Search Results (10,017)

Rice, a dietary staple for over half of the global population, is highly susceptible to bacterial and fungal diseases such as bacterial blight, brown spot, and leaf smut, which can severely reduce yields. Traditional manual detection is labor-intensive and often results in delayed intervention and excessive chemical use. Although deep learning models like convolutional neural networks (CNNs) achieve high accuracy, their computational demands hinder deployment in resource-limited agricultural settings. We propose MCP-X, an ultra-compact CNN with only 0.21 million parameters for real-time, on-device rice disease classification. MCP-X integrates a shallow encoder, multi-branch expert routing, a bi-level recurrent simulation encoder–decoder (BRSE), an efficient channel attention (ECA) module, and a lightweight classifier. Trained from scratch, MCP-X achieves 98.93% accuracy on PlantVillage and 96.59% on the Rice Disease Detection Dataset, without external pretraining. Mechanistically, expert routing diversifies feature branches, ECA enhances channel-wise signal relevance, and BRSE captures lesion-scale and texture cues—yielding complementary, stage-wise gains confirmed through ablation studies. Despite slightly higher FLOPs than MobileNetV2, MCP-X prioritizes a minimal memory footprint (~1.01 MB) and deployability over raw speed, running at 53.83 FPS (2.42 GFLOPs) on an RTX A5000. It achieves 16.7×, 287×, 420×, and 659× fewer parameters than MobileNetV2, ResNet152V2, ViT-Base, and VGG-16, respectively. When integrated into a multi-resolution ensemble, MCP-X attains 99.85% accuracy, demonstrating exceptional robustness across controlled and field datasets while maintaining efficiency for real-world agricultural applications. Full article

This paper presents the design and implementation of a vision-based score detection system tailored for smart IoT basketball applications. The proposed architecture leverages a compact, low-cost device comprising a high-resolution overhead camera and a Raspberry Pi 5 microprocessor equipped with a hardware accelerator for real-time object detection. The detection pipeline integrates convolutional neural networks (YOLO-based) and custom preprocessing techniques to localize the basketball hoop and track the ball trajectory. A scoring event is confirmed when the ball enters the defined scoring zone with downward motion over multiple frames, effectively reducing false positives caused by occlusions, multiple balls, or irregular shot directions. The system is part of a scalable IoT analytics platform known as Koško, which provides real-time statistics, leaderboards, and user engagement tools through a web-based interface. Field tests were conducted using data collected from various public and school courts across Niš, Serbia, resulting in a robust and adaptable solution for automated basketball score monitoring in both indoor and outdoor environments. The methodology supports edge computing, multilingual deployment, and integration with smart coaching and analytics systems. Full article

Background/Objectives: Muscle aging is a complex, dynamic process that impairs overall metabolism and physiological function. The molecular mechanisms underlying declines in muscle performance and metabolic efficiency remain poorly understood, largely due to the time and resource demands of traditional model organisms. The hawk moth Manduca sexta offers a promising alternative, with a short adult lifespan (~10 days) and notable similarities to vertebrate muscle systems, making it well-suited for time-course molecular dissection of muscle aging. Methods: In this study, we performed high-resolution temporal analysis of muscle tissues from aging M. sexta, spanning the physiomuscular aging process from middle age to advanced age. Results: We observed decreased expression of genes involved in fatty acid β-oxidation, ATP synthase subunits, superoxide dismutase, glutathione S-transferases, and heat shock proteins. In contrast, genes associated with proteolysis, catabolic processes, insulin signaling, akirin, titin, high-affinity choline transporters, and vesicular acetylcholine transporters were increased in expression. Conclusions: These changes suggest a shift toward increased proteolysis and protein catabolism with age. Our findings support the use of M. sexta as a complementary model for muscle aging research. However, it remains unclear whether the observed gene expression changes are driven by intrinsic, sex-specific age-related muscle aging or confounded by potential starvation effects in older males. Full article

LiDAR technology has undergone significant advancement in recent years, establishing itself as a technique for long-range, high-precision detection. As its use expands into more intricate scenarios, the need to overcome blind spots in the scanning field and enhance system stability has become increasingly critical. This paper introduces a novel coaxial LiDAR system featuring a double-clad optical fiber-based receiver which consists of a single-mode fiber core for the emission of the laser beam and a multimode inner cladding for the collection and transmission of the back-reflected beam. The real-time system is specifically engineered to measure distances in both near and far fields, eliminating blind spots. Experimental evaluations demonstrate that our system achieves a detection range of 0.2–70.7 m, with a distance accuracy of 3.4 cm and an angular resolution of 0.018°. Compared with conventional LiDAR systems, our approach eliminates the need for complex optical pathway designs and algorithmic compensation. It offers a simplified structure, enhanced stability, and high accuracy. Full article

