Search Results (487)

Mapless deep reinforcement learning (DRL) navigation in dynamic indoor environments is difficult under single-frame 2D LiDAR, which reports where an obstacle is but not whether it is approaching. We introduce the Positional-Velocity Spatio-Temporal Attention Module (PV-STAM), a compact perception block (19,968 trainable parameters, under 3% of network capacity) that combines a per-sector scan-difference channel with a two-head self-attention mechanism and a 384→48 compression bottleneck. Policies are trained with Soft Actor-Critic across a seven-phase progressive curriculum with bidirectional demotion, scaling from static goal-seeking to fifteen simultaneously moving obstacles at 0.18 m/s, with a Hardware-Calibrated Training Mode applied in the final phase. Seven configurations were evaluated on three zero-shot benchmark arenas over 300 episodes each across three random seeds, and key control variants were validated in 130 valid physical trials on a TurtleBot3 Waffle Pi across two matched corridor scenarios. The simulation and physical evaluations give different orderings, and this dissociation is the paper’s principal result. In simulation, SAC-PV-STAM, SAC-R-PV-STAM and SAC-MLP-FS lie within 4.6 percentage points of one another on two of three benchmarks and are not separable; on hardware, they separate with large margins. SAC-R-PV-STAM reached the goal in 19 of 20 static trials and 20 of 20 trials with a moving obstacle, with no threshold violations across 40 trials, against SAC-MLP-FS (stratified p = 0.00068) and SAC-PV-STAM (stratified p = 1.1 × 10⁻⁷). A non-recurrent variant stalled with a clear floor ahead—SAC-PV-STAM commanded no forward velocity above 0.005 m/s in any of the 2746 samples of the final 91.3 s of a 122.3 s stall while the LiDAR reported 3.50 m directly ahead—and a prediction stated before the experiment, that introducing a moving obstacle would remove the stall, was confirmed (timeouts 9/10 to 0/20, p = 7 × 10⁻⁷). We further report a scan-difference contamination analysis showing up to 12.0× more spurious spikes during self-rotation on hardware, a width-matched comparison separating the effect of Huber critic loss from that of critic width, and a recurrent LSTM baseline that remains significantly inferior across all three benchmarks and unstable across seeds. Full article

Core wall leakage is a common seepage-related defect occurring during the operation of core wall dams. Clarifying the characteristics and evolution of seepage field distortion under leakage conditions is an important basis for assessing dam safety status and evaluating leakage risks. In this study, short-duration, small-scale physical model tests of an asphalt concrete core wall dam were conducted under leakage heads of 14 cm (S1) and 20 cm (S2) to characterize the spatial response of the seepage field. Finite element simulations were subsequently validated against the experimental results and extended through parametric analyses of leakage channel widths of 1–10 mm and burial depths of 5–29 m. The experiments showed that the distortion profiles on horizontal sections were approximately semi-elliptical, with the maximum extent occurring at the elevation of the leakage center and progressively decreasing with vertical distance. The maximum dam-axis diffusion width increased from 9.00 cm under S1 to 11.05 cm under S2, while the maximum streamwise diffusion distance increased from 4.70 to 6.00 cm, corresponding to increases of 22.8% and 27.7%, respectively. Three-dimensional reconstruction further showed that the distorted region exhibited a quasi-semi-ellipsoidal morphology centered on the leakage location. Quantitative comparison between the experimental and numerical results yielded MAEs of 1.325 and 0.550 cm, RMSEs of 1.506 and 0.628 cm, and MAPEs of 18.01% and 13.90% for the dam-axis diffusion width and maximum streamwise diffusion distance, respectively, indicating that the numerical model captured the observed spatial response with moderate quantitative discrepancies. Engineering-scale simulations further showed that increasing leakage channel width generally enlarged the distorted region, whereas the effect of burial depth was non-monotonic and controlled by the downstream phreatic surface. Above the phreatic surface, seepage field distortion increased with leakage channel burial depth; once the leakage channel reached or extended below the phreatic surface, the distortion weakened with further increases in depth. These findings quantitatively characterize the spatial response of the internal seepage field to local core wall leakage and provide a physical basis for interpreting leakage-induced anomalies in seepage monitoring of core wall dams. Full article

