Search Results (235)

In shearography-based tire testing, so-called “Mode Hops”, abrupt phase changes caused by laser mode changes, can lead to significant disturbances in the interference image analysis. These artifacts distort defect assessment, lead to retesting or false-positive decisions, and, thus, represent a significant hurdle for the automation of the shearography-based tire inspection process. This work proposes a deep learning workflow that combines a pretrained, optimized ResNet-50 classifier with Grad-CAM, providing a practical and explainable solution for the reliable detection and localization of Mode Hops in shearographic tire inspection images. We trained the algorithm on an extensive, cross-machine dataset comprising more than 6.5 million test images. The final deep learning model achieves a classification accuracy of 99.67%, a false-negative rate of 0.48%, and a false-positive rate of 0.24%. Applying a probability-based quadrant-repeat decision rule within the inspection process effectively reduces process-level false positives to zero, with an estimated probability of repetition of ≤0.084%. This statistically validated approach increases the overall inspection accuracy to 99.83%. The method allows the robust detection and localization of relevant Mode Hops and represents a significant contribution to explainable, AI-supported tire testing. It fulfills central requirements for the automation of shearography-based tire testing and contributes to the possible certification process of non-destructive testing methods in safety-critical industries. Full article

Monitoring reclaimed wellsites in boreal forest environments requires accurate, scalable, and repeatable methods for assessing vegetation recovery. This study evaluates the use of uncrewed aerial vehicle (UAV)-based light detection and ranging (LiDAR) and multispectral (MS) imagery for individual tree detection, crown delineation, and classification across five reclaimed wellsites in Alberta, Canada. A deep learning workflow using 3D convolutional neural networks was applied to LiDAR and MS data collected in spring, summer, and autumn. Results show that LiDAR alone provided high accuracy for tree segmentation and height estimation, with a mean intersection over union (mIoU) of 0.94 for vegetation filtering and an F1-score of 0.82 for treetop detection. Incorporating MS data improved deciduous/coniferous classification, with the highest accuracy (mIoU = 0.88) achieved using all five spectral bands. Coniferous species were classified more accurately than deciduous species, and classification performance declined for trees shorter than 2 m. Spring conditions yielded the highest classification accuracy (mIoU = 0.93). Comparisons with ground measurements confirmed a strong correlation for tree height estimation (R² = 0.95; root mean square error = 0.40 m). Limitations of this technique included lower performance for short, multi-stemmed trees and deciduous species, particularly willow. This study demonstrates the value of integrating 3D structural and spectral data for monitoring forest recovery and supports the use of UAV remote sensing for scalable post-disturbance vegetation assessment. The trained models used in this study are publicly available through the TreeAIBox plugin to support further research and operational applications. Full article

Large-scale vegetation loss induced by surface coal mining constitutes a critical driver of regional ecological degradation. However, the applicability of existing change detection methodologies based on remote sensing within complex mining areas under diverse climatic conditions remains systematically unverified. To address this gap and reveal nationwide disturbance patterns, this study systematically evaluates the performance of two algorithms—Continuous Change Detection and Classification (CCDC) and Landsat-based Detection of Trends in Disturbance and Recovery (LandTrendr)—in identifying vegetation loss across three major climatic zones of China (the humid, semi-humid, and semi-arid zones). Based on the optimal algorithm, the vegetation loss year and loss magnitude across all of China’s surface coal mining areas from 1990 to 2020 were accurately identified, enabling the reconstruction of the comprehensive, nationwide spatio-temporal pattern of mining-induced vegetation loss over the past 30 years. The results show that: (1) CCDC demonstrated superior stability and significantly higher accuracy (OA = 0.82) than LandTrendr (OA = 0.31) in identifying loss years across all zones. (2) The cumulative vegetation loss area reached 1429.68 km², with semi-arid zones accounting for 86.76%. Temporal analysis revealed a continuous expansion of the loss area from 2003 to 2013, followed by a distinct inflection point and decline during 2014–2016 attributable to policy-driven regulations. (3) Further analysis revealed significant variations in the average magnitude of loss across different climatic zones, namely semi-arid (0.11), semi-humid (0.21), and humid (0.25). These findings underscore the imperative for region-specific restoration strategies to ensure effective conservation outcomes. This study provides a systematic quantification and analysis of long-term, nationwide evolution patterns and regional differentiation characteristics of vegetation loss induced by surface coal mining in China, offering critical support for sustainable development decision-making in balancing energy development and ecological conservation. Full article

