Search Results (8,004)

This study investigates the feasibility of acoustic signal analysis for the assessment of weld bead quality in the shielded metal arc welding (SMAW) process. The work focuses on comparing time-domain acoustic signals and time–frequency spectrogram representations for the classification of welds as accepted or rejected according to standard welding inspection criteria. Two key acoustic descriptors, the fundamental frequency (F₀) and the harmonics-to-noise ratio (HNR), were extracted and analyzed to evaluate statistical differences between the two weld quality classes. Statistical tests, including Anderson–Darling, Levene, ANOVA, and Kruskal–Wallis (α = 0.05), revealed significant differences between accepted and rejected welds. Accepted welds exhibited a bimodal HNR distribution associated with transient arc instability at the beginning and end of the bead, whereas rejected welds showed more uniform acoustic behavior throughout the process. Subsequently, the acoustic data were represented using both audio signals and spectrograms and used as inputs for ten supervised machine learning models, including Support Vector Classifier (SVC), Logistic Regression (LR), k-Nearest Neighbors (KNN), Decision Tree (DT), Random Forest (RF), Extra Trees (ET), Gradient Boosting (GB), and Naïve Bayes (NB). The results demonstrate that spectrogram-based representations significantly outperform time-domain signals, achieving accuracies of 0.95–0.96, ROC-AUC values above 0.95, and false positive and false negative rates below 6%. These findings indicate that, while scalar acoustic descriptors provide statistically significant insight into weld quality, time–frequency representations combined with machine learning enable a more robust and reliable framework for automated non-destructive evaluation, particularly in manual SMAW processes under realistic operating conditions. Full article

Vigna umbellate, a typical edible and medicinal crop, is rich in polyphenolic compounds with antioxidant, antibacterial, anti-inflammatory, and lipid-regulating activities. However, traditional methods for polyphenol content detection rely on chemical analysis, which is cumbersome and time-consuming, making it difficult to meet the demands of high-throughput rapid detection. Although hyperspectral imaging technology offers the potential for non-destructive and rapid detection, existing analytical methods are often limited by issues such as high spectral band redundancy, insufficient feature extraction, and inadequate model stability, which constrain prediction accuracy and practical application potential. To address this, this study proposes a multi-scale residual convolutional neural network (MS-RCNN) based on competitive adaptive reweighted sampling (CARS) for feature band selection, combined with near-infrared hyperspectral imaging technology, to construct a rapid and non-destructive prediction model for the polyphenol content of Vigna umbellata. The model employs a parallel multi-scale convolutional module to extract spectral features with different receptive fields, and incorporates residual connections and adaptive pooling mechanisms to enhance feature reuse and robustness. Experiments compared the performance of partial least squares regression (PLSR), least squares support vector machine (LS-SVM), multi-scale convolutional neural network (MS-CNN), and MS-RCNN models. The results indicate that the MS-RCNN model based on CARS screening achieved the best prediction performance, with a coefficient of determination (R²) of 0.9467, a root mean square error of prediction (RMSEP) of 0.0448, and a residual predictive deviation (RPD) of 4.33. Compared with the optimal PLSR and LSSVM models, its R² values were improved by 0.2078 and 0.1119, respectively. In summary, the MS-RCNN model proposed in this study enables rapid, non-destructive, and accurate prediction of polyphenol content in Vigna umbellata, providing an efficient technical approach for quality detection of edible and medicinal crops. Full article

Cotton, a globally significant crop grown in over 100 countries, sustains a $40 billion market and provides employment for over 350 million people worldwide. However, plastic contamination remains a persistent challenge within the industry, degrading cotton fiber quality and disrupting ginning. Manual inspection and optical machine-vision systems struggle when plastic fragments are concealed by fibers or lack sufficient color contrast. To address these challenges, we developed an ultrasonic phased-array imaging system operating at 40 kHz under field-programmable gate array (FPGA) control. Transmitter elements emit pulsed ultrasound along radial paths, separate reflection receivers record echo amplitudes to form acoustic images, and a set of transmission receivers captures signal attenuation, which is overlaid onto the reflection-based image to highlight potential contaminants. In preliminary laboratory-based tests on both seed cotton and lint samples, the system successfully detected visually obscured plastic fragments as small as $2\mathrm{cm}\times 2\mathrm{cm}$ with an angular resolution limit of $\pm 3^{\circ }$. Distinct reflection peaks and corresponding attenuation overlays were produced across the field of view, validating the system’s detection capabilities. These results demonstrate the feasibility of using ultrasonic imaging to reveal concealed plastics in cotton processing. Integrating this approach with existing optical methods could enhance contaminant-removal workflows and improve overall fiber quality and processing efficiency. Full article

