Search Results (261)

Geographic object-based image analysis (GEOBIA) has demonstrated substantial utility in remote sensing tasks. However, its integration with deep learning remains largely confined to image-level classification. This is primarily due to the irregular shapes and fragmented boundaries of segmented objects, which limit its applicability in semantic segmentation. While convolutional neural networks (CNNs) excel at local feature extraction, they inherently struggle to capture long-range dependencies. In contrast, Transformer-based models are well suited for global context modeling but often lack fine-grained local detail. To overcome these limitations, we propose GOFENet (Geo-Object Feature Enhanced Network)—a hybrid semantic segmentation architecture that effectively fuses object-level priors into deep feature representations. GOFENet employs a dual-encoder design combining CNN and Swin Transformer architectures, enabling multi-scale feature fusion through skip connections to preserve both local and global semantics. An auxiliary branch incorporating cascaded atrous convolutions is introduced to inject information of segmented objects into the learning process. Furthermore, we develop a cross-channel selection module (CSM) for refined channel-wise attention, a feature enhancement module (FEM) to merge global and local representations, and a shallow–deep feature fusion module (SDFM) to integrate pixel- and object-level cues across scales. Experimental results on the GID and LoveDA datasets demonstrate that GOFENet achieves superior segmentation performance, with 66.02% mIoU and 51.92% mIoU, respectively. The model exhibits strong capability in delineating large-scale land cover features, producing sharper object boundaries and reducing classification noise, while preserving the integrity and discriminability of land cover categories. Full article

This study investigated the influence of various annealing treatments on the microstructure and corrosion behavior of 20MnCr5 steel in a 3.5% NaCl solution. A combination of microstructural analysis, hardness testing, and electrochemical techniques was used to comprehensively characterize each condition. To enhance data interpretability, a correlation analysis was performed and visualized through a correlation diagram, enabling statistical assessment of the relationships between grain features, phase distribution, mechanical properties, and corrosion indicators. The results demonstrated that corrosion resistance in 20MnCr5 steel is not governed by a single parameter but by the interplay between grain size, morphology, and phase balance. Excessive pearlite content or coarse, irregular grains were consistently associated with higher corrosion rates and lower electrochemical stability. In contrast, a moderate phase ratio and equiaxed grain structure, achieved through normalization, resulted in better corrosion resistance, confirmed by the highest polarization resistance and lowest corrosion current density values among all samples. Although increased grain refinement improved the hardness, it did not always correlate with a better corrosion performance, especially when morphological uniformity was lacking. This highlights the importance of balancing mechanical and corrosion properties through carefully controlled thermal processing. Full article

Plastid division regulatory genes play a crucial role in the morphogenesis of chloroplasts and amyloplasts. Chloroplasts are the main sites for photosynthesis and metabolic reactions, while amyloplasts are the organelles responsible for forming and storing starch granules. The proper division of chloroplasts and amyloplasts is essential for plant growth and yield maintenance. Therefore, this study aimed to examine the J175 (FtsZ2-2) gene, cloned from an ethyl methanesulphonate (EMS) mutant involved in chloroplast and amyloplast division in maize, through map-based cloning. We found that J175 encodes a cell division protein, FtsZ (filamentous temperature-sensitive Z). The FtsZ family of proteins is widely distributed in plants and may be related to the division of chloroplasts and amyloplasts. The J175 protein is localized in plastids, and its gene is expressed across various tissues. From the seedling stage, the leaves of the j175 mutant exhibited white stripes, while the division of chloroplasts was inhibited, leading to a significant increase in volume and a reduction in their number. Measurement of the photosynthetic rate showed a significant decrease in the photosynthetic efficiency of j175. Additionally, the division of amyloplasts in j175 grains at different stages was impeded, resulting in irregular polygonal starch granules. RNA-seq analyses of leaves and kernels also showed that multiple genes affecting plastid division, such as FtsZ1, ARC3, ARC6, PDV1-1, PDV2, and MinE1, were significantly downregulated. This study demonstrates that the maize gene j175 is essential for maintaining the division of chloroplasts and amyloplasts and ensuring normal plant growth, and provides an important gene resource for the molecular breeding of maize. Full article