Wireless sensor networks in agricultural fields face abrupt node failures from environmental stress and physical damage, yet most clustering protocols assume fault-free operation. This paper proposes SMA-WSN, a failure-aware clustering protocol inspired by the Slime Mold Algorithm, selecting cluster heads via a tri-objective fitness that combines residual energy, base-station proximity, and local coverage. Evaluated in NS-3.47 against 11 protocols over six agricultural topology archetypes—spatial node-layout patterns derived from Algerian field geometries and densities, not biophysical canopy or soil-moisture propagation models—and five failure rates (18,000 runs, 50 seeds), SMA-WSN achieves the highest per-hop delivery ratio at every failure rate and ranks first overall on that metric, with its largest margins on the heterogeneous, oasis, and sparse-Saharan terrains (T4–T6). We also report an end-to-end delivery ratio, counting a datum only when it reaches the base station: under that stricter definition, all 12 protocols change rank, and SMA-WSN places seventh. We report both and treat the divergence between them as a result in its own right. SMA-WSN achieves the highest per-hop PDR while remaining statistically tied with the best protocols in energy fairness (Jain index 0.922, within one confidence half-width of the leaders GWO-LEACH, PSO-LEACH, and FGO-QL). An ablation over the three fitness terms and their weights shows this advantage is robust to objective weighting and arises from the SMA selection dynamics rather than any single fitness term: no term, including coverage, is individually decisive, so the advantage does not depend on a finely tuned weighting. Disabling the stochastic wrap phase, by contrast, costs 11.4 percentage points of per-hop PDR, identifying it as the mechanism actually responsible for the gain. On First Node Death under failure, SMA-WSN is mid-field rather than dominant; we report this honestly and discuss it as an open question for follow-up work distinguishing energy-exhaustion deaths from abrupt hardware failures. Full article

Morphometric information on ambling horses raised in Türkiye remains limited. This study aimed to characterize the morphometric traits of ambling horses according to breeding origin, age, and sex and to provide baseline phenotypic information on the studied population. A total of 102 naturally ambling horses were evaluated, including indigenous horses raised in Bursa (n = 28), indigenous horses raised in Nazilli, Aydın Province (n = 65), and Afghan-origin horses raised in Bursa (n = 9). Morphometric measurements of the head, body, and limbs were recorded, and the effects of breeding origin, age, and sex were analyzed using a general linear model. Significant differences among breeding origins were observed for several morphometric traits (p < 0.05). Afghan-origin horses generally exhibited higher values for several body measurements, particularly withers height, body length, chest depth, chest width, and heart girth, reflecting morphometric differences within the studied sample. Age- and sex-related differences were also observed for several morphometric traits. The strongest correlation was observed between CH and BH (r = 0.920, p < 0.001), followed by WH and BH (r = 0.873, p < 0.001) and WH and CH (r = 0.860, p < 0.001). ITH was also positively correlated with BH (r = 0.825), CH (r = 0.808), and WH (r = 0.744) (all p < 0.001). Overall, this study provides baseline morphometric information on ambling horses from different breeding origins raised in Türkiye and contributes to the phenotypic characterization of the studied population. However, because the Afghan-origin group included only nine horses and the distribution of animals among some groups was unbalanced, the observed differences should be interpreted with caution. Further studies involving larger and more balanced populations are needed to confirm these findings. Full article