The dielectric properties of cement-based composites (CBC) are highly sensitive to environmental humidity, which seriously restricts the quantitative interpretation accuracy of ground-penetrating radar (GPR) in the non-destructive testing of cement concrete pavement. In view of the lack of targeted prediction models due to the unclear mechanism of humidity influence in existing research, the core innovations of this study are: (1) the synergistic mechanism of water vapor dipole polarization and adsorbed water multi-layer polarization is clarified, revealing the intrinsic reason for the accelerated growth of permittivity in the high humidity range; (2) the constructed four-component dielectric model of “cement mortar–aggregate–water vapor–adsorbed water” achieves high-precision prediction within the range of 50~100% RH (R² > 0.94, relative error < 5%), and shows good predictive ability within the test scope of this study; (3) a GPR humidity correction protocol based on the model is proposed, which can effectively improve the accuracy of nondestructive testing of cement concrete structures. In this study, CBC samples with water–cement ratios of 0.4~0.6 were prepared using P.O 32.5/P.O 42.5 cement and limestone aggregate. Under the conditions of 20 ± 0.5 °C, relative humidity (RH) of 50~100%, and 2 GHz (common GPR frequency), the permittivity was measured using an Agilent P5001A network analyzer to verify the model. The results show that the permittivity increases monotonically with humidity, and the growth rate in the high humidity range (70~100%) is 2.2 times that of the low humidity range (50~70%); The higher the water–cement ratio, the shorter the age, and the lower the cement strength grade, the stronger the humidity sensitivity of CBC dielectric properties. This model provides a reliable humidity correction tool for GPR detection, and significantly improves the accuracy of nondestructive evaluation of cement concrete structures. Full article

Composite materials are extensively used in the aeronautical and aerospace industries for their high strength-to-weight ratios but are vulnerable to barely visible impact damage (BVID), which can severely compromise structural integrity. Digital shearography (DS) provides a non-contact, full-field solution for subsurface inspection; however, low signal-to-noise ratios in raw phase maps often hinder precise damage identification. This study explores a post-processing methodology utilizing a band-pass filtering algorithm and temporal summation to isolate damage-related spatial frequencies. An in-house digital shearography system was used to inspect a carbon-fiber-reinforced polymer (CFRP) plate subjected to 13.5 J and 26.2 J impacts. Twelve phase maps, acquired during the thermal cooling stage, were processed using a multi-pass filters to systematically analyze different frequency ranges. Results demonstrate that summing multiple filtered phase maps significantly enhances the contrast of damage signatures compared to single phase maps or traditional unwrapping techniques. Furthermore, quantitative assessment using image quality metrics, such as the generalized contrast-to-noise ratio (gCNR), confirmed that optimal frequency selection is essential for an accurate damage delineation. This approach provides a robust framework for improving the reliability and sensitivity of non-destructive testing in composite structures. Full article

The increasing demand for high-quality grapes necessitates rapid and objective quality assessment methods to overcome the limitations of traditional subjective and destructive techniques. This study investigated the feasibility of using hyperspectral imaging combined with machine learning for non-destructive quality evaluation of fresh grapes. Hyperspectral data were acquired from four table grape varieties (“Rose”, “Yongyou”, “Xiahei”, and “Jumbo”), and their Soluble Solids Content (SSC) was measured, which varied significantly among varieties. We extracted texture features using the Gray-Level Co-occurrence Matrix (GLCM) from images at key wavelengths, which were a combination of those selected by the Successive Projections Algorithm (SPA) and sensitive wavelengths. Comparative models for variety classification (qualitative) and SSC prediction (quantitative) were built using Extreme Learning Machine (ELM), Convolutional Neural Network (CNN), and Partial Least Squares (PLS) with full-range spectra and texture features as inputs. The results showed that the ELM model using full-range spectra was superior for both tasks, achieving a classification accuracy of 97.56% and, for SSC prediction, an $R_p^2$ of 0.75 and an RMSEP of 0.81. Notably, CNN models also showed considerable robustness. Our findings confirm that combining hyperspectral imaging with machine learning is a viable strategy for fresh grape quality assessment. Full article