Since March 2021, a series of volcanic eruptions on Iceland’s Reykjanes Peninsula has repeatedly triggered wildfires in moss-dominated heathlands—an unprecedented phenomenon in this environment. These fires have consumed extensive organic material, posing emerging health risks and long-term ecological impacts. Using high-resolution multispectral satellite data from the Copernicus program, we present the first quantitative assessment of the spatial and temporal dynamics of volcanic wildfire activity. Our analysis reveals a cumulative burned area extending 11.4 km² beyond the lava flows, primarily across low-relief terrain. Time series of the Normalized Difference Vegetation Index (NDVI) capture both localized fire scars and diffuse, landscape-scale burn patterns, followed by slow and spatially heterogeneous recovery. Complementary ground surveys conducted in August 2024 document diverse post-fire successional pathways, with vegetation regrowth and species composition strongly governed by microtopography and substrate texture. Together, these results demonstrate that volcanic wildfires represent a novel and consequential secondary disturbance in Icelandic volcanic systems, highlighting the complex and protracted recovery dynamics of moss heath ecosystems following fire-induced perturbation. Full article

Memory forensics is an essential cybersecurity tool that comprehensively examines volatile memory to detect the malicious activity of fileless malware that can bypass disk analysis. Image-based detection techniques provide a promising solution by visualizing memory data into images to be used and analyzed by image processing tools and machine learning methods. However, the effectiveness of image-based data for detection and classification requires high computational efforts. This paper investigates the efficacy of texture-based methods in detecting and classifying memory-resident or fileless malware using different image resolutions, identifying the best feature descriptors, classifiers, and resolutions that accurately classify malware into specific families and differentiate them from benign software. Moreover, this paper uses both local and global descriptors, where local descriptors include Oriented FAST and Rotated BRIEF (ORB), Scale-Invariant Feature Transform (SIFT), and Histogram of Oriented Gradients (HOG) and global descriptors include Discrete Wavelet Transform (DWT), GIST, and Gray Level Co-occurrence Matrix (GLCM). The results indicate that as image resolution increases, most feature descriptors yield more discriminative features but require higher computational efforts in terms of time and processing resources. To address this challenge, this paper proposes a novel approach that integrates Local Interpretable Model-agnostic Explanations (LIME) with deep learning models to automatically identify and crop the most important regions of memory images. The LIME’s ROI was extracted based on ResNet50 and MobileNet models’ predictions separately, the images were resized to 128 × 128, and the sampling process was performed dynamically to speed up LIME computation. The ROIs of the images are cropped to new images with sizes of (100 × 100) in two stages: the coarse stage and the fine stage. The two generated LIME-based cropped images using ResNet50 and MobileNet are fed to the lightweight neural network to evaluate the effectiveness of the LIME-based identified regions. The results demonstrate that the LIME-based MobileNet model’s prediction improves the efficiency of the model by preserving important features with a classification accuracy of 85% on multi-class classification. Full article

Optical tracking of unknown space objects requires both spatial accuracy and temporal stability to enable high-resolution identification through narrow field-of-view sensors. Traditional performance indices such as RMS error and persistence time (PT) have been used for controller tuning, but they each capture only a subset of the requirements for successful optical identification. This paper proposes a new composite metric, the Persistence-Weighted Tracking Index (PWTI), which combines spatial precision and segment-level continuity into a single measure. The metric assigns a frame-level score based on positional error and accumulates weighted scores over the longest continuous in-threshold segment. Using PWTI as the optimization objective, a genetic algorithm (GA) is employed to tune the PID gains of a frame-by-frame offset correction controller. Comparative simulations under various observation scenarios demonstrate that the PWTI-based approach outperforms RMS- and PT-based tuning methods in both alignment accuracy and consistency. The results validate the proposed metric as a more suitable performance indicator for optical identification tasks involving unknown or uncataloged targets. Full article

Accurate, scalable mapping of coral reef habitats is essential for monitoring ecosystem health and detecting change over time. In this study, we introduce a novel mathematically based nonlinear spectral unmixing method for benthic habitat classification, which provides sub-pixel estimates of benthic composition, capturing the mixed benthic composition within individual pixels. We compare its performance against two machine learning approaches: semi-supervised K-Means clustering and AdaBoost decision trees. All models were applied to high-resolution PlanetScope satellite imagery and ICESat-2-derived terrain metrics. Models were trained using a ground truth dataset constructed from benthic photoquadrats collected at Heron Reef, Australia, with additional input features including band ratios, standardized band differences, and derived ICESat-2 metrics such as rugosity and slope. While AdaBoost achieved the highest overall accuracy ($93.3\%$) and benefited most from ICESat-2 features, K-Means performed less well ($85.9\%$) and declined when these metrics were included. The spectral unmixing method uniquely captured sub-pixel habitat abundance, offering a more nuanced and ecologically realistic view of reef composition despite lower discrete classification accuracy ($64.8\%$). These findings highlight nonlinear spectral unmixing as a promising approach for fine-scale, transferable coral reef habitat mapping, especially in complex or heterogeneous reef environments. Full article