This study investigated and optimized the operational process parameters of an extrusion-based 3D food printing system to maximize the dimensional fidelity of newly developed Wolffia globosa (duckweed) starch hydrogel constructs relative to a nominal target specification of 30 × 30 × 30 mm. Prior to parameter optimization, synchrotron X-ray tomographic microscopy (SR-XTM) was used to characterize Wolffia globosa particle size and dispersion within the starch matrix, showing that grinding eliminated large particle agglomerates (up to approximately 150 µm) and was necessary for smooth, continuous extrusion; the ground formulation was accordingly selected for all printing trials. A full factorial experimental configuration was executed to examine the synchronized effects of three core process parameters: print-head traverse speed (5–15 mm/s), extrusion speed (5–15 steps/mm), and layer height (1.9–3.7 mm). Experimental responses were evaluated via Three-Way Analysis of Variance (ANOVA) and Response Surface Methodology (RSM) using triplicate measurements (n = 3) at each of the 27 tested parameter combinations. Residual diagnostics indicated an approximately normal distribution for the height model (Shapiro–Wilk p = 0.716), whereas the width and length models showed some departure from normality (p < 0.01), consistent with the significant lack-of-fit detected for these two responses. Three-way ANOVA confirmed that print-head (nozzle) speed was the dominant factor governing the in-plane dimensions (width and length; partial η² ≈ 0.98), while height was jointly governed by all three factors, with layer height and print-head speed contributing the largest effects. With the statistical power afforded by replicate measurements, all two- and three-way interactions among the three factors were also statistically significant for width and length (p < 0.001), refining the single-replicate interaction pattern reported previously. Empirical second-order polynomial equations explained a substantial share of the variance in each dimension (R² = 0.70–0.85), although formal lack-of-fit testing indicated that higher-order interactions not captured by the quadratic terms remained statistically significant, and the equations should therefore be interpreted as descriptive rather than as precise predictive tools. Based on the triplicate means, a print-head speed of 5 mm/s, extrusion speed of 15 steps/mm, and a layer height of 1.9 mm minimized the mean cumulative absolute error to 4.01 mm, yielding a mean dimensional profile of 28.63 ± 0.78 mm width, 27.76 ± 0.96 mm length, and 30.41 ± 0.70 mm height (mean ± SD, n = 3). Full article

Structural cracks that develop within semi-rigid asphalt pavement structures may remain concealed beneath the pavement surface, making their opening widths, depths, and morphologies difficult to determine nondestructively. Although ground-penetrating radar (GPR) can localize subsurface reflective cracks, the hyperbolic anomalies in a B-scan, a two-dimensional cross-sectional radar profile obtained from sequential measurements along a survey line, do not directly represent the actual geometrical dimensions of the cracks. To solve this problem, this paper proposes a geometrical dimension detection and morphological inversion (GDMI) method, called the GPR-GDMI. The GPR-GDMI comprises two independently trained networks: GPR-GCDNet for geometrical dimensions detection and GPR-CMIGAN for morphology inversion. GPR-GCDNet utilizes trapezoidal boxes to represent crack dimensions in B-scan, enhanced by directional crack attention (DCA) for hyperbolic pattern extraction, a SetTrap Head for trapezoidal proposal generation, and multi-scale feature fusion for improved receptive field and coverage. GPR-CMIGAN is an unsupervised framework consisting of two independently designed U-Net generators and discriminators, with Gaussian noise injected to prevent mode collapse. The crack-dimension-constrained cycle-consistency loss incorporates GPR-GCDNet’s detected dimensions as physical constraints in the inversion, while Wasserstein loss mitigates over-constraints and improves stability during training. On the test set, GPR-GCDNet achieved an AP of 0.84, an F1-score of 0.89, and an mIoU of 0.74 under an IoU threshold of 0.5. GPR-CMIGAN achieved a performance on MS-SSIM, LPIPS, and FID of 0.9984, 0.0566, and 2.3145, respectively. Ablation studies confirm the contribution of each component. By quantitative experiments, GPR-GDMI achieves accurate detection of crack geometrical dimensions and morphological inversion based on field-collected GPR data. Full article

