Search Results (669)

Convolutional Neural Networks (CNNs) and Vision Transformers (ViTs) have achieved significant success in various computer vision applications, including the classification of high-resolution imagery. However, a notable limitation of these deep learning approaches is their tendency to inadequately preserve the precise edges and shapes of target objects. In contrast, Object-Based Image Analysis (OBIA) offers a methodology that emphasizes the preservation of object boundaries by segmenting images into meaningful objects. Combining CNNs and ViTs with OBIA leverages the feature extraction capabilities of these deep learning algorithms and the boundary-preserving advantages of OBIA, leading to enhanced classification accuracy and improved delineation of object boundaries in high-resolution images. Still, the main challenge for combining these methods lies in effectively aligning the irregularly shaped image objects produced by OBIA with the regular image patches required by CNNs and ViT architectures. In this study, we propose a novel approach that integrates superpixel segmentation with CNNs and ViTs for the automatic classification of benthic habitats using high-resolution orthomosaic images. Initially, the Simple Linear Iterative Clustering (SLIC) algorithm was applied to segment the high-resolution orthomosaic images into superpixels. Subsequently, the central points of the resulting superpixels were utilized to generate square image patches. These patches performed as inputs for ConvNeXt-Base and EfficientNet-B0 pre-trained CNNs to extract fine-grained features and Dinov2 ViTs to extract high-level features. Then, a Support Vector Machine (SVM) classifier was trained using these attributes to classify benthic habitats. Eventually, the classification label derived from the SVM defined the class of each superpixel segment. This method achieved an average overall accuracy of 0.96 in classifying benthic habitats. Overall, we demonstrate that combining CNNs, ViTs, and superpixel segmentation is an effective approach to benthic habitats classification, providing accurate high-resolution maps of heterogeneous reef environments. Full article

Blueberry production has rapidly expanded over the past decade, accompanied by growing demand for efficient and accurate methods to monitor the flowering and fruiting phases of blueberry development, which has a direct impact on yield potential. Accurate determination of blueberry phenology enables growers to make data-driven decisions on freeze protection applications and harvest windows. In addition, objective phenology data of blueberry mapping populations will provide high-quality phenotype data for the discovery of genetic mechanisms regulating blueberry flowering and fruiting times. Traditional approaches, such as manual counting and visual ratings, are labor-intensive and subjective in capturing variation across genotypes. Recent progress in computer vision and deep learning has enabled automated flower detection, but most existing studies on blueberries remain restricted to narrow flowering windows or close-up images, limiting their application at the bush level and across the seasonal development. In this study, we developed BerryFlowerNet, a customized YOLO-based model to detect and count blueberry flower clusters from bud to green fruit stages. A comprehensive dataset was collected on three dates using a field phenotyping robot, covering five flowering stages. The integration of CFNet, a custom module fusing shallow spatial features, and PIoU loss improved the detection performance. Additionally, the Slicing Aided Hyper Inference algorithm was employed to address small-object detection in bush-level images. Experimental results demonstrated that BerryFlowerNet outperformed the baseline YOLO model and three additional detectors, achieving an average mAP0.5 of 0.644 across five independent training runs. The model achieved an accuracy of 0.88 when predicting blueberry flowering stages, indicating its effectiveness and accuracy. Additionally, the results of the bush-level image analysis showed the capability of the model to capture genotype-level differences in flowering dynamics. Overall, this approach offers new opportunities for growers and breeders to determine blueberry phenological development that is critical for optimizing on-farm management strategies and advancing precision phenotyping to facilitate the development of climate-resilient blueberries. Full article