To enable non-destructive quantitative characterization of constituent content in C/C–SiC ceramic-matrix composites, this study develops a physics-guided framework based on multispectral photon-counting X-ray detection. In practical photon-counting measurements, multispectral attenuation features are jointly distorted by detector-response non-idealities, including charge sharing, K-escape, and finite energy resolution, as well as by beam-hardening effects from the polychromatic X-ray source. To address this coupled problem, a Geant4 11.2-based detector-response model was incorporated into a unified correction workflow together with beam-hardening compensation, so that physically consistent multispectral attenuation vectors could be recovered for subsequent constituent inversion rather than merely for spectrum restoration. On this basis, a fine-grained theoretical database covering different SiC mass fractions was established, and quantitative constituent inversion was achieved by matching the corrected attenuation features to the database. Experimental results show that the proposed framework effectively suppresses thickness-dependent bias in attenuation measurements and yields an average relative error below 3% for pure aluminum. For C/C–SiC composites, the SiC mass fraction can be quantified with an accuracy better than 3 wt%. These results demonstrate that the proposed method provides a practical non-destructive route for constituent-content characterization in heterogeneous ceramic-matrix composites and is valuable for manufacturing quality control and in-service assessment. Full article

Nitrogen (N) and phosphorus (P) are essential nutrients that play critical roles in plant physiological processes and the accumulation of N and P in broccoli head was significantly correlated with yield. Therefore, there is a need for a rapid, non-destructive diagnosis of crop status by detecting deficiencies in essential nutrients. This study evaluated the effects of N and P deficiency on field grown broccoli (Brassica oleracea L. var. italica Plenk) using a plant-induced electrical signal (PIES) sensor, in which needle electrodes are inserted into the stem to measure electrical conductivity reflecting plant water and ion status. Four treatments were established, including the control (N100P100) with sufficient N and P supply, N deficiency (N0P100), P deficiency (N100P0), and combined N–P deficiency (N0P0). For sufficient supply, urea and fused phosphate (FP) were applied at rates of 122 kg N ha⁻¹ and 71 kg P ha⁻¹, respectively. Soil, stem, and leaf nutrient contents, growth parameters, and stress related indicators were analyzed and their relationship with PIES values were evaluated. PIES was highest in control (N100P100) and lowest under N–P deficiency (N0P0). Higher PIES values were observed during the vegetative stage, whereas values declined during the reproductive stage, reflecting changes in physiological activity. Growth parameters such as shoot and root weight and stem diameter were generally superior in the control (N100P100) treatment, while leaf calcium (Ca), magnesium (Mg), and potassium (K) concentrations showed no significant differences among treatments. Total N content in leaves was higher in N fertilized treatments (control and P deficiency). Photosynthesis-related parameters, including soil plant analysis development (SPAD), Fv/Fm, and chlorophyll content, were lowest under N–P deficiency, which was reflected in the PIES. Principal component analysis (PCA) showed that the PIES was closely associated with growth and photosynthetic parameters and clearly distinguished N sufficient treatments (control and P deficiency) from N deficient treatments (N0P100, N0P0). Overall, these findings suggest that PIES monitoring can serve as a sensitive physiological indicator of nutrient stress and may be applied as an early diagnostic tool before visible growth inhibition occurs in broccoli cultivation. Full article

Non-invasive magnetic resonance imaging (MRI), as an extension of nuclear magnetic resonance (NMR) technology, enables detailed characterization of lithium-ion batteries (LIBs) in model systems. This review summarizes the fundamental principles of MRI and its applications in liquid/solid electrolytes, electrodes, and limited commercial diagnostics. Key capabilities include quantifying ion diffusion coefficients and mobility numbers in electrolytes, visualizing dendrite growth in lithium metal, and tracking lithium distribution in porous electrodes such as graphite and LiCoO₂. However, spatial and temporal resolution (typically 10–100 μm with acquisition times ranging from minutes to hours) and metal-induced shielding effects severely limit direct imaging in complete commercial batteries. Indirect methods like magnetic field imaging (MFI) show potential for defect detection. Future work should focus on sequence optimization and multimodal fusion, while emphasizing MRI’s primary role in fundamental research rather than conventional industrial testing. Full article