Jadeite transparency grading is challenging because adjacent categories exhibit high visual similarity and commercial images often contain complex backgrounds and acquisition-related variations. This study proposes a lightweight grayscale-guided visual perception network based on YOLOv10s for joint jadeite localization and transparency-category grading. First, the Jade Texture Layered Convolution (Jade-TLC) module replaces the initial CBS layers and combines asymmetric convolution for slender linear textures with dilated grouped convolution for regional granular features. The two branches are adaptively fused using the whole-image grayscale mean as an auxiliary image-level statistic rather than a physical measurement of transmittance. Second, an improved SimBAM module applies lightweight dual-pooling spatial attention before each detection head to suppress background interference. Third, the Anisotropic Minimum Point Distance IoU (A-MPDIoU) loss normalizes horizontal and vertical coordinate deviations by the ground-truth width and height to improve localization of boxes with unequal scales. Experiments on 4325 images show that the proposed model achieves 94.25% mAP@50, 93.34% precision, 92.11% recall, and 92.72% F1-score, with 7.82 M parameters, 23.4 G FLOPs, and 95.21 FPS. Controlled ablation, sensitivity, and replacement experiments support the complementary contributions of the proposed components. Further validation on independently collected datasets is required. Full article

Silphium perfoliatum L. is a promising perennial bioenergy and forage crop; however, its agricultural expansion is severely hindered by primary physiological dormancy and non- uniform seedling emergence This study evaluated the effects of chemical priming treatments (distilled H₂O, H₂O₂, GA₃, KNO₃, ethephon, and sodium nitrophenolate (SNP)) and the seed mass classification (small: 0.011–0.015 g; medium: 0.016–0.019 g; large: 0.020–0.024 g) on germination dynamics, allometry, and seedling vigor. Under controlled conditions, priming agents demonstrated highly differentiated morpho-regulatory capacities. Gibberellic acid (GA₃) significantly overcame physiological dormancy, maximizing the final germination percentage (77.33%), germination index (2.54), and seedling vigor index II (SV II) (4.62) compared to the control (23.33%, 0.99, and 1.11, respectively). Conversely, KNO₃ and ethephon optimized germination synchrony (Z) (0.27 and 0.29) and suppressed early fragile shoot etiolation to enhance dry matter accumulation (0.064 g). Furthermore, multivariate morphometric analysis of seed mass categories revealed a distinct transverse allometric expansion in larger seeds (length-to-width ratio decreasing from 1.58 in small seeds to $1.50$ in large seeds). While the ultimate germination capacity remained statistically unaffected by seed size (37.60% to 47.20%), large seeds exhibited a pronounced early kinetic divergence, achieving a developmental head start (23.2% cumulative germination by day 8 versus (7.2%) for small seeds. Larger seed mass directly translated into superior structural biomass accumulation, yielding significantly robust seedlings with elevated fresh weight (1.059 g vs. 0.789 g in small seeds), dry weight (0.073 g vs. 0.041 g), and total seedling length (91.32 mm vs. 73.70 mm), without altering the radicle-to-shoot allocation ratio (0.89 vs. 0.76). We conclude that early crop establishment in S. perfoliatum is driven by dormancy-breaking chemical signaling and reserve-dependent physical seed quality. Full article

Piano key weirs (PKWs) increase the effective crest length and provide high discharge capacity within limited spillway widths; however, their hydraulic efficiency depends on key geometry. Addressing the limited understanding of key-curvature effects, this study systematically investigates the influence of inlet- and outlet-key curvature on the discharge coefficient (C_d) and crest-wise discharge distribution of type-A PKWs. Nine configurations are examined experimentally in a flume at the Hydraulics Research Institute, Egypt. These include a reference model with linear inlet and outlet keys and modified geometries incorporating C- and S-shaped keys in both normal and inverted forms. Experiments are conducted under free-flow conditions for total-head ratios H_T/P up to 0.23, where H_T is the total upstream head above the crest and P is the PKW height. A three-dimensional numerical model using FLOW-3D HYDRO is validated against these experimental data and subsequently extended to H_T/P ≈ 0.50–0.60. All reported results and conclusions are based on the numerical simulations over the full investigated range. Results show that C_d increases to a peak near H_T/P ≈ 0.08 and then decreases gradually for all configurations. Both inlet- and outlet-key curvature influence hydraulic performance by modifying flow guidance and redistribution along the crest. However, only the inverted C-shaped inlet configuration (M2) improves efficiency, increasing C_d by about 5.46% on average over the investigated head range and enhancing the total discharge by approximately 4.36% at H_T/P ≈ 0.12 relative to the reference case. Discharge-distribution analysis indicates that this improvement results from inlet feeding and efficient flow conveyance along the sidewall crest. A predictive relation is developed for the optimal configuration to support hydraulic design. Full article