Down quality is the core evaluation indicator of thermal insulation products, and its grade determination strictly complies with the down content index specified in the national standard GB/T 17685-2016 Feather and Down. Traditional down quality inspection adopts manual sorting and weighing methods, which are plagued by low efficiency, strong subjectivity and high error rates, thereby restricting the intelligent upgrading of the down industry. This study aims to develop an automatic down detection and quantitative grading method conforming to national standards based on deep learning. A down dataset consisting of 632 RGB images is constructed, with each image containing 5–10 individual down samples and covering five categories: mature down clusters, immature down clusters, down filaments, feathers, and yellow-tail down. Three mainstream frameworks including YOLOv8, YOLOv11 and YOLOv26 are trained for performance comparison. Precision, recall, mAP@50 and mAP@50-95 are adopted as evaluation metrics. In addition, this paper proposes a research idea for down content calculation and automatic classification and grading of down quality in accordance with relevant national standards. The experimental results demonstrate that the latest models do not necessarily achieve the optimal performance. The newly released YOLOv26n and YOLOv26m exhibit relatively low accuracy in the down detection task, with mAP@50 values of only 0.98556 and 0.99077, and recall rates of 0.95032 and 0.97848, respectively, failing to outperform their previous-generation counterparts. In contrast, YOLOv11n achieves the best comprehensive performance, with an mAP@50 of 0.99416, a precision of 0.99544, a recall of 0.99722, and an mAP@50-95 of 0.63464. Meanwhile, the model has only 2.58 M parameters, a computational complexity of 6.3 GFLOPs, and a single training time of approximately 6.7 min, achieving an optimal balance between detection accuracy and computational efficiency. All models show the highest detection accuracy for mature down clusters and yellow-tailed down, while slight confusion exists between immature down clusters and down filaments. This study verifies the feasibility of the YOLO series models in down quality inspection in accordance with national standards, and reveals that model architecture iteration does not necessarily lead to performance improvement on specific industrial datasets. The lightweight and robustly designed YOLOv11n presents greater practical value. The intelligent detection scheme proposed in this paper can assist in optimizing the traditional manual quality inspection workflow, alleviating the burden of manual counting and reducing subjective errors. It provides new ideas and technical references for the rapid screening and objective determination of down quality. Furthermore, the proposed research framework for automatic classification and grading of down quality is expected to promote the development of down quality inspection toward standardization, intelligence, and automation in the future. Full article

This paper investigates whether event-level pitch-bearing fatigue damage can be estimated directly from turbine measurements, and whether these mechanical damage metrics leave measurable fingerprints in the generator DC-link voltage and current. To achieve this, a case study was performed using SCADA and structural load data from the 45 kW Chalmers (Björkö) research turbine. This data was segmented into 223 park-run-park pitch events. For each event, blade-root flapwise and edgewise bending moments were converted into radial and axial loads at the pitch bearing; an equivalent dynamic bearing load $P_{\mathrm{eq}}\left(t\right)$ was reconstructed using SKF and DG03 formulations; and rainflow counting with an S–N curve and Palmgren–Miner’s rule was used to compute event-level damage indices compatible with the International Standard Organization basic rating life concepts. In parallel, DC-link voltage and current were summarized into time-domain features, combined with operating-condition descriptors, and clustered using PCA-based k-means. The resulting clusters captured distinct electrical regimes that, across several event batches, corresponded to different levels of accumulated fatigue damage: regimes with sustained high DC-link voltage and longer duration tended to exhibit higher mean damage indices than lower, steadier DC regimes, indicating an electromechanical link. The results show that physics-based lifetime estimation and unsupervised analysis of existing electrical traces can be combined into a hybrid workflow for pitch-bearing condition assessment without additional sensors. Full article

