that the proposed methodology yields substantial gains in precision. Specifically, our model achieves 40.0% and 40.4% AP (Average Precision), respectively, establishing a new state-of-the-art for localizing highly dynamic, micro-scale objects. Full article

Weed identification and quantification are processes that are usually manual, subjective, and error-prone. Weeds compete with crops for nutrients, minerals, physical space, sunlight, and water. Thus, weed identification is a crucial component of precision agriculture for autonomous removal and site-specific treatments, efficient weed control, and sustainability. Convolutional Neural Networks (CNNs) are very common in weed identification. This work implemented CNN models for semantic segmentation based on the U-Net architecture for automatically segmenting and quantifying weeds in potato crops using RGB images acquired by a drone at 9–10 m height, flying at 1 m/s. Remote sensing images are affected by factors that degrade image quality and the model’s accuracy. Five U-Net variants were evaluated: the original U-Net, Residual U-Net, Double U-Net, Modified U-Net, and AU-Net. The models were trained using the TensorFlow/Keras frameworks on Google Colab Pro+, following the Knowledge Discovery in Databases (KDD) methodology for image analysis. Each model was trained using a diverse custom dataset in uncontrolled environments, considering six classes: background, Broadleaf dock (Rumex obtusifolius), Dandelion (Taraxacum officinale), Kikuyu grass (Cenchrus clandestinum), other weed species, and the crop potato (Solanum tuberosum L.). The models’ segmentation was widely assessed using Mean Dice Coefficient, Mean IoU, and Dice Loss metrics. The results showed that the Residual U-Net model performed the best in multi-class segmentation, achieving a Mean IoU of 0.8021, a performance comparable to or superior to that reported by other authors. Additionally, a Student’s t-test was applied to complement the data analysis, suggesting that the model is reliable for weed quantification. Full article

Multi-Object Tracking (MOT) is a fundamental research topic in the field of computer vision, with broad application potential in unmanned aerial vehicle (UAV) videos. However, existing methods still face significant challenges in detection discriminability and identity association stability due to the small scale and weak appearance of objects under aerial viewpoints, as well as complex background interference. To address these issues, we propose an Edge-Graph Enhanced Network (EGEN) for UAV aerial MOT, aiming to improve the performance of small object detection (SOD) and tracking in complex scenes. The framework follows a one-step tracking paradigm and consists of three main components: object detection, embedding feature extraction, and data association. In the detection stage, we design an Edge-Guided Gaussian Enhancement Module (EGGEM), which models edge relationships between objects and backgrounds from a global perspective and selectively enhances Gaussian features guided by edge information, thereby strengthening key structural features of small objects while suppressing background interference. In the embedding feature extraction stage, we develop a Graph-Guided Embedding Enhancement Module (GGEEM), which explicitly represents re-identification (ReID) embeddings as a graph structure and jointly models nodes and their neighborhood relationships to fully capture inter-object associations and enhance embedding discriminability. In the data association stage, we introduce a hierarchical two-stage association strategy to match objects with different confidence levels separately, improving tracking stability and robustness. Extensive experiments on the VisDrone, UAVDT, and self-constructed WildDrone datasets demonstrate that the proposed method significantly outperforms state-of-the-art approaches in both SOD and MOT, demonstrating strong generalization and practical applicability. Full article

Identifying the plant species comprising a pasture, among other aspects, is crucial for assessing its nutritional value for grazing animals and facilitating its effective management. Traditionally, it requires labor-intensive visual inspection. Artificial Intelligence (AI) offers a solution for automatic classification, yet robust datasets for training such models in natural, uncontrolled environments are scarce. This data descriptor presents a dataset of 741 images collected in pasture lands in the Centre of Portugal using standard cameras at a height of 50 cm. A semi-automated annotation pipeline was employed, utilizing a Faster R-CNN model followed by manual verification and refinement. The dataset contains 1744 annotations across four categories: ‘Shrubs’, ‘Grasses’, ‘Legumes’, and ‘Others’. It includes diverse morphological variations and captures real-world challenges such as occlusion and lighting variability. This dataset serves as a benchmark for training object detection models in agricultural settings, facilitating the development of automated monitoring systems for precision agriculture. Such a mechanism could be incorporated into a mobile application, mounted on a drone, or embedded in an animal-worn device, enabling automated sampling and identification of the plant composition within a pasture. Full article