Background: The zygomaticus major (ZMa) is a key mimetic muscle with reported anatomical variability, but in vivo descriptions based on routinely acquired imaging remain limited. Methods: This retrospective single-centre study evaluated whether clinically indicated head-and-neck CT angiography (CTA) can support a bone-anchored, consensus-based descriptive assessment of ZMa morphology. Seventy-five CTA examinations were analysed bilaterally, yielding 150 hemifaces from 42 males and 33 females. Multiplanar reconstructions were reviewed using skeletal landmarks. Results: ZMa morphology was categorised as Type I (single-bellied), including Type Ia with independent zygomatic origin and Type Ib with proximal fusion with the zygomaticus minor; Type II (double-bellied); or Type III (three-bellied). The ZMa was identifiable in all hemifaces. Type I was observed in 86.7% of hemifaces, including Type Ia in 78.7% and Type Ib in 8.0%; Type II occurred in 12.0% and Type III in 1.3%. Patient-level asymmetry was present in 22.7%. Mean ZMa length was 41.77 ± 6.21 mm, proximal attachment width 5.95 ± 1.36 mm, and distal attachment width 4.39 ± 1.41 mm. These findings suggest that routinely acquired CTA may provide an opportunistic, bone-referenced approach to the descriptive in vivo assessment of visible ZMa morphology and cohort-specific morphometric characteristics. Conclusions: CTA should not be considered a reference-standard modality for facial muscle assessment, and further validation using dedicated soft-tissue imaging and formal reliability testing is required. Full article

The rice field eel (Monopterus albus) is an economically important aquaculture species widely distributed in Southeast Asia. However, the lack of nationally approved varieties and the increasing demand for high-quality eel products have highlighted the necessity of developing improved germplasm resources for aquaculture production. In this study, morphometric morphology and muscle texture traits were evaluated in 367 wild eel individuals, and a genome-wide association study (GWAS) based on whole-genome resequencing was conducted to identify genetic variants and candidate genes associated with economically important traits, including body length (BL), tail length (TL), tail height (TH), head length (HL), head width (HW), head height (HH), hardness, chewiness, springiness, and resilience. The GWAS identified three significant and seventeen suggestive SNPs associated with body shape traits, with most loci located on chromosome 4. Candidate gene analysis of the associated genomic regions revealed several candidate genes potentially associated with growth regulation, nutrient metabolism, and skeletal development, including b4galnt4a, tmem86a, cpt1b, mrpl23, lama1, gcm2, wnt3a, thrab, and bmpr1a. Additionally, four suggestive SNPs associated with muscle texture traits were detected, and candidate genes related to muscle development and extracellular matrix regulation, including fstl1b, col5a3a, and pnn, were identified. These findings provide new insights into the genetic basis underlying morphometric variation and muscle texture diversity in M. albus. The identified SNP markers and candidate genes provide potential molecular resources for understanding the genetic basis of body morphology and muscle texture variation in rice field eel. Full article

This paper reports the design, implementation, and experimental evaluation of a laboratory autonomous parking platform that supports both perpendicular and parallel parking. The platform integrates three components. First, a depth-image processing pipeline comprising grayscale conversion, Gaussian filtering, Canny edge detection, and Hough transform line extraction is used for parking space detection, together with a pixel-width criterion for distinguishing perpendicular from parallel spaces. Second, a turning-radius trajectory planning strategy based on Ackermann steering geometry determines the steering positions for each parking mode. Third, a fuzzy correction scheme, using membership functions for lateral position and heading angle estimated from web camera imagery, refines the final parking pose. Throughout all maneuvers, LiDAR provides 360° environmental scanning for obstacle detection and emergency stop. Seven test cases covering all supported parking modes were carried out; the parking mode was selected correctly and the maneuver was completed in all seven cases. The evaluation is qualitative, and the design parameters reported here are specific to the hardware configuration used. The contribution is accordingly a transparent and fully documented reference implementation rather than a performance advance. Full article