With the large-scale integration of offshore wind farms (OWFs), harmonic issues caused by the interaction between high-frequency switching of converters and complex network impedances pose severe challenges to power quality. Traditional harmonic monitoring heavily relies on post-event fixed-threshold alarm mechanisms, which struggle to achieve early warning during the low-distortion sub-health operation stage and lack the capability for multi-dimensional tracing of harmonic degradation sources. To address these limitations, this paper proposes a deep warning network for grid harmonics combining multi-step regression and multi-dimensional tracing within a unified multi-task learning (MTL) architecture. First, a deep shared feature encoder, integrating a bi-directional long short-term memory (Bi-LSTM) network with a multi-head self-attention (MHSA) mechanism, is utilized to extract high-order temporal coupling features between meteorological evolution and multi-node electrical states. Subsequently, the main task branch executes a k-step-ahead multivariate time-series regression to accurately predict the evolution trend of total harmonic distortion (THD) at both the point of common coupling (PCC) and the turbine terminal. Simultaneously, the auxiliary task branch performs multi-label micro-state classification based on relative degradation thresholds, achieving fine-grained multi-dimensional tracing covering spatial nodes, electrical attributes, and their joint micro-states. Experimental results on real-world OWF operational data demonstrate that through the joint optimization of regression and tracing tasks, the proposed MultiDimKStepMTL model significantly improves time-series prediction accuracy, achieving a $10.3\%$ relative improvement over single-task baselines, while substantially reducing computational overhead. This research successfully advances grid harmonic monitoring from passive response to proactive micro-state early warning, providing a solid, highly interpretable data-driven foundation for active filter control of offshore wind clusters. Full article

Accurate crop row detection is identified as a critical prerequisite for autonomous agricultural navigation, yet it remains challenging in complex field environments. To achieve a balance between segmentation accuracy, robustness, and real-time performance, an improved crop row segmentation and navigation method based on the DeepLabV3+ framework was developed. MobileNetV2 was adopted as the backbone to minimize computational costs, while feature representation was enhanced through integrated attention mechanisms and multi-scale fusion. Specifically, split-attention convolution was integrated into the backbone, a DenseASPP + SP module was employed for multi-scale contextual capture, and a Convolutional Block Attention Module (CBAM) was added to refine feature responses. Experimental results demonstrated that the proposed method outperformed mainstream models, achieving a mean Intersection over Union (mIoU) of 93.42% and an f1-score of 96.8%. The model maintained a lightweight architecture with 8.35 M parameters and a real-time speed of 32 FPS. Furthermore, crop row anchor points were extracted and processed via DBSCAN clustering and RANSAC fitting to generate high-precision navigation lines. Validation showed that the middle crop row yielded the highest fitting accuracy with minimal angular and lateral errors. This study provides an efficient visual perception solution for intelligent field operations. Full article

Visual scene matching navigation (VSMN) for unmanned aerial vehicles (UAVs) boasts advantages such as high precision, high reliability, and autonomy. The biggest challenge lies in the tension between local fine-grained information and global semantics, as well as limited generalization ability in real-world environments. While existing Transformer-based cross-view geolocation methods enhance global context modeling capabilities, they still generally face issues such as high demands on training data and computational resources, insufficient fusion of local fine-grained information and global semantics, and real-time performance in real-world complex environment. To address these problems, we propose a scene matching and localization algorithm based on the Siamese-ViT. For feature extraction, we use the ViT model to extract global features and K-means clustering to aggregate local features. Combined with the global features extracted by the ViT, a robust local–global feature representation vector is generated. For feature matching, incremental principal component analysis (IPCA) is used to reduce the dimensionality of the high-dimensional feature space, and a KD-tree is constructed for fast feature retrieval to improve matching efficiency. We validated our algorithm on the University-1652 dataset and a dataset of real-world satellite-drone image pairs. The results show that our Siamese-ViT outperforms other models in both Recall and AP. We conduct flight experiments in real-world environments, capturing drone images of complex scenes, including farmland, urban buildings, and waterways. The results show that, at a flight altitude of 350 m, our algorithm achieves an average absolute value of 6.2063 m for latitude, 6.7552 m for longitude, and 10.1922 m for horizontal error. Therefore, our Siamese-ViT demonstrates ideal overall positioning accuracy. Full article