Phosphorus (P) deficiency severely limits maize growth and yield, yet early detection remains challenging, as visible symptoms appear only after prolonged starvation. This study evaluated the capability of hyperspectral imaging (HSI) combined with machine learning to detect P deficiency in maize seedlings at both symptomatic and pre-symptomatic stages. Two greenhouse experiments were conducted: a long-term pot system under high and low P conditions and a short-term hydroponic experiment with three P concentrations of 500, 100, and 0 μmol/L phosphate (Pi). After long-term P deficiency, significant reductions in shoot biomass and Pi content were observed, while root biomass increased and nutrient profiles were altered. Hyperspectral signatures revealed distinct wavelength-specific differences across visible, red-edge, and near-infrared (NIR) regions, with P-deficient leaves showing lower reflectance in green and NIR regions but higher reflectance in the red band. A multilayer perceptron machine learning model achieved 99.65% accuracy in discriminating between P treatments. In the short-term experiment, P deficiency significantly reduced tissue Pi content within one week without affecting pigment composition or photosynthetic parameters. Despite the absence of visible symptoms, hyperspectral measurements detected subtle spectral changes, particularly in older leaves, enabling classification accuracies of 80.71–84.56% in the first week and 85.88–90.98% in the second week of P treatment. Conventional vegetation indices showed weak correlations with Pi content and failed to detect early P deficiency. These findings demonstrate that HSI combined with machine learning can effectively detect P deficiency before visible symptoms emerge, offering a non-destructive, rapid diagnostic tool for precision nutrient management in maize production systems. Full article

Eggshell thickness is a critical indicator for evaluating egg breakage resistance and hatchability, yet traditional measurement methods remain destructive and inefficient. To address this, this study proposes a robust prediction approach by integrating Gradient Boosting Decision Tree (GBDT) feature optimization with an improved CatBoost algorithm. First, a joint strategy of Standard Normal Variate (SNV) and Multiplicative Scatter Correction (MSC) was employed to eliminate spectral scattering noise and enhance organic matrix fingerprint information. Subsequently, GBDT was introduced for nonlinear feature evaluation to adaptively screen the top 50 wavelengths, effectively mitigating the “curse of dimensionality” and multicollinearity in full-spectrum data. A CatBoost regression model was then constructed using an Ordered Boosting mechanism, supported by a dual anti-overfitting strategy that merged 10-fold nested cross-validation with Bootstrap resampling. Experimental results demonstrate that this method significantly outperforms traditional algorithms in both prediction accuracy and generalization. The coefficients of determination (R²) for the calibration and prediction sets reached 0.930 and 0.918, respectively, with a root mean square error of prediction (RMSEP) of 0.008 mm. Residual analysis confirms that prediction errors follow a zero-mean Gaussian distribution, indicating that systematic bias was effectively eliminated. This research provides a reliable theoretical foundation and technical support for the intelligent grading of poultry egg quality. Full article

Silver films are key building blocks for plasmonic and nanophotonic devices, whose optical performance and device reliability are highly sensitive to particulate contamination introduced during fabrication and operation. Herein, a non-destructive surface cleaning strategy specifically applicable to silver film systems is proposed, based on the synergistic regulation of the mechanical properties of a polymer layer and its interfacial adhesion to the silver film. Such regulation is achieved by tuning hydrogen-bond-mediated interactions within a modified poly(vinyl alcohol) (PVA) layer, enabling effective control over the locus of fracture during peeling, such that fracture preferentially occurs at the polymer/silver interface. Unlike conventional polymer-assisted cleaning methods that suffer from an inherent trade-off between bulk cohesion and interfacial adhesion, this approach decouples the two properties through molecular-level hydrogen-bond redistribution. As a result, particulate contaminants can be efficiently removed from the silver surface while preserving the structural integrity of the silver film. The proposed method achieves a particle removal efficiency of up to 98% for contaminants larger than 30 nm and can be stably applied to silver films with lateral dimensions ranging from 1 inch to 12 inches, demonstrating excellent scalability. By further adjusting the processing parameters and compositional ratios of the polymer layer, this strategy is expected to be adaptable to silver films with different thicknesses and structural configurations, providing a reliable surface cleaning solution for improving the performance and reliability of plasmonic and optoelectronic thin-film devices. Full article