Reliable identification of pest developmental stages provides a foundation for investigating aphid development and adaptive mechanisms and informs the selection of timely interventions in integrated pest management. In this study, eight morphological indicators of Uroleucon gobonis were distinguishing under a stereomicroscope. Additionally, six candidate genes identified from transcriptomic data were selected to characterize their expression profiles across five developmental instars. These results showed that the body length enabled distinguishing of the aphids from the first instar to the fourth instar, while antennal length and cauda length facilitated distinction between the second–fourth instar nymphs; cornicle length allowed separation of the third instar and subsequent instars. In contrast, body width permitted discrimination primarily between the first and second instars; head width, foreleg length, and hindleg length exhibited substantial overlap across instars, limiting their utility for instar discrimination. Elevated expression of DN1031, DN1019, and DN1093 was a prominent feature of first instar nymphs. DN1098 was persistently increased from the first to the third instar, which facilitated discrimination among these three developmental stages. The fourth instar was characterized by the concurrent down-regulation of DN1098 and DN1031 relative to the third instar, whereas adults exhibited a distinct expression peak of DN136 compared to other developmental stages. This study establishes an integrated instar-identification system for U. gobonis, combining rapid morphological screening with molecular characterization. This framework provides a valuable methodological basis for investigating developmental plasticity and supports the development of targeted pest management strategies. Full article

Rail light-strip morphology reflects the wheel-rail contact condition. Reliable automatic analysis remains difficult. The strip is narrow and has weak boundaries, while specular reflection, rail-head texture and trackside background interfere with color inspection images. This study proposes a segmentation-guided geometric method for rail light-strip abnormality analysis. An improved SegFormer jointly segments the background, rail-head and light-strip regions. A boundary detail enhancement module refines weak rail-head and light-strip contours. Focal Loss emphasizes minority and hard boundary pixels. The rail-head mask provides the geometric reference for extracting the light-strip centerline, eccentricity, width sequence and connected-component morphology. The predicted masks are ordered using the corrected mileage record. Every 1000 original-resolution rows then form a consecutive 1 m detection unit. When a geometric rule is triggered, the method reports that unit’s 1 m mileage interval together with its eccentricity, width-change or local-integrity measurement. The model achieves 95.67% mean Intersection over Union (mIoU) on 3520 annotated images. It detects 845 of 876 positive units, with 96.46% recall, 89.23% precision and 92.70% F1-score. The resulting records identify abnormal 1 m mileage intervals and report the corresponding eccentricity, width-change, or local-integrity measurements for targeted manual review. Full article

Body measurements provide a non-invasive basis for estimating live weight, yet their predictive value and age-related variation remain poorly characterized in female Jiangyue donkeys. Body weight (BW) and 11 morphometric traits were recorded in 484 females: withers height (WH), body length (BL), chest circumference (CC), cannon circumference (CAC), head length (HL), neck length (NL), chest width (CW), chest depth (CD), rump height (RH), rump length (RL), and rump width (RW). Descriptive statistics, correlation analyses, and regression analyses were performed using SPSS 27.0; stepwise regression was then applied to derive and validate BW prediction equations. Additionally, four nonlinear functions—Logistic, Gompertz, Brody, and von Bertalanffy—were fitted to cross-sectional BW data from 241 females sampled at different ages to characterize the population-level relationship between age and weight. BW was positively correlated with all morphometric traits (p < 0.01), with CC showing the strongest association. The optimal equations explained 91.8% and 84.4% of the variation in BW among growing and adult donkeys, respectively (both p < 0.01). Of the four nonlinear functions evaluated, the Brody model provided the closest fit (R² = 0.99960) and yielded an estimated asymptotic mature weight of 191.049 kg. Because the age–weight analysis was based on cross-sectional observations, the resulting curve represents population-level variation rather than within-animal growth trajectories. Collectively, these models offer quantitative tools for live-weight estimation, growth assessment, and herd management in female Jiangyue donkeys. Full article