As the number of unmanned aerial vehicles (UAVs) and the volume of computational tasks increase in UAV ad hoc networks (UAVANET), the solution space for task allocation strategies grows exponentially. In practical emergency scenarios with concurrent multi-user access, multi-UAV systems equipped with mobile edge computing (MEC) devices face challenges such as limited computing resources and imbalanced task distribution during task offloading. To address these challenges, this paper proposes an adaptive task allocation algorithm named AUSTA-DQHO (Adaptive UAV Swarm Task Allocation using Deep Q-networks and Genetic Algorithms Hybrid Optimization), which combines Deep Q-Network (DQN) with Genetic Algorithm (GA), aiming to optimize computational task scheduling and minimize both the total task delay and the variance in task delays. First, we introduce a multi-UAV-assisted MEC application framework. In this framework, UAVs equipped with high-performance computing modules are deployed as airborne servers in the target area, providing data offloading and task computation support for IoT devices. Next, to tackle the optimization problem, we replace the random action selection process in DQN with a hybrid strategy that incorporates heuristic methods—specifically, GA and greedy algorithms—to perform global search and make more effective decisions for optimal task allocation for each offloading request. Furthermore, to accelerate the convergence of the AUSTA-DQHO policy while ensuring global optimality, we introduce a pre-clustering mechanism and a dynamic weighting factor for randomly generated task offloading requests in the target area. These mechanisms effectively reduce the solution space and ensure that optimal actions are learned at different stages of the training process. Experimental results demonstrate that the proposed algorithm achieves a task latency reduction of 18.72% and a load balancing improvement of 98.72%, surpassing the performance of the other algorithms. Additionally, we explore the optimal number of UAVs under given environmental conditions to minimize the waste of computing resources. Full article

While the demand for free-form architecture (FFA) has increased with advancements in computer-aided design (CAD) technology, the rationalization of complex surfaces into fabricable panels remains a significant challenge due to high production costs and technical complexity. Practical pain points, such as the prohibitive cost of unique molds and the inefficiency of manual data processing during design iterations, pose substantial economic risks. This study proposes an intelligent surface rationalization framework that integrates parametric design with machine learning algorithms in Autodesk^TM Dynamo Studio, a plug-in to Revit. A data-driven classification workflow was developed using four key geometric parameters—planarity, principal curvature (PC), Gaussian curvature (GC), and mean curvature (MC). Two unsupervised learning algorithms, a Gaussian mixture model and K-means clustering, were compared for their classification performance. As a result of two case studies, free-form surface classification by a Gaussian mixture model (CGMM) demonstrated flexibility in modeling complex surface data by probabilistically managing the uncertainty of the curvature distribution, and free-form surface classification by K-means clustering (CKC) was confirmed to be effective for the rapid classification of large-scale panel data. Optimizing the proportion of flat and single-curved panels through the proposed workflow contributes to deriving a reasonable balance between design intent and construction costs/constructability at the early design stage, and strengthening risk management capabilities for FFA. Full article

The generation of Digital Surface Models (DSMs) from large-format, high-resolution satellite imagery constitutes a critical component of photogrammetry and computer vision. Achieving efficient, robust, and high-quality DSM reconstruction has therefore become a prominent research focus. However, with the continuous improvement in satellite image resolution and the increasing diversity of image sources, satellite image matching—serving as the fundamental step in DSM generation—still faces significant challenges, including the uneven distribution of feature points and insufficient registration stability in large-scale imagery. To address these issues, this paper presents a refined DSM generation method for high-resolution satellite imagery, termed SAT-MAK. The framework consists of three main stages: (1) sparse matching based on MAK (Multi-type Aggregated Keypoints) extraction; (2) a density-weighted clustering matching optimization strategy; and (3) DSM generation following a conventional photogrammetric pipeline. Experiments were conducted on multiple sets of high-resolution satellite imagery, and the proposed method was compared with four commonly used satellite image 3D reconstruction algorithms. The results demonstrate that, compared with state-of-the-art methods, the proposed SAT-MAK approach improves DSM completeness by 5.29% while maintaining competitive RMSE performance, highlighting its strong potential for practical applications. Full article

The design of differentiated fiscal instruments for industrial sustainability requires robust, data-driven tools capable of capturing the heterogeneity of environmental performance across manufacturing units—a challenge that conventional econometric approaches address only partially, given the non-linear nature of operational–environmental interactions in reconfigurable production systems. This study introduces a two-phase computational framework that integrates unsupervised machine learning and supervised classification to generate evidence-based sustainability profiles for fiscal policy targeting. Its principal contribution is the combination of K-Means clustering with a binary artificial neural network (ANN) classifier, operationalized through an accessible decision-support interface that enables differentiated incentive allocation without requiring programming expertise from policymakers. A dataset of 1000 manufacturing records comprising seven operational and technological input variables—material usage, production capacity, reconfiguration time, downtime, AI optimization, IoT connectivity, and predictive maintenance—and three environmental output indicators—energy consumption, carbon emissions, and waste generation—was analyzed. In Phase One, K-Means segmentation with k = 6, selected through multi-criteria convergence (Silhouette = 0.102; Elbow, Davies–Bouldin, and Calinski–Harabasz indices), identified six distinct sustainability profiles with marked environmental differentiation. In Phase Two, a binary ANN classifier (architecture: 7 → 64 → 32 → 1 neurons; ReLU and sigmoid activations) was trained to distinguish the reference cluster C0 (low environmental impact: energy 145.1 kWh, emissions 45.2 CO₂-eq) from the high-impact cluster C1 (emissions 67.8 CO₂-eq, waste 41.5 kg). The trained classifier achieved an overall accuracy of 75.4% and an AUC-ROC of 0.774 on the held-out test set, with a macro-averaged F1-score of 0.753 and a Cohen’s kappa coefficient of 0.508, indicating moderate-to-substantial agreement beyond chance. Class C1 (high-impact establishments) achieved a precision of 0.794 and a recall of 0.730, supporting reliable identification of manufacturing units that would most benefit from targeted fiscal support. The framework is deployed through a Gradio-based graphical interface incorporating a traffic-light sustainability classification (green/yellow/red), enabling direct and interactive application by tax authorities and industrial policymakers. The modular architecture supports adaptation to larger or sector-specific datasets, making it transferable across industrial policy contexts. Full article

Hyperspectral image (HSI) analysis plays a central role in remote sensing tasks requiring fine-grained material discrimination, vegetation health assessment, and post-disturbance monitoring. Yet, the high dimensionality and strong spectral redundancy in HSIs often reduce the efficiency and reliability of machine learning models. These challenges are especially important in wildfire science and prescribed-fire monitoring, where spectral responses vary due to burn severity, char deposition, canopy structure, and early vegetation recovery. Benchmark datasets such as Indian Pines and Pavia University and others provide controlled environments for algorithms’ evaluation, but real-world post-fire forest conditions pose additional complexity. This study presents a unified and comprehensive evaluation of five dimensionality reduction strategies: Principal Component Analysis (PCA), Spatial–Spectral Edge Preservation (SSEP), Spectral-Redundancy Penalized Attention (SRPA), and a Deep Reinforcement Learning (DRL)-based selector together with a clustering based baseline, K-Means Clustering-Based Band Selection (KMCBS). These strategies are combined with classical machine learning and deep learning classifiers: Random Forest (RF), Support Vector Machines (SVMs), K-Nearest Neighbors (KNNs), and 3D Convolutional Neural Networks (3D-CNN). The full pipeline includes exploratory data analysis, preprocessing, patch-based spatial–spectral modeling, consistent train–validation protocols, and multi-dataset evaluation across Indian Pines, Pavia University, and a new custom VNIR hyperspectral dataset collected after prescribed burns at the Lubrecht Experimental Forest in Montana, USA. By systematically comparing statistical, edge-aware, attention-guided, and reinforcement learning-based band-selection strategies, this work identifies compact yet informative spectral subsets that enhance classification performance while reducing computational cost. Importantly, the inclusion of the Montana prescribed-burn dataset provides a unique real-world testbed for understanding band selection behavior in fire-affected forest environments. Overall, this study contributes a generalizable and extensible framework for HSI dimensionality reduction and classification, laying the groundwork for future applications in wildfire assessment, vegetation recovery monitoring, and remote sensing. Full article

This study presents an experimental platform that integrates computer vision and machine learning to support approximate pH estimation and endpoint detection in titration experiments for science education. A Raspberry Pi-based setup was used to capture real-time solution images, which were converted into RGB data for analysis. Grid-based image preprocessing reduced artifacts caused by ripples and localized color variations. Cluster analysis identified three RGB-based solution categories that were correlated with pH. Regression analysis, including Random Forest modeling, achieved high predictive accuracy with low error. Machine learning classification models were also evaluated, with Random Forest and K-Nearest Neighbors showing strong performance for the non-linear relationship between pH and RGB values. The results support the feasibility of using BTB within its transition range for approximate pH estimation and endpoint detection in an educational setting. The system can also be used as an educational platform through which students engage with automated data collection, machine learning, and real-time analysis. By reducing subjective visual observation and improving experimental reproducibility, this approach supports the use of digital technologies in science education. Full article

Airfield lighting control (ALC) is critical for ensuring safe, efficient, and compliant airport operations, especially under low-visibility conditions. However, current centralized control architectures cannot adequately meet the real-time responsiveness, scalability, and reliability requirements of Advanced Surface Movement Guidance and Control Systems (A-SMGCS) Level IV. To overcome these limitations, this paper proposes a novel cloud–edge–end collaborative architecture for a mobile ALC scenario, in which we formulate a joint task computing and energy consumption optimization problem to maximize long-term system utility under latency, computation, and communication constraints. In this way, the mobile airfield lighting (MAL) system can also quickly adapt its optimal formation pattern based on the airport environment, lighting conditions, and the type of aircraft taking off or landing via efficient computation, thereby achieving the best navigational assistance effect. For solving such an optimization problem, a framework that combines K-medoids with the Improved Twin Delayed Deep Deterministic Policy Gradient (ITD3) is proposed to integrate the efficiency of clustering for rough allocation and the high-precision dynamic optimization capability of the improved TD3. The training depends on edge nodes and the cloud to achieve online performance. Finally, the extensive simulation proved that our novel algorithm is efficient. Full article

Background/Objectives: Selective inhibition of JAK1 remains a major challenge in cytokine-signaling therapeutics due to the high structural similarity of the JAK family. Here, we present an integrated computational framework that combines large-scale binding-site conformational analysis, ensemble docking, and protein–ligand interaction fingerprinting (PLIF) to elucidate the structural determinants of JAK1 selectivity and prioritize JAK1-biased scaffolds. Methods: A curated set of JAK1 and JAK2 catalytic-domain structures was clustered to capture binding-site diversity, and representative conformers were evaluated using >2300 annotated ligands. Docking performance was assessed via AUC, early enrichment metrics, and structural pose validation against experimentally resolved complexes. The workflow was subsequently applied to a library of ~6000 drug-like compounds to prioritize candidates with predicted JAK1 preference. Results: Across the ensemble, the most predictive features reliably separated active from inactive ligands (AUC = 0.78–0.82) and captured subtle, systematic rank shifts supporting the reported JAK1 bias. Interaction fingerprint analysis revealed a conserved hinge-binding motif required for potency, alongside a JAK1-enriched hotspot adjacent to Glu aD.55 that contributes to isoform discrimination. Applied to a library of ~6000 drug-like molecules, the workflow yielded 174 candidates predicted to exhibit preferential JAK1 recognition and reduced JAK2 engagement. Conclusions: These findings define the structural and physicochemical features underlying JAK1 selectivity and illustrate how ensemble-based modeling can guide the discovery of next-generation selective kinase inhibitors. Full article