We developed a machine vision-based robotic system to address automation challenges in pharmaceutical pill sorting and packaging. The hardware platform integrates a high-resolution industrial camera with an HSR-CR605 robotic arm. Image processing leverages the VisionMaster 4.3.0 platform for color classification and positioning. Coordinate mapping between camera and robot is established through a three-point calibration method, with real-time communication realized via the Modbus/TCP protocol. Experimental validation demonstrates that the system achieves 95% recognition accuracy under conditions of pill overlap ≤ 30% and dynamic illumination of 50–1000 lux, ±0.5 mm picking precision, and a sorting efficiency of108 pills per minute. These results confirm the feasibility of integrating domestic hardware and algorithms, providing an efficient automated solution for the pharmaceutical industry. This work makes three key contributions: (1) demonstrating a cost-effective domestic hardware-software integration achieving 42% cost reduction while maintaining comparable performance to imported alternatives, (2) establishing a systematic validation methodology under industrially-relevant conditions that provides quantitative robustness metrics for pharmaceutical automation, and (3) offering a practical implementation framework validated through multi-scenario experiments that bridges the gap between laboratory research and production-line deployment. Full article

Family dogs stand out with regard to their special (human-like) attachment behavior towards their owners. This dog–owner attachment bond, analogous to the human infant–mother relationship, has been extensively documented at the behavioral level. Capitalizing on the fully non-invasive polysomnography protocol, the current study compares family dogs’ sleep structure when sleeping in the company of their owners versus an experimenter (a friendly stranger human). Subjects (N = 9) participated in three recording sessions, each lasting 3 h. The first session served as an adaptation to the recording environment, while the second and third were the test sessions analyzed for the present paper. On these two occasions, dogs slept, in a counterbalanced order, once in the company of their owner, while on the other occasion they slept in the company of an experimenter, while the owner was outside the room. Polysomnography recordings were used to extract high-resolution information (in 20 sec epochs) on the time dogs spend awake and in each of the sleep stages (drowsiness, non-REM, and REM). Our results show a robust difference between dogs’ sleep structure with and without the owner. In addition to an increased sleep latency and worsened sleep efficiency, dogs spent considerably less time in deep sleep (non-REM) when their owner was absent. These findings add to the increasing body of literature dealing with dog-to-owner attachment and provide unique physiological evidence for the phenomenon, complementing the widely reproduced behavioral data. Full article

Background: The growing demand for automated microorganism classification in the context of Laboratory 4.0 highlights the potential of convolutional neural networks (CNNs) for accurate and efficient image analysis. However, their effectiveness remains limited by the scarcity of large, labeled datasets. This study addresses a key gap in the literature by investigating how commonly used image preprocessing techniques, such as lossy compression, non-uniform scaling (typically applied to fit input images to CNN input layers), and data augmentation, affect the performance of CNNs in automated microorganism classification. Methods: Using two well-established CNN architectures, AlexNet and DenseNet-121, both frequently applied in biomedical image analysis, we conducted a series of computational experiments on a standardized dataset of high-resolution bacterial images. Results: Our results demonstrate under which conditions these preprocessing strategies degrade or improve CNN performance. Using the findings from this research to optimize hyperparameters and train the CNNs, we achieved classification accuracies of 98.61% with AlexNet and 99.82% with DenseNet-121, surpassing the performance reported in current state-of-the-art studies. Conclusions: This study advances laboratory digitalization by reducing data preparation effort, training time, and computational costs, while improving the accuracy of microorganism classification with deep learning. Its contributions also benefit broader biomedical fields such as automated diagnostics, digital pathology, clinical decision support, and point-of-care imaging. Full article

The rising global demand for high-quality animal protein has driven the development of advanced technologies in high-density livestock farming. Rabbits, with their rapid growth, high reproductive efficiency, and excellent feed conversion, play an important role in modern animal agriculture. However, large-scale rabbit farming poses challenges in timely health inspection and environmental monitoring. Traditional manual inspections are labor-intensive, prone-to-error, and inefficient for real-time management. To address these issues, we propose Livestock Environmental Monitoring System–Gantry (LIVEMOS-G), an intelligent gantry-based monitoring system tailored for large-scale rabbit farms. Inspired by plant phenotyping platforms, the system integrates a three-axis motion module with multi-source imaging (RGB, depth, near-infrared, thermal infrared) and an environmental sensing module. It autonomously inspects around the farm, capