Search Results (1,323)

We present a (Light)weight (S)atellite–(A)erial feature (M)atching framework (LiteSAM) for robust UAV absolute visual localization (AVL) in GPS-denied environments. Existing satellite–aerial matching methods struggle with large appearance variations, texture-scarce regions, and limited efficiency for real-time UAV applications. LiteSAM integrates three key components to address these issues. First, efficient multi-scale feature extraction optimizes representation, reducing inference latency for edge devices. Second, a Token Aggregation–Interaction Transformer (TAIFormer) with a convolutional token mixer (CTM) models inter- and intra-image correlations, enabling robust global–local feature fusion. Third, a MinGRU-based dynamic subpixel refinement module adaptively learns spatial offsets, enhancing subpixel-level matching accuracy and cross-scenario generalization. The experiments show that LiteSAM achieves competitive performance across multiple datasets. On UAV-VisLoc, LiteSAM attains an RMSE@30 of 17.86 m, outperforming state-of-the-art semi-dense methods such as EfficientLoFTR. Its optimized variant, LiteSAM (opt., without dual softmax), delivers inference times of 61.98 ms on standard GPUs and 497.49 ms on NVIDIA Jetson AGX Orin, which are 22.9% and 19.8% faster than EfficientLoFTR (opt.), respectively. With 6.31M parameters, which is 2.4× fewer than EfficientLoFTR’s 15.05M, LiteSAM proves to be suitable for edge deployment. Extensive evaluations on natural image matching and downstream vision tasks confirm its superior accuracy and efficiency for general feature matching. Full article

Understanding heavy metal behavior in arid saline soils is critical for phytoremediation in degraded lands. This study investigated metal distribution and plant enrichment in the Tarim Basin using 323 soil and 55 plant samples (Populus euphratica, Tamarix ramosissima, cotton, jujube). Analyses included redundancy analysis (RDA) and bioconcentration factor (BCF) assessments. Key findings reveal that elevated salinity (total salts, TS > 200 g/kg) and alkalinity (pH > 8.5) immobilized As, Cd, Cu, and Zn. Precipitation and competitive leaching reduced metal mobility by 42–68%. Plant enrichment strategies diverged significantly: P. euphratica hyperaccumulated Cd (BCF = 1.59) and Zn (BCF = 2.41), while T. ramosissima accumulated As and Pb (BCF > 0.05). Conversely, cotton posed Hg transfer risks (BCF = 2.15), and jujube approached Cd safety thresholds in phosphorus-rich soils. RDA indicated that pH and total salinity (TS) jointly suppressed metal bioavailability, explaining 57.6% of variance. Total phosphorus (TP) and soil organic carbon (SOC) enhanced metal availability (36.8% variance), with notable TP-Cd synergy (Pearson’s r = 0.42). We propose a dual-threshold management framework: (1) leveraging salinity–alkalinity suppression (TS > 200 g/kg + pH > 8.5) for natural immobilization; and (2) implementing TP control (TP > 0.8 g/kg) to mitigate crop Cd risks. P. euphratica demonstrates targeted phytoremediation potential for degraded saline agricultural systems. This framework guides practical management by spatially delineating zones for natural immobilization versus targeted remediation (e.g., P. euphratica planting in Cd/Zn hotspots) and implementing phosphorus control in high-risk croplands. Full article

In this study, the oxidation of industrial hemp staple fibers by the TEMPO/NaBr/NaClO system was explored by the real-time monitoring of the changes in reaction rate, selective oxidative conversion, and reaction time under different operating conditions such as TEMPO usage, NaBr usage, NaClO usage, reaction time, and reaction temperature. We propose a variable-speed competition mechanism between NaClO and TEMPO, which provides experimental support for the long-standing hypothesis that hypochlorite delays acid formation through modulation of the HOCl/OCl⁻ and HOBr/OBr⁻ equilibrium dynamics. The innovative use of combined analysis for several consecutive first-order reactions to investigate the rate-limiting reactions of TEMPO, TEMPO⁺, and TEMPOH over a range of concentrations revealed that the reaction that generates TEMPOH is the key rate-limiting reaction. We characterize the apparent oxidation kinetics of industrial hemp staple fiber in the TEMPO/NaBr/NaClO system using a pseudo-first-order kinetic model, revealing distinct apparent reaction rates across both primary and secondary bast fiber regions. This paper explained the difference in reaction rate between the two aspects of microfibril spatial structure and cellulose crystal structure. The single-factor analysis indicates that reaction time and temperature exert the most significant influence on the conversion rate of selective oxidation within this system. Full article

Against the backdrop of global population growth and intensified resource competition, the sustainable development of the water-energy-food system (WEF) is facing challenges. Wetlands, as key ecological hubs, play a crucial role in regulating water cycles, energy metabolism, and food production, thus serving as a breakthrough point for resolving the bottleneck of resource synergy. Incorporating wetlands into the WEF framework helps us comprehensively understand and optimize the interrelationships among water, energy, and food. This paper proposes an indicator system based on WEFW to study the coupling of water-energy-food-wetland systems and analyzes the evolution of the comprehensive development index of WEFW and its coupling relationship in Jiangxi Province from 2001 to 2022. It uses the grey correlation model to explore the sustainable development capacity of wetland resources, water resources, energy resources, and food resources in Jiangxi Province, and employs a geographical detector model to quantify the contribution of wetlands to WEFW. The research results show that (1) the comprehensive evaluation of WEFW systems in various cities in Jiangxi Province has generally improved, but there is imbalance in regional development. Cities such as Nanchang and Jiujiang have performed well, while cities like Jingdezhen and Xinyu need to enhance resource integration and sustainable development. (2) The coupling coordination degree (CCD) has experienced a process of “stability-fluctuation-recovery”, with a significant increase after 2014, and the spatial differentiation characteristics are obvious. (3) Wetlands play a dominant role in the spatial differentiation of CCD, and their interaction with water, energy, and food resources significantly enhance the explanatory power of their impact on CCD. (4) The grey model indicates that the CCDs of WEFW systems in most cities of Jiangxi Province have a projected annual growth rate of 1.8% (2022–2032), reaching 0.71–0.73 in leading cities. These results emphasize the importance of wetland protection and sustainable resource management in promoting regional coordinated development. The research and prediction of the coupling coordination relationship of water-energy-food-wetland systems can provide a scientific basis for the sustainable development of Jiangxi Province and also offer important scientific references for other regions to achieve a balance between ecological protection and resource utilization. Full article

YKL-40 is a chitinase-like glycoprotein implicated in various pathological processes, yet its glycosaminoglycan (GAG) binding profile beyond heparin has not been examined. In this study, we performed a Microscale Thermophoresis (MST) analysis on the heparin-binding glycoprotein YKL-40 using low molecular weight GAG oligosaccharides. We identified two new GAG ligands, dermatan sulfate (DS) and hyaluronan (HA), while chondroitin sulfate (CS) showed no detectable binding affinity. The results show that heparin is bound with the strongest affinity, followed by DS and HA. To further investigate these differences, molecular docking was used to evaluate possible binding modes. Molecular docking results indicated that both heparin and DS interacted with the same site on YKL-40, the heparin-binding site at residues 143–149, suggesting a multifunctional binding region that may act as a competitive switch or integration hub for spatially regulated signaling. Together, these findings expand the known ligand profile of YKL-40 and offer new insights into its ECM-context-dependent roles, with implications for targeting YKL-40 in diseases involving chronic inflammation, fibrosis, and cancer progression. Full article

The spatial distribution of plant species is a crucial indicator of the mechanisms driving competition or coexistence both within and between populations and communities. Analyzing these patterns provides essential insights into fundamental ecological processes and aids in evaluating ecological hypotheses. To study the spatial distribution of dominant tree species and their associations both within and among species, we established a 25-hectare forest plot in Lushan Mountain. We employed the g(r) function alongside three null models—complete spatial randomness (CSR), heterogeneous Poisson (HP), and antecedent condition (AC)—to analyze spatial patterns and assess species interactions at various life stages. Additionally, we examined the relationships between spatial distributions and environmental factors such as soil properties and topography using Berman’s test. Our results showed that all 12 dominant tree species exhibited significant aggregation under the CSR model; however, the scales of aggregation were reduced under the HP model. We also found evidence of aggregation among multiple species across different life stages and tree layers under CSR. Notably, this pattern persisted under the AC model but was limited to specific spatial scales. Furthermore, elevation, topographical convexity, and the total content of soil nitrogen (N) and carbon (C) were identified as statistically significant predictors of species distributions. Overall, these findings highlight that both biological and environmental factors play a vital role in shaping plant spatial patterns across different scales. Full article

Remote sensing scene image classification has extensive application scenarios in fields such as land use monitoring and environmental assessment. However, traditional methodologies based on convolutional neural networks (CNNs) face considerable challenges caused by uneven image quality, imbalanced sample distribution, intra-class similarities and limited computing resources. To address such issues, this study proposes a hierarchical lightweight attention-enhanced network (HLAE-Net), which employs a hierarchical feature collaborative extraction (HFCE) strategy. By considering the differences in resolution and receptive field as well as the varying effectiveness of attention mechanisms across different network layers, the network uses different attention modules to progressively extract features from the images. This approach forms a complementary and enhanced feature chain among different layers, forming an efficient collaboration between various attention modules. In addition, an improved lightweight attention module group is proposed, including a lightweight dual coordinate spatial attention module (DCSAM), which captures spatial and channel information, as well as the lightweight multiscale spatial and channel attention module. These improved modules are incorporated into the featured average sampling (FAS) bottleneck and basic bottlenecks. The experiments were studied on four public standard datasets, and the results show that the proposed model outperforms several mainstream models from recent years in overall accuracy (OA). Particularly in terms of small training ratios, the proposed model shows competitive performance. Maintaining the parameter scale, it possesses both good classification ability and computational efficiency, providing a strong solution for the task of image classification. Full article

The surface quality of industrial gaskets directly impacts sealing performance, operational reliability, and market competitiveness. Inadequate or unreliable defect detection in silicone gaskets can lead to frequent maintenance, undetected faults, and security risks in downstream systems. This paper presents VST-YOLOv8, a trustworthy and secure defect detection framework built upon an enhanced YOLOv8 architecture. To address the limitations of C2F feature extraction in the traditional YOLOv8 backbone, we integrate the lightweight Mobile Vision Transformer v2 (ViT v2) to improve global feature representation while maintaining interpretability. For real-time industrial deployment, we incorporate the Gating-Structured Convolution (GSConv) module, which adaptively adjusts convolution kernels to emphasize features of different shapes, ensuring stable detection under varying production conditions. A Slim-neck structure reduces parameter count and computational complexity without sacrificing accuracy, contributing to robustness against performance degradation. Additionally, the Triplet Attention mechanism combines channel, spatial, and fine-grained attention to enhance feature discrimination, improving reliability in challenging visual environments. Experimental results show that VST-YOLOv8 achieves higher accuracy and recall compared to the baseline YOLOv8, while maintaining low latency suitable for edge deployment. When integrated with secure industrial control systems, the proposed framework supports authenticated, tamper-resistant detection pipelines, ensuring both operational efficiency and data integrity in real-world production. These contributions strengthen trust in AI-driven quality inspection, making the system suitable for safety-critical manufacturing processes. Full article

In the context of the digital economy reshaping the global competitive landscape, digital industry clusters have become the key driving force to overcome the diminishing returns of traditional inputs and realize sustainable economic development in the digital era. However, the internal mechanisms and spatial effects through which digital industrial clusters drive high-quality development and thereby foster sustainable regional economic growth remain unclear. Based on China’s provincial panel data from 2012 to 2023, this study constructs time-fixed spatial Durbin model and mediation effect model to systematically examine the impact mechanism of digital industry clusters on high-quality economic development, and to analyze their direct effects, spatial spillover effects and mediation transmission effects. The following effects have been found: (1) digital industry clusters can directly promote the high-quality development of the region’s economy (0.070), and can also significantly promote the high-quality development of the region’s economy through the mediating effect of innovative talent agglomeration (0.021); (2) the spatial spillover effect of digital industry clusters consists of the negative siphoning effect of innovative talent and positive technology diffusion and driving effect, which makes the total effect of digital industry clusters on neighboring regions uncertain; (3) Technology-intensive areas, as well as the eastern and northeastern regions, have effectively transformed the advantages of digital industry clusters into momentum for high-quality economic development, whereas central and western regions have not yet fully unleashed the driving effect of digital industry on the high-quality development of the economy, due to the constraints of the industrial structure, innovation factors and infrastructure. Based on the empirical results, the article suggests accelerating the construction of digital industry innovation hubs, establishing cross-regional technology sharing platforms, constructing a negative externality compensation mechanism for talent loss areas, and implementing differentiated regional development strategies. The study addresses a gap in existing research by analyzing the spatial mediation effects of digital industrial agglomeration on high-quality economic development. It extends theoretical insights into industrial clustering within the digital economy and offers actionable policy pathways for developing countries to promote sustainable economic growth through digital industrial clusters. Full article

Construction site layout planning (CSLP) plays a pivotal role in determining the overall efficiency and cost-effectiveness of construction projects. Material handling operations, which constitute a significant portion of indirect project costs, heavily depend on the spatial arrangement of temporary facilities such as site offices, storage yards, and equipment zones. Poorly planned layouts can lead to excessive travel distances, increased material handling times, and operational delays, all of which contribute to inflated costs and reduced productivity. Therefore, optimizing the layout of construction sites to minimize transportation distances and enhance workflow is a critical task for project managers, contractors, and other stakeholders. The challenge in CSLP lies in the complexity of simultaneously satisfying multiple, often conflicting, requirements such as space constraints, safety regulations, and functional proximities. This complexity is compounded by the dynamic nature of construction activities and the presence of numerous facilities to be allocated within limited and irregularly shaped site boundaries. Mathematically, this problem can be formulated as a Quadratic Assignment Problem (QAP), a well-known NP-hard combinatorial optimization problem. The QAP seeks to assign a set of facilities to specific locations in a manner that minimizes the total cost, typically modeled as the sum of products of flows (e.g., material movement) and distances between assigned locations. However, due to the computational complexity of QAP, exact solutions become impractical for medium to large-scale site layouts. In recent years, metaheuristic algorithms have gained traction for effectively tackling such complex optimization problems. Among these, the Advanced Jaya Algorithm (A-JA), a recent population-based metaheuristic, stands out for its simplicity, parameter-free nature, and robust search capabilities. This study applies the A-JA to solve the CSLP modeled as a QAP, aiming to minimize the total weighted travel distance of material handling within the site. The algorithm’s performance is validated through two realistic case studies, showcasing its strong search capabilities and competitive results compared to traditional optimization methods. This promising approach offers a valuable decision-support tool for construction managers seeking to enhance site operational efficiency. Full article

Defect detection in lithium-ion battery (LIB) welding presents unique challenges, including scale heterogeneity, subtle texture variations, and severe class imbalance. We propose a multi-scale convolutional framework that integrates EfficientNet-B0 for lightweight representation learning, PANet for cross-scale feature aggregation, and a YOLOv8 detection head augmented with multi-head attention. Parallel dilated convolutions are employed to approximate self-similar receptive fields, enabling simultaneous sensitivity to fine-grained microstructural anomalies and large-scale geometric irregularities. The approach is validated on three datasets including RIAWELC, GC10-DET, and an industrial LIB defects dataset, where it consistently outperforms competitive baselines, achieving 8–10% improvements in recall and F1-score while preserving real-time inference on GPU. Ablation experiments and statistical significance tests isolate the contributions of attention and multi-scale design, confirming their role in reducing false negatives. Attention-based visualizations further enhance interpretability by exposing spatial regions driving predictions. Limitations remain regarding fixed imaging conditions and partial reliance on synthetic augmentation, but the framework establishes a principled direction toward efficient, interpretable, and scalable defect inspection in industrial manufacturing. Full article

Drivable area detection and segmentation are critical tasks for autonomous mobile robots in complex and dynamic environments. RGB-based methods offer rich semantic information but suffer in unstructured environments and under varying lighting, while LiDAR-based models provide precise spatial measurements but often require high-resolution sensors and are sensitive to sparsity. In addition, most fusion-based systems are constrained by fixed sensor setups and demand retraining when hardware configurations change. This paper presents a real-time, modular RGB–LiDAR fusion framework for robust drivable area recognition and mapping. Our method decouples RGB and LiDAR preprocessing to support sensor-agnostic adaptability without retraining, enabling seamless deployment across diverse platforms. By fusing RGB segmentation with LiDAR ground estimation, we generate high-confidence drivable area point clouds, which are incrementally integrated via SLAM into a global drivable area map. The proposed approach was evaluated on the KITTI dataset in terms of intersection over union (IoU), precision, and frames per second (FPS). Experimental results demonstrate that the proposed framework achieves competitive accuracy and the highest inference speed among compared methods, confirming its suitability for real-time autonomous navigation. Full article

Chinese fir, as a crucial fast-growing tree species in the hilly regions of southern China, exhibits spatial structure characteristics that directly influence both the ecological functionality and productivity of its stands. This study focused on Chinese fir plantations in the Yangkou State-Owned Forest Farm, Fujian Province. Using UAV-LiDAR point cloud data, individual tree parameters such as height and crown width were extracted, and a DBH inversion model was constructed by integrating machine learning algorithms. Spatial structure parameters were quantified through weighted Voronoi diagrams. A comprehensive evaluation system was established based on the combined weighting method and fuzzy evaluation model to systematically analyze spatial structure characteristics and their evolutionary patterns across different age classes. The results demonstrated that growth environment indicators (openness and openness ratio) progressively declined with the stand’s age, reflecting deteriorating light conditions due to increasing canopy closure. Growth superiority (size ratio and angle competition index) exhibited a “V”-shaped trend, with the most intense competition occurring in the middle-aged stands before stabilizing in the over-mature stage. The resource utilization efficiency (uniform angle and forest layer index) showed continuous optimization, reaching optimal spatial configuration in over-mature stands. This study developed a spatial structure evaluation system for Chinese fir plantations by combining UAV data and cloud modeling, elucidating structural characteristics and developmental patterns across different growth stages, thereby providing theoretical foundations and technical support for close-to-nature management and the precision quality improvement of Chinese fir plantations. Full article

Livestock interactions with wildlife have been a concern for managers historically. Invasive feral swine represent an additional management concern in the realm of resource competition as well as zoonotic disease spread between livestock and wildlife. Our study deployed game cameras on a ranch in the Rolling Plains of North Texas to obtain a better understanding of the possibility of interspecies interactions among cattle, feral swine, and white-tailed deer across spatial, temporal, and seasonal variables. Species’ use of bottomlands, shallow uplands, and deep uplands within the ranch were monitored continuously over the course of a year. Cattle and feral swine exhibited high diel activity overlap with the greatest overlap estimates occurring in bottomlands (Δ = 0.889) and wintertime (Δ = 0.875). Cattle and deer exhibited lower diel overlap (Δ = 0.596–0.836, depending on the season and vegetation type), which could be a sign of niche partitioning between the two ungulates. Image captures and overlap estimates suggest interactions between cattle and the other two species occur less frequently in shallow upland sites relative to the other vegetation types. Though image captures of the three species were 17–69% lower in summer relative to fall, indirect interactions may remain high due to competition for shared resources and greater reliance on watering sites. Results suggest that land managers should focus on bottomland sites for feral swine eradication efforts and as areas of increased contact among species. Results can be used to guide livestock and wildlife management and herd health decisions, which can improve ranch economic, environmental, and social sustainability. Full article

With the increasing availability of high-dimensional hyperspectral data from modern remote sensing platforms, accurate and efficient classification methods are urgently needed to overcome challenges such as spectral redundancy, spatial variability, and the curse of dimensionality. The current hyperspectral image classification technique has become a crucial tool for analyzing material information in images. However, traditional classification methods face limitations when dealing with multidimensional data. To address these challenges and optimize hyperspectral image classification algorithms, this study employs a novel fusion method that combines principal component analysis (PCA) based on null spectral information and 2D convolutional neural networks (CNNs). First, the original spectral data are downscaled using PCA to reduce redundant information and extract essential features. Next, 2D CNNs are applied to further extract spatial features and perform feature fusion. The powerful adaptive learning capabilities of CNNs enable effective classification of hyperspectral images by jointly processing spatial and spectral features. The findings reveal that the proposed algorithm achieved classification accuracies of 98.98% and 97.94% on the Pavia and Indian Pines datasets, respectively. Compared to traditional methods, such as support vector machines (SVMs) and extreme learning machines (ELMs), the proposed algorithm achieved competitive performance with 98.81% and 98.64% accuracy on the same datasets, respectively. This approach not only enhances the accuracy and efficiency of the hyperspectral image classification but also provides a promising solution for remote sensing data processing and analysis. Full article