In recent years, the availability of automated change detection algorithms in Google Earth Engine has permitted the cloud-based processing of large quantities of satellite imagery. Models such as the Continuous Change Detection and Classification (CCDC), CCDC-Spectral Mixture Analysis (CCDC-SMA), and Landsat-based Detection of Trends in Disturbance and Recovery (LandTrendr) allow users to exploit decades of Earth Observations (EOs), leveraging the Landsat archive and data from other sensors to detect disturbances in forest ecosystems. Despite the wide adoption of these methods, robust documentation, and a growing community of users, little research has systematically detailed their tuning process in mangrove environments. This work aims to identify the best practices for applying these models to monitor changes within mangrove forest cover, which has been declining gradually in Belize the last several decades. Partnering directly with the Belizean Forest Department, our team developed a replicable, efficient methodology to annually update the country’s mangrove extent, employing EO-based change detection. We ran a series of model variations in both CCDC-SMA and LandTrendr to identify the parameterizations best suited to identifying change in Belizean mangroves. Applying the best performing model run to the starting 2017 mangrove extent, we estimated a total loss of 540 hectares in mangrove coverage by 2024. Overall accuracy across thirty variations in model runs of LandTrendr and CCDC-SMA ranged from 0.67 to 0.75. While CCDC-SMA generally detected more disturbances and had higher precision for true changes, LandTrendr runs tended to have higher recall. Our results suggest LandTrendr offered more flexibility in balancing precision and recall for true changes compared to CCDC-SMA, due to its greater variety of adjustable parameters. Full article

Emergency braking detection plays a vital role in enhancing road safety by identifying potentially hazardous driving behaviors. While existing methods rely heavily on artificial intelligence and computationally intensive algorithms, this paper proposes a lightweight, real-time algorithm for distinguishing emergency braking from non-emergency events using accelerometer and gyroscope signals. The proposed approach applies magnitude calculations and a moving average filters algorithm to preprocess inertial data collected from a six-axis IMU sensor. By analyzing peak values of acceleration and angular velocity, the algorithm successfully separates emergency braking from other events such as regular braking, passing over speed bumps, or traversing damaged roads. The results demonstrate that emergency braking exhibits a unique short-pulse pattern in acceleration and low angular velocity, distinguishing it from other high-oscillation disturbances. Furthermore, varying the window length of the moving average impacts classification accuracy and computational cost. The proposed method avoids the complexity of neural networks while retaining high detection accuracy, making it suitable for embedded and real-time vehicular systems, such as early warning applications for fleet management. Full article

Induction motors (IM) play essential tasks in distinct production sectors because of their low cost and robustness. Considering that most of the energy demand in industry is allocated for powering up IM, recent research has focused on detecting and predicting faults to avoid severe disturbances. Broken rotor bars (BRB) in IM cause a significant deficit of energy, above all in those applications where constant changes in speed are required, increasing the probability of a catastrophic failure. Variable speed drives (VSD) introduce harmonic components to the power supply current controlling the IM rotating speed, which make it difficult to identify BRB. Therefore, in this work, an innovative methodology is proposed for detecting BRB in VSD-fed IM with a wide rotating-speed bandwidth during their start-up transient. The introduced procedure performs a statistical analysis for computing the mean, median, mode, variance, skewness, and kurtosis, to identify slight changes on the acquired current signal. These values are fed into an artificial neural network (ANN), which carries out the IM operational condition classification as healthy (HLT) or with BRB. Experimentally obtained results corroborate the effectiveness of the proposed approach to detecting BRB even for dynamically varying rotating speed, reaching a high accuracy of 99%, similar to recently reported techniques. Full article

Abandoned coal mine lands, particularly waste coal piles that predate the Surface Mining Control and Reclamation Act (SMCRA) of 1977, pose significant environmental and safety risks. Unlike sites mined after SMCRA—where operators are legally mandated to conduct reclamation—there is no legal obligation for companies or individuals to restore lands disturbed before the law’s enactment. As a result, these historical sites remain largely unmanaged and understudied. This study develops a satellite imagery-based analytical workflow to identify and monitor such historical waste coal piles. Using supervised classification of Sentinel-2 imagery with four machine learning models, we identified waste coal piles in both active mining areas and regions disturbed prior to SMCRA. Among the models tested, Random Forest achieved the highest accuracy for classifying waste coal, with a precision of 86% and a recall of 77%. A subsequent time-series analysis revealed that historical waste coal piles have undergone gradual but consistent vegetation recovery since 1986, indicating a natural reclamation process. These areas showed minimal changes in disturbance magnitude, suggesting the absence of significant disturbing events. In contrast, active mining regions showed substantial disturbance consistent with ongoing operations. The combined classification and change detection approach successfully distinguished historical waste coal piles from those in active mining regions, with a precision of 78% and recall of 100%. These findings highlight the potential of remote sensing and temporal analysis to support the identification and assessment of historical waste coal piles. The proposed approach can help prioritize reclamation efforts and inform policy decisions addressing the long-term environmental impacts of historical coal mining. Full article

Background: Brain tumors can severely impair neurological function, leading to symptoms such as headaches, memory loss, motor coordination deficits, and visual disturbances. In severe cases, they may cause permanent cognitive damage or become life-threatening without early detection. Methods: To address this, we propose an interpretable deep ensemble model for tumor detection in Magnetic Resonance Imaging (MRI) by integrating pre-trained Convolutional Neural Networks—EfficientNetB7, InceptionV3, and Xception—using a soft voting ensemble to improve classification accuracy. The framework is further enhanced with a Light Gradient Boosting Machine as a meta-learner to increase prediction accuracy and robustness within a stacking architecture. Hyperparameter tuning is conducted using Optuna, and overfitting is mitigated through batch normalization, L2 weight decay, dropout, early stopping, and extensive data augmentation. Results: These regularization strategies significantly enhance the model’s generalization ability within the BR35H dataset. The framework achieves a classification accuracy of 99.83 on the MRI dataset of 3060 images. Conclusions: To improve interpretability and build clinical trust, Explainable Artificial Intelligence methods Grad-CAM++, LIME, and SHAP are employed to visualize the factors influencing model predictions, effectively highlighting tumor regions within MRI scans. This establishes a strong foundation for further advancements in radiology decision support systems. Full article

Efficient and explainable vehicle damage inspection is essential due to the increasing complexity and volume of vehicular incidents. Traditional manual inspection approaches are not time-effective, prone to human error, and lead to inefficiencies in insurance claims and repair workflows. Existing deep learning methods, such as CNNs, often struggle with generalization, require large annotated datasets, and lack interpretability. This study presents a robust and interpretable deep learning framework for vehicle damage classification, integrating Vision Transformers (ViTs) and ensemble detection strategies. The proposed architecture employs a RetinaNet backbone with a ViT-enhanced detection head, implemented in PyTorch using the Detectron2 object detection technique. It is pretrained on COCO weights and fine-tuned through focal loss and aggressive augmentation techniques to improve generalization under real-world damage variability. The proposed system applies the Weighted Box Fusion (WBF) ensemble strategy to refine detection outputs from multiple models, offering improved spatial precision. To ensure interpretability and transparency, we adopt numerous explainability techniques—Grad-CAM, Grad-CAM++, and SHAP—offering semantic and visual insights into model decisions. A custom vehicle damage dataset with 4500 images has been built, consisting of approximately 60% curated images collected through targeted web scraping and crawling covering various damage types (such as bumper dents, panel scratches, and frontal impacts), along with 40% COCO dataset images to support model generalization. Comparative evaluations show that Hybrid ViT-RetinaNet achieves superior performance with an F1-score of 84.6%, mAP of 87.2%, and 22 FPS inference speed. In an ablation analysis, WBF, augmentation, transfer learning, and focal loss significantly improve performance, with focal loss increasing F1 by 6.3% for underrepresented classes and COCO pretraining boosting mAP by 8.7%. Additional architectural comparisons demonstrate that our full hybrid configuration not only maintains competitive accuracy but also achieves up to 150 FPS, making it well suited for real-time use cases. Robustness tests under challenging conditions, including real-world visual disturbances (smoke, fire, motion blur, varying lighting, and occlusions) and artificial noise (Gaussian; salt-and-pepper), confirm the model’s generalization ability. This work contributes a scalable, explainable, and high-performance solution for real-world vehicle damage diagnostics. Full article