Road safety depends heavily on the timely identification and repair of potholes; however, detecting potholes is challenging due to various lighting and weather conditions. This work presents an attention-enhanced object detection framework for aerial pothole detection design that relies on a pre-trained backbone, YOLOv8, and a custom feature-extraction network, the Feature Pyramid Network (FPN). An enhanced detection head is used to make the model aware of discriminative areas in space to get accurate localization of a pothole to overcome the major limitations of the standard YOLOv8 used in aerial road inspection, irrespective of the road surface. The underlying architecture incorporates a purpose-built data layer and a preprocessing engine that can accommodate scenarios such as seasonal changes and bad weather. To further enhance learning dynamics, a customized loss function and a new optimizer framework are incorporated to improve convergence towards overall detection reliability. Specifically, a custom differential optimizer that uses layer-wise adaptive learning rates and momentum-based gradient updates to help suppress false positives and accelerate convergence. Conversely, the IoU-based personal loss function, combined with real-time validation, stabilizes training across a range of road conditions. A major feature of the proposed system is its ability to process aerial imagery from unmanned drone platforms. Empirical analysis proves a good result: an average precision of 0.980 with the IoU of 0.5 and an F1-score of 0.97 with a confidence threshold of 0.30. Precision is high (0.97 at the 90-percent confidence level). These metrics show how well the model will be able to balance false positives and false negatives—a critical need in a safety-critical deployment. The results make the framework a potential, scalable, and reliable candidate for integrating smart transportation systems and autonomous vehicle navigation. Full article

Apple snail infestation poses a persistent threat to rice production in open-field environments, where long-term coexistence with this species is unavoidable. This study presents a drone-based precision control approach that integrates high-resolution micro-topographic mapping with site-specific pesticide application. A lightweight mapping unmanned aerial vehicle was deployed to produce centimeter-level microtopographic data across paddy fields, facilitating the identification of deep-water areas preferred by apple snails. From these elevation-derived water risk patterns, prescription maps were generated to guide downstream management decisions, and agricultural drones equipped for granular application subsequently performed targeted pesticide delivery only in these high-risk areas. Over 2 years of field experiments, the proposed method achieved rice yields comparable to those under conventional management while reducing pesticide use by 44.1–63.0%, with lower estimated crop damage in regions with high apple snail occurrence. Designed with robustness and scalability in mind, the system demonstrated considerable potential for practical implementation in general farming households and broader applications in precision pest management. Full article

The spotted lanternfly (SLF, Lycorma delicatula) is an invasive planthopper causing severe agricultural and environmental damage in 20 U.S. states. SLF control remains constrained by (1) overreliance on broad-spectrum pesticides that harm nearby ecosystems, (2) inefficiency and ecological risk of alternative methods, and (3) underutilization of SLF behavioral traits and artificial intelligence (AI) in IPM. This study introduces AI-LyD, an AI-driven IPM framework integrating behavioral ecology, predictive modeling, image-based detection, and low-cost physical controls. Incorporating SLF behavioral constraints, including cold-exposure requirements for egg hatching, into ecological models improved prediction accuracy (AUC = 0.821, Sensitivity = 0.888, Kappa = 0.642) and reconstructed SLF distributions consistent with current proliferation trends. A YOLO-based detection model leveraging SLF clustering behavior improved identification accuracy from 84% to 96% and reduced false positives from 42% to 8% in real-world drone-collected imagery. Exploiting SLF crawling, jumping, and hydrophobic behaviors, the novel Aquabex water-moat device with an optimized 60° opening trapped 85% of Stage I–IV nymphs and reduced adult invasions by 67%, at an estimated cost below USD $0.50 per unit. Field deployments across four locations in Hunterdon County, New Jersey, achieved a 91% population reduction (95% CI: 90.1–92.0%). Together, these results establish AI-LyD as the first operational, scalable SLF IPM system, and this paradigm can be applied to controlling other invasive species. Full article

Unmanned aerial vehicle (UAV) applications are growing fast in different sectors, such as agricultural, commercial, academic, leisure, and health fields. However, drones pose a significant threat to public safety due to their ability to transmit information, particularly when used in an unauthorized or malicious manner. In fact, in order to protect citizens’ privacy and prevent accidents in high-traffic areas due to poorly controlled flights, no-fly zones for drones have been established in the legislation of a number of countries. Most common UAV detection techniques are based on radio frequencies, which identify drones and their models by monitoring radio frequency signals. However, differentiating between multiple UAVs of the same model is their main limitation. This article fills this gap by proposing a method for physically tracking the communication frames of a registered UAV in the presence of another UAV of the same model. A measurement campaign was conducted to collect real-world RF communication signals from two DJI MAVIC 2 Zoom, two DJI Air2S, and two DJI Phantom drones. This measurement was performed inside and outside an anechoic chamber in order to study the UAV’s communication without any interference and in the presence of other communications. Through detailed statistical analysis, we characterized features such as communication duration, time intervals between communications, signal strength, and patterns in communication timing sequences. Our analysis revealed unique, identifiable patterns for each UAV, even within identical models. Based on these results, we developed an automated system that links communication frames to the corresponding registered drones. The proposed method fills gaps in drone detection and surveillance models, providing valuable information for applications in the fields of security and airspace management. This research lays the foundation for drone identification solutions, thereby addressing a major limitation of current detection technologies. Full article