Mold infection by Aspergillus and Penicillium spp. in Sithong sweet tamarind (Tamarindus indica L.) during commercial postharvest storage poses quality and food safety risks. However, the current visual detection method, which involves randomly cracking open the pods, is both destructive and laborious. The integration of near-infrared spectroscopy (NIRS) with artificial neural networks (ANN) enables rapid and non-destructive detection while capturing non-linear biochemical–spectral relationships, offering advantages over conventional destructive and linear analytical methods. It was tested as a mold classifier in sweet tamarind pods preserved in commercial ambient conditions (25 °C, 60% relative humidity) for five weeks. Six hundred pods were examined weekly using interactance spectroscopy (800–2500 nm) with six measurement points per pod and four spectral preprocessing methods. The ANN outperformed partial least squares discriminant analysis (PLS-DA) across all storage weeks, peaking at Week 2 with standard normal variate (SNV) preprocessing (prediction accuracy: 85.00%; sensitivity: 0.84; specificity: 0.86; F1-score: 0.85). Advanced tissue degeneration caused spectral heterogeneity, which decreased performance at Week 4 (prediction accuracy: 71.82–76.36%). Principal component loadings identified mold-induced water redistribution and carbohydrate depletion wavelengths at 938, 975–980, and 1035 nm. Week-adaptive calibration is essential for implementation because of the large difference between week-specific model accuracy (up to 85%) and overall storage model accuracy (63.53%). These findings provide a mechanistic underpinning for smaller wavelength-selective sensors and temporally adaptive mold screening systems in commercial tamarind storage. Full article

Coffee processing requires continuous optimization to preserve sensory quality while improving process efficiency. Although hyperspectral imaging has been widely applied for food quality evaluation, its use for predicting coffee cup score during controlled drying remains limited. This study aimed to evaluate the effect of drying temperature on the drying kinetics of Caturra coffee and to develop a predictive model for cup score using hyperspectral imaging combined with Partial Least Squares Regression (PLSR). Coffee samples were dried at four constant temperatures (30, 40, 50, and 60 °C) in forced-convection ovens, and hyperspectral reflectance images (400–1000 nm) were acquired using a Specim FX10 camera. Sensory evaluation was conducted by six certified Q Arabica Graders. Drying times were 52, 34, 30, and 20 h at 30, 40, 50, and 60 °C, respectively, with corresponding cup scores of 83.21, 83.50, 83.60, and 83.26 points. Effective moisture diffusivity ranged from ${10}^{-13}$ to ${10}^{-12}$ m²/s, while mass transfer coefficients were on the order of ${10}^{-9}$ m/s, with activation energies of 28.016 and 19.272 kJ/mol. No significant differences in cup score were observed among drying temperatures ($p>0.05$). A PLSR-based model was developed to estimate cup score from hyperspectral data, achieving $R^2$ values of 0.770 and 0.855 and RMSE values of 0.515 and 0.518 for calibration and validation, respectively. Key wavelengths at 480, 600, 720, and 940 nm were identified as the most influential spectral regions associated with chemical compounds affecting sensory quality. These findings demonstrate the potential of integrating drying kinetics and hyperspectral imaging as a rapid and non-destructive approach for objective prediction of coffee sensory quality during processing. Full article

As a branch of nondestructive testing (NDT), Pulsed Eddy Current Testing (PECT) is characterized by its wide frequency spectrum and high penetration depth. After years of development, it has been widely applied to defect detection and material characterization of key components in industries such as petrochemicals, new energy, and aerospace. With the large-scale application of new energy sources like liquefied natural gas (LNG), methanol, and liquid hydrogen, the demand for NDT of non-ferromagnetic materials (e.g., austenitic stainless steel) has surged. However, challenges such as electromagnetic leakage caused by low magnetic permeability and the lift-off effect induced by protective layers impose stricter requirements on inspection technologies, driving the evolution of PECT towards adaptability in complex scenarios. This paper systematically reviews the latest advances in PECT technology, covering detection sensors, modeling methods, detection signal processing, and engineering applications. With a particular emphasis on research outcomes from the past decade, this paper also proposes potential directions for future development, aiming to provide a reference for innovative research and the industrial promotion of PECT technology. Full article