Background/Objectives: Early diagnosis of Parkinson’s Disease (PD) is essential for initiating interventions that may slow its progression and enhance patient quality of life. Gait analysis provides a non-invasive means of capturing subtle motor disturbances, enabling the prediction of both disease presence and severity. This study evaluates and contrasts Bayesian-optimized convolutional neural network (CNN) and long short-term memory (LSTM) models applied directly to Vertical Ground Reaction Force (VGRF) signals for Parkinson’s disease detection and staging. Methods: VGRF recordings were segmented into fixed-length windows of 5, 10, 15, 20, and 25 s. Each segment was normalized and supplied as input to CNN and LSTM network. Hyperparameters for both architectures were optimized via Bayesian optimization using five-fold cross-validation. Results: The Bayesian-optimized LSTM achieved a peak binary classification accuracy of 99.42% with an AUC of 1.000 for PD versus control at the 10-s window, and 98.24% accuracy with an AUC of 0.999 for Hoehn–Yahr (HY) staging at the 5-s window. The CNN model reached up to 98.46% accuracy (AUC = 0.998) for binary classification and 96.62% accuracy (AUC = 0.998) for multi-class severity assessment. Conclusions: Bayesian-optimized CNN and LSTM models trained on VGRF data both achieved high accuracy in Parkinson’s disease detection and staging, with the LSTM exhibiting a slight edge in capturing temporal patterns while the CNN delivered comparable performance with reduced computational demands. These results underscore the promise of end-to-end deep learning for non-invasive, gait-based assessment in Parkinson’s disease. Full article

The increasing penetration of renewable energy into power systems has intensified transient power quality (PQ) disturbances, demanding efficient detection and classification methods to enable timely operational decisions. This paper introduces a hybrid framework combining an Improved Adaptive S-Transform (IAST) with a Random Forest (RF) classifier to address these challenges. The IAST employs a globally adaptive Gaussian window as its kernel function, which automatically adjusts window length and spectral resolution based on real-time frequency characteristics, thereby enhancing time–frequency localization accuracy while reducing algorithmic complexity. To optimize computational efficiency, window parameters are determined through an energy concentration maximization criterion, enabling rapid extraction of discriminative features from diverse PQ disturbances (e.g., voltage sags and transient interruptions). These features are then fed into an RF classifier, which simultaneously mitigates model variance and bias, achieving robust classification. Experimental results show that the proposed IAST–RF method achieves a classification accuracy of 99.73%, demonstrating its potential for real-time PQ monitoring in modern grids with high renewable energy penetration. Full article

Accurate crop row detection is crucial for determining weed pressure (weeds item per square meter). However, this task is complicated by the similarity between crops and weeds, the presence of missing plants within rows, and the varying growth stages of both. Our hypothesis was that in drone imagery captured at altitudes of 20–30 m—where individual plant details are not discernible—weed presence among crops can be statistically detected, allowing for the generation of a weed distribution map. This study proposes a computer vision detection method using images captured by unmanned aerial vehicles (UAVs) consisting of six main phases. The method was tested on 208 images. The algorithm performs well under normal conditions; however, when the weed density is too high, it fails to detect the row direction properly and begins processing misleading data. To investigate these cases, 120 artificial datasets were created with varying parameters, and the scenarios were analyzed. It was found that a rate variable—in-row concentration ratio (IRCR)—can be used to determine whether the result is valid (usable) or invalid (to be discarded). The F1 score is a metric combining precision and recall using a harmonic mean, where “1” indicates that precision and recall are equally weighted, i.e., β = 1 in the general Fβ formula. In the case of moderate weed infestation, where 678 crop plants and 600 weeds were present, the algorithm achieved an F1 score of 86.32% in plant classification, even with a 4% row disturbance level. Furthermore, IRCR also indicates the level of weed pressure in the area. The correlation between the ground truth weed-to-crop ratio and the weed/crop classification rate produced by the algorithm is 98–99%. As a result, the algorithm is capable of filtering out heavily infested areas that require full weed control and capable of generating weed density maps on other cases to support precision weed management. Full article