Accurate spatial information on individual tree locations is essential for precision forestry, the integration of field and remote sensing data, and tree-level forest analyses. This study compared the positional accuracy and tree identification performance of four tree-mapping approaches: legacy paper maps, a pseudolite-based field positioning system (TerraHärp), drone-based laser scanning, and mobile laser scanning (MLS). The analysis was conducted in five long-term experimental forest sites in Lithuania, comprising pine- and spruce-dominated stands with varying stand densities. Tree locations derived from legacy maps and the TerraHärp system were compared to assess systematic and random positional discrepancies. TerraHärp-derived tree positions were subsequently used as a reference to evaluate the laser scanning-based methods. Positional accuracy was assessed using Hotelling’s T² test, root-mean-square error, and the National Standard for Spatial Data Accuracy (NSSDA), while spatial autocorrelation of deviations was examined using Moran’s I. The results indicated that discrepancies between TerraHärp and legacy maps were dominated by systematic horizontal shifts in the historical maps, whereas random positional variability was relatively small and consistent across stand types. Drone-based laser scanning showed a strong dependence of tree identification accuracy on stand density and mean tree diameter. Overall, CHM-based segmentation yielded more accurate tree identification than 3D point cloud segmentation, with mean F1-scores of 0.78 and 0.72, respectively. Positional accuracy varied by method, with the largest errors from CHM apexes and highest 3D point cloud points (mean NSSDA ≈ 1.8–2.0 m), improved accuracy using the lowest 3D cluster points (1.45–1.72 m), and the highest accuracy achieved using mobile laser scanning (mean NSSDA 0.76–0.90 m; >95% of trees within 1 m). These results demonstrate that pseudolite-based field mapping provides a reliable reference for high-precision tree location and for integrating field and laser scanning data in managed conifer stands. Full article

In recent years, unmanned aerial vehicle (UAV) technology has advanced rapidly, leading to its widespread deployment. However, this proliferation has been accompanied by a rise in unauthorized “black flight”, which poses a series of security risks to low-altitude airspace. Therefore, it is imperative to develop effective drone detection and identification techniques for airspace security management. This paper presents a radio frequency (RF)-based drone recognition method via variational mode decomposition (VMD) and multi-domain feature fusion. First, the collected RF signals exchanged between drones and their controllers are preprocessed using VMD. Subsequently, a multi-domain feature extraction method is introduced, which extracts time-domain, frequency-domain and time–frequency-domain features from the modes after VMD. To reduce feature dimensionality, a two-stage feature selection scheme based on ReliefF is then proposed. Finally, a support vector machine (SVM) is constructed for UAV classification. Experimental results on the open-source CardRF dataset show that the proposed method achieves superior performance compared to existing schemes, with an average identification accuracy of over 74.7% at SNRs greater than −10 dB. Full article

Uncovering latent structures from complex, degraded data is a central challenge in modern unsupervised learning, with critical implications for downstream tasks. This principle is exemplified in the domain of aerial imagery, where the quality of images captured by drones is often compromised by complex, flight-induced degradations, thereby raising the information entropy and obscuring essential semantic patterns. Conventional super-resolution methods, trained on generic data, fail to restore these unique artifacts, thereby limiting their effectiveness for vessel identification, a task that fundamentally relies on clear pattern recognition. To bridge this gap, we introduce a novel adaptive super-resolution framework for ship images captured by drones. The approach integrates a static stage for foundational feature extraction and a dynamic stage for adaptive scene reconstruction, enabling robust performance in complex aerial environments. Furthermore, to ensure the super-resolution model’s generalizability and effectiveness, we optimize the design of degradation methods based on the characteristics of drone aerial images and construct a high-resolution dataset of ship images captured by drones. Extensive experiments demonstrate that our method surpasses existing state-of-the-art algorithms, confirming the efficacy of our proposed model and dataset. Full article

The dairy sheep, vital to the Mediterranean economy, struggles to balance productivity, sustainability, and animal welfare, especially in extensive, small-scale systems. Precision livestock farming (PLF) technologies offer new opportunities by enabling continuous, non-invasive, and data-driven monitoring across diverse farming conditions. Despite rapid progress in sensors, computer vision, wearable devices, and artificial intelligence (AI), a comprehensive synthesis focused on dairy sheep remains limited. This review provides an updated overview of PLF applications in dairy sheep farming, based on a literature review. The 2018–2025 timeframe was chosen to capture recent advances in Internet of Things (IoT), AI, and sensor technologies that have achieved practical relevance only in recent years. The review identifies core technological domains such as automated weight and body condition monitoring, biometric identification, wearable and IoT-based sensors, localization systems, behavioral and thermal monitoring, virtual fencing, drone-assisted herding, and advanced decision-support tools. Innovations including lightweight deep-learning models, multimodal sensing frameworks, and digital twins highlight the growing potential for scalable, real-time applications. While technological progress is substantial, practical adoption is hindered by economic, technical, interoperability, and ethical barriers. This review consolidates current evidence and identifies future priorities to guide the development of integrated, welfare-focused PLF solutions for dairy sheep farming. Full article

Unmanned Aircraft Systems (UASs) offer a promising future for aviation operations, even though it suffers larger cyber-related challenges. As such, cyber-resilience becomes a core property for drones’ operations. This paper presents a systematic meta-review of the scientific literature on Unmanned Aircraft Systems cyber-resilience, starting from 28 literature reviews and surveys in the field. This study examines three areas: the typologies of cyber threats being investigated, the cyber-resilience aspects and functions, and how proposed mitigation strategies align with and support these resilience functions. Overall, 69 cyber threats were identified, where Global Positioning System (GPS) spoofing and jamming were the most frequent ones, underscoring the vulnerability of GPS-based navigation systems in UAS. In terms of cyber-resilience functions, the largest focus remains on the identification, protection, and detection of cyber threats, while limited attention emerges to incident handling and post-event recovery. This is confirmed by the higher frequency of preventive, rather than recovery-oriented, mitigation strategies. Overall, the findings point towards a still limited cyber-resilience implementation for Unmanned Aircraft Systems, witnessing the need for more systemic efforts to guarantee truly resilient UAS operations. Full article

As the threat from malicious UAVs continues to intensify, accurate identification of individual UAVs has become a critical challenge in regulatory and security domains. Existing single-signal analysis methods suffer from limited recognition accuracy. To address this issue, this paper proposes a high-precision individual identification method for UAVs based on FFS-SPWVD and DIR-YOLOv11. The proposed method first employs a frame-by-frame search strategy combined with the smoothing pseudo-Wigner–Ville distribution (SPWVD) algorithm to obtain effective time–frequency feature representations of flight control signals. Building on this foundation, the YOLOv11n network is adopted as the baseline architecture. To enhance the extraction of time–frequency texture features from UAV signals in complex environments, a Multi-Branch Auxiliary Multi-Scale Fusion Network is incorporated into the neck network. Meanwhile, partial space–frequency selective convolutions are introduced into selected C3k2 modules to alleviate the increased computational burden caused by architectural modifications and to reduce the overall number of model parameters. Experimental results on the public DroneRFb-DIR dataset demonstrate that the proposed method effectively extracts flight control frames and performs high-resolution time–frequency analysis. In individual UAV identification tasks, the proposed approach achieves 96.17% accuracy, 97.82% mAP50, and 95.29% recall, outperforming YOLOv11, YOLOv12, and YOLOv13. This study demonstrates that the proposed method achieves both high accuracy and computational efficiency in individual UAV recognition, providing a practical technical solution for whitelist identification and group size estimation in application scenarios such as border patrol, traffic control, and large-scale events. Full article

Hilly and mountainous areas are important agricultural production regions globally. Their dramatic topography, dense fruit tree planting, and steep slopes severely restrict the application of traditional plant protection machinery. Pest and disease control has long relied on manual spraying, resulting in high labor intensity, low efficiency, and pesticide utilization rates of less than 30%. Plant protection UAVs, with their advantages of flexibility, high efficiency, and precise application, provide a feasible technical approach for plant protection operations in hilly and mountainous areas. However, steep slopes and dense orchard environments place higher demands on key technologies such as drone positioning and navigation, attitude control, trajectory planning, and terrain following. Achieving accurate identification and adaptive following of the undulating fruit tree canopy while maintaining a constant spraying distance to ensure uniform pesticide coverage has become a core technological bottleneck. This paper systematically reviews the key technologies and research progress of plant protection UAVs in hilly and mountainous operations, focusing on the principles, advantages, and limitations of core methods such as multi-sensor fusion positioning, intelligent SLAM navigation, nonlinear attitude control and intelligent control, three-dimensional trajectory planning, and multimodal terrain following. It also discusses the challenges currently faced by these technologies in practical applications. Finally, this paper discusses and envisions the future of plant protection UAVs in achieving intelligent, collaborative, and precise operations on steep slopes and in dense orchards, providing theoretical reference and technical support for promoting the mechanization and intelligentization of mountain agriculture. Full article