Wheat disease detection is a crucial component of intelligent agricultural systems in modern agriculture. However, at present, its detection accuracy still has certain limitations. The existing models hardly capture the irregular and fine-grained texture features of the lesions, and the results of spatial information reconstruction caused by standard upsampling operations are inaccuracy. In this work, the GDFC-YOLO method is proposed to address these limitations and enhance the accuracy of detection. This method is based on YOLOv11 and encompasses three key aspects of improvement: (1) a newly designed Ghost Dynamic Feature Core (GDFC) in the backbone, which improves the efficiency of disease feature extraction and enhances the model’s ability to capture informative representations; (2) a redesigned neck structure, Disease-Focused Neck (DF-Neck), which further strengthens feature expressiveness, to improve multi-scale fusion and refine feature processing pipelines; and (3) the integration of the Powerful Intersection over Union v2 (PIoUv2) loss function to optimize the regression accuracy and convergence speed. The results showed that GDFC-YOLO improved the average accuracy from 0.86 to 0.90 when the cross-overmerge threshold was 0.5 (mAP@0.5), its accuracy reached 0.899, its recall rate reached 0.821, and it still maintained a structure with only 9.27 M parameters. From these results, it can be known that GDFC-YOLO has a good detection performance and stronger practicability relatively. It is a solution that can accurately and efficiently detect crop diseases in real agricultural scenarios. Full article

The heating rates and forming temperatures during the hot forming process of titanium alloys cause significant differences in phase transformation, grain size, and dislocation evolution. The formability and service performance of titanium alloy formed components are affected by these factors. This study investigated the hot flow behaviors of Ti-6Al-4V at temperatures ranging from 800 to 900 °C and heating rates ranging from 0.1 to 10 °C/s. These were tested via Gleeble hot tensile experiments, and the grain size and phase evolution were quantitatively characterized via EBSD and XRD. The results suggest that a higher heating rate decreases the β-phase transformation and dislocation density and inhibits grain coarsening, leading to better formability. The heating rate was introduced into the viscoplastic constitutive model for the first time to achieve accurate predictions of the microstructure and hot flow behavior under different heating rates. The prediction accuracy of the hot flow behavior and phase volume fraction reaches 92.93% and 94.97%. The current-assisted hot stamping experiments and finite element (FE) simulations of Ti-6Al-4V irregular cross-section components were carried out at temperatures of 800 and 900 °C and at heating rates of 1 and 3 °C/s. The results show that the rapidly heated formed components exhibit better thickness uniformity and yield strength. The FE simulation guided by the optimized constitutive model has achieved a 96.96% and 92.76% prediction accuracy for the thickness distribution and β-phase volume fraction, respectively. Full article

This study used integrated in vitro and in vivo approaches to investigate how the starch source (glutinous rice, indica rice, maize, or high-amylose rice) influences starch digestion kinetics and, consequently, the apparent nutrient utilization and amino acid metabolism in goslings. Four diets were formulated using glutinous rice, indica rice, maize, and high-amylose rice, and in vitro digestion and animal experiments were carried out. The data showed the particle sizes of the four starches: glutinous rice ≈ indica rice < corn < amylose. The glutinous rice starch grain is a porous polyhedron with an angular surface, the corn starch grain is an ellipsoid with a smooth surface, the indica rice starch grain is a polyhedron with a smooth and compact surface, and the high-amylose starch grain is an irregular polyhedron with a smooth surface. Starch digestibility was relatively stable for the indica and corn-based diets, and starch digestibility was higher for the indica rice diet compared to the corn- and high-amylose starch-based diets. The utilization of Asp, Ser, Glu, Gly, and Phe was higher for the glutinous rice diet compared to the maize and high-amylose diets. Furthermore, with this diet, the availability of Thr and Ala was observed to be higher than with the indica rice and high-amylose diets. In conclusion, the particle size and structure of starch from different sources (glutinous rice, indica rice, corn, and high-amylose rice) were different, significantly affecting the starch digestion rate. The glutinous rice diet enables a fast digestion rate for starch, which is rapidly digested in the proximal intestine. The inadequate supply of glucose in the distal intestine enhances the oxidative energy supply from dietary amino acids in that region, thereby improving the apparent digestibility of both starch and crude protein. Full article

Traditional coal quality assessment methods rely exclusively on the laboratory testing of physical samples, which impedes detailed stratigraphic evaluation and limits the integration of intelligent precision mining technologies. To resolve this challenge, this study investigates geophysical logging as an innovative method for coal quality prediction. By integrating scanning electron microscopy (SEM), X-ray analysis, and optical microscopy with interdisciplinary methodologies spanning mathematics, mineralogy, and applied geophysics, this research analyzes the coal quality and mineral composition of the Shanxi Formation coal seams in northern Jiangsu, China. A predictive model linking geophysical logging responses to coal quality parameters was established to delineate relationships between subsurface geophysical data and material properties. The results demonstrate that the Shanxi Formation coals are gas coal (a medium-metamorphic bituminous subclass) characterized by low sulfur content, low ash yield, low fixed carbon, high volatile matter, and high calorific value. Mineralogical analysis identifies calcite, pyrite, and clay minerals as the dominant constituents. Pyrite occurs in diverse microscopic forms, including euhedral and semi-euhedral fine grains, fissure-filling aggregates, irregular blocky structures, framboidal clusters, and disseminated particles. Systematic relationships were observed between logging parameters and coal quality: moisture, ash content, and volatile matter exhibit an initial decrease, followed by an increase with rising apparent resistivity (LLD) and bulk density (DEN). Conversely, fixed carbon and calorific value display an inverse trend, peaking at intermediate LLD/DEN values before declining. Total sulfur increases with density up to a threshold before decreasing, while showing a concave upward relationship with resistivity. Negative correlations exist between moisture, fixed carbon, calorific value lateral resistivity (LLS), natural gamma (GR), short-spaced gamma-gamma (SSGG), and acoustic transit time (AC). In contrast, ash yield, volatile matter, and total sulfur correlate positively with these logging parameters. These trends are governed by coalification processes, lithotype composition, reservoir physical properties, and the types and mass fractions of minerals. Validation through independent two-sample t-tests confirms the feasibility of the neural network model for predicting coal quality parameters from geophysical logging data. The predictive model provides technical and theoretical support for advancing intelligent coal mining practices and optimizing efficiency in coal chemical industries, enabling real-time subsurface characterization to facilitate precision resource extraction. Full article

Structural timber is an important building material, but surface defects such as cracks and knots seriously affect its load-bearing capacity, dimensional stability, and long-term durability, posing a significant risk to structural safety. Conventional inspection methods are unable to address the issues of multi-scale defect characterization, inter-class confusion, and morphological diversity, thus limiting reliable construction quality assurance. To overcome these challenges, this study proposes WDNET-YOLO: an enhanced deep learning model based on YOLOv8n for high-precision defect detection in structural wood. First, the RepVGG reparameterized backbone utilizes multi-branch training to capture critical defect features (e.g., distributed cracks and dense clusters of knots) across scales. Second, the ECA attention mechanism dynamically suppresses complex wood grain interference and enhances the discriminative feature representation between high-risk defect classes (e.g., cracks vs. knots). Finally, CARAFE up-sampling with adaptive contextual reorganization improves the sensitivity to morphologically variable defects (e.g., fine cracks and resin irregularities). The analysis results show that the mAP50 and mAP50-95 of WDNET-YOLO are improved by 3.7% and 3.5%, respectively, compared to YOLOv8n, while the parameters are increased by only 4.4%. The model provides a powerful solution for automated structural timber inspection, which directly improves building safety and reliability by preventing failures caused by defects, optimizing material utilization, and supporting compliance with building quality standards. Full article

Background/Objectives: A critical part of the maize life cycle takes place during the summer, and due to climate change, its growth and development are increasingly exposed to the irregular and unpredictable effects of drought stress. Developing and using new cultivars with increased drought tolerance for farmers is the easiest and cheapest solution. One of the concepts to screen for drought tolerance is to expose germplasm to various growth scenarios (environments), expecting that random drought will occur in some of them. Methods: In the present study, thirty-two maize hybrids belonging to four FAO maturity groups were tested for grain yield at six locations over two consecutive years. In parallel, data of the basic meteorological elements such as air temperature, relative humidity and precipitation were collected and used to compute two indices, scPDSI (Self-calibrating Palmer Drought Severity Index) and VPD (Vapor Pressure Deficit), that were assessed as indicators of drought (water deficit) severity during the vegetation period. Practical implementation of these indices was carried out indirectly by first analyzing yield data using a factor analytic model to detect latent environmental variables affecting yield and then correlating those latent variables with drought indices. Results: The first latent variable, which explained 47.97% of the total variability, was correlated with VPD (r = −0.58); the second latent variable explained 9.57% of the total variability and was correlated with scPDSI (r = −0.74). Furthermore, latent regression coefficients (i.e., genotypic sensitivities to latent environmental variables) were correlated with genotypic drought tolerance. Conclusions: This could be considered an indication that there were two different acting mechanisms in which drought affected yield. Full article

Ultrasonic pulse velocity (UPV) is a widely used technique for diagnosis and structural safety assessment of existing buildings. The main difficulties in UPV tests on-site are due to one-sided accessibility of materials and degraded/irregular surfaces. Pulse-echo ultrasonic tomography (PE-UT) can overcome the problem. Though it has been widely applied for detecting inhomogeneities within concrete, few works use the instrument to assess UPV. The present paper aims to fill the gap by comparing PE-UT results with those of through-transmission ultrasonic tests (TT-UT) commonly used for UPV characterization. TT-UT measurements were performed with cylindrical and exponential transducers. The latter are used on irregular surfaces or when coupling gel is forbidden. Few data are in the literature comparing exponential and cylindrical transducers’ results. This is a further element of novelty of the paper. PE-UT and TT-UT results were compared considering the effect of material compositeness, water, transmission mode, and transducer type. It was found that PE-UT allows for reliable and rapid one-sided measurements on concrete and stone in different conditions. The differences between PE-UT and TT-UT results were between 1 and 3%. Exponential transducers gave reliable results on fine-grained stone in direct transmission, with differences lower than 4% with cylindrical transducer results. Full article

The electrochemical behavior of copper (II) on glassy carbon from an eutectic mixture of choline chloride (ChCl) and ethylene glycol (EG) was investigated using cyclic voltammetry (CV). The redox and deposition processes were studied for electrolyte concentrations of 0.01 M and 0.5 M Cu(II), with particular attention paid to the effects of different Cu(II) concentrations on the copper deposition potential and morphology of the copper deposits. The CV results showed that the Cu(II) species are reduced to Cu(0) via two separate steps. Higher Cu(II) concentrations in the electrolyte triggered the formation of differently coordinated Cuⁿ⁺ complexes next to the electrode, which shifted the electrodeposition potential of Cu(I)/Cu(0) couples towards more positive values. The Cu deposits were obtained potentiostatically from 0.01 M and 0.5 M Cu(II)-ChCl:EG electrolyte and analyzed using scanning electron microscopy, energy-dispersive X-ray spectroscopy, and X-ray diffraction. The different copper concentrations in electrolytes induced different morphologies of electrodeposited copper, where the mixture of irregular grains and carrot or needle-like dendrites was obtained from 0.01 M, and rose-like forms were obtained from 0.5 M electrolytes. This study is the first to identify these rose-like forms and the mechanism of their formation, which is discussed in detail. Full article

Functionally graded materials (FGMs) are fabricated on Ti-6Al-4V alloy surfaces to improve insufficient surface hardness and wear resistance. Microstructure and mechanical properties and strengthening–toughening mechanisms of FGMs were investigated. The FGM cladding layer exhibits distinct gradient differentiation, demonstrating gradient variations in the nanoindentation hardness, wear resistance, and Al/V elemental composition. Molten pool dynamics analysis reveals that Marangoni convection drives Al/V elements toward the molten pool surface, forming compositional gradients. TiN-AlN eutectic structures generated on the FGM surface enhance wear resistance. Rapid solidification enables heterogeneous nucleation for grain refinement. The irregular wavy interface morphology strengthens interfacial bonding through mechanical interlocking, dispersing impact loads and suppressing crack propagation. FGMs exhibit excellent wear resistance and impact toughness compared with Ti-6Al-4V titanium alloy. The specific wear rate is 1.17 × 10⁻² mm³/(N·m), dynamic compressive strength reaches 1701.6 MPa, and impact absorption energy achieves 189.6 MJ/m³. This work provides theoretical guidance for the design of FGM strengthening of Ti-6Al-4V surfaces. Full article

LiDAR-based 3D object detection is fundamental in autonomous driving but remains challenging due to the irregularity, unordered nature, and non-uniform density of point clouds. Existing methods primarily rely on either graph-based or tree-based representations: Graph-based models capture fine-grained local geometry, while tree-based approaches encode hierarchical global semantics. However, these paradigms are often used independently, limiting their overall representational capacity. In this paper, we propose density-aware tree–graph cross-message passing (DA-TGCMP), a unified framework that exploits the complementary strengths of both structures to enable more expressive and robust feature learning. Specifically, we introduce a density-aware graph construction (DAGC) strategy that adaptively models geometric relationships in regions with varying point density and a hierarchical tree representation (HTR) that captures multi-scale contextual information. To bridge the gap between local precision and global contexts, we design a tree–graph cross-message-passing (TGCMP) mechanism that enables bidirectional interaction between graph and tree features. The experimental results of three large-scale benchmarks, KITTI, nuScenes, and Waymo, show that our method achieves competitive performance. Specifically, under the moderate difficulty setting, DA-TGCMP outperforms VoPiFNet by approximately 2.59%, 0.49%, and 3.05% in the car, pedestrian, and cyclist categories, respectively. Full article

This study focuses on the structural and electrochemical behavior of the compound ZnLaFeO₄ as a negative electrode material for nickel–metal hydride (Ni-MH) batteries. The material was synthesized by a sol–gel hydrothermal method to assess the influence of lanthanum doping on the ZnFe₂O₄ spinel structure. X-ray diffraction revealed the formation of a dominant LaFeO₃ perovskite phase, with ZnFe₂O₄ and La₂O₃ as secondary phases. SEM analysis showed agglomerated grains with an irregular morphology. Electrochemical characterization at room temperature and a discharge rate of C/10 (full charge in 10 h) revealed a maximum discharge capacity of 106 mAhg⁻¹. Although La³⁺ doping modified the microstructure and slowed the activation process, the electrode exhibited stable cycling with moderate polarization behavior. The decrease in capacity during cycling is due mainly to higher internal resistance. These results highlight the potential and limitations of La-doped spinel ferrites as alternative negative electrodes for Ni-MH systems. Full article

The detection of tunnel cracks plays a vital role in ensuring structural integrity and driving safety. However, tunnel environments present significant challenges for crack detection, such as uneven lighting and shadow occlusion, which can obscure surface features and reduce detection accuracy. To address these challenges, this paper proposes a novel crack detection network named STCYOLO. First, a dynamic snake convolution (DSConv) mechanism is introduced to adaptively adjust the shape and size of convolutional kernels, allowing them to better align with the elongated and irregular geometry of cracks, thereby enhancing performance under challenging lighting conditions. To mitigate the impact of shadow occlusion, a Shadow Occlusion-Aware Attention (SOAA) module is designed to enhance the network’s ability to identify cracks hidden in shadowed regions. Additionally, a tiny crack upsampling (TCU) module is proposed, which reorganizes convolution kernels to more effectively preserve fine-grained spatial details during upsampling, thereby improving the detection of small and subtle cracks. The experimental results demonstrate that, compared to YOLOv8, our proposed method achieves a 2.85% improvement in mAP and a 3.02% increase in the F score on the crack detection dataset. Full article