Castor fiber birulai is a subspecies of the Eurasian beaver that has a relatively small population size compared to other Castor subspecies. There is limited genetic research on this subspecies. In this study, mitochondrial cytochrome b (Cytb) and D-loop sequences were analysed in genetic samples obtained from 19 individuals residing in the Buergen River Basin, Xinjiang, China. The Cytb region presented a single haplotype, whereas three haplotypes were identified in the D-loop region. The genetic diversity within the Chinese population was low (D-loop Hd = 0.444; Pi = 0.0043), markedly lower than that observed in other geographical populations of C. fiber. Phylogenetic reconstructions and haplotype network analyses revealed substantial genetic differentiation between C. f. birulai and other Eurasian lineages (Fst > 0.95), supporting the status of C. f. birulai as a distinct evolutionary lineage. Although the genetic distance between the Chinese and Mongolian populations was relatively small (distance = 0.00269), significant genetic differentiation was detected (Fst = 0.67055), indicating that anthropogenic disturbances—such as hydraulic infrastructure and fencing along the cross-border Bulgan River—may have impeded gene flow and dispersal. Demographic analyses provided no evidence of recent population expansion (Fu’s Fs = 0.19152), suggesting a demographically stable population. In subsequent studies, we recommend increasing nuclear gene data to verify whether the C. f. birulai population meets the criteria for Evolutionarily Significant Unit classification, and strengthening cross-border protection and cooperation between China and Mongolia. Full article

Reliable and accurate docking remains a fundamental challenge for autonomous mobile robots (AMRs) operating in complex industrial environments with dynamic lighting, motion blur, and occlusion. This study proposes a novel multi-sensor fusion-based docking framework that significantly enhances robustness and precision by integrating YOLOv8-based AprilTag detection, depth-aided 3D localization, and LiDAR-based orientation correction. A key contribution of this work is the construction of a custom AprilTag dataset featuring real-world visual disturbances, enabling the YOLOv8 model to achieve high-accuracy detection and ID classification under challenging conditions. To ensure precise spatial localization, 2D visual tag coordinates are fused with depth data to compute 3D positions in the robot’s frame. A LiDAR group-symmetry mechanism estimates heading deviation, which is combined with visual feedback in a hybrid PID controller to correct angular errors. A finite-state machine governs the docking sequence, including detection, approach, yaw alignment, and final engagement. Simulation and experimental results demonstrate that the proposed system achieves higher docking success rates and improved pose accuracy under various challenging conditions compared to traditional vision- or LiDAR-only approaches. Full article

Depressive disorders significantly impact individuals worldwide, including children and adolescents. Despite their widespread occurrence, early and precise diagnosis of depressive disorders remains a complex and challenging task, particularly in younger populations. This study proposes a novel machine learning framework leveraging kinematic handwriting analysis to enhance the detection of psychomotor disturbances indicative of psychomotor retardation in youths with depression. The handwriting data were acquired from 20 youths with depression and 20 healthy controls. All participants completed a simple repetitive handwriting task: continuous writing of the small cursive Latin letter “l”. Segmentation of the handwriting data into individual “Letters” was conducted, and 177 kinematic features were extracted and analyzed. Statistical methods were used to identify significant features. After recursive feature elimination, classification was achieved through machine learning algorithms: logistic regression, support vector machine, and random forest. After the identification of 40 significant features, logistic regression, utilizing an optimal three-feature subset, achieved the highest accuracy in classifying individual letters of 76.7% and the highest accuracy in classifying subjects of 82.5%. The feature selection process revealed that velocity-related features were most effective in distinguishing patients with depression from controls, expectedly reflecting a slowdown in psychomotor functioning among the patients. The findings demonstrate that kinematic handwriting analysis, when combined with machine learning techniques, offers a promising tool to support objective recognition of psychomotor speed, providing insight into psychomotor retardation in youth with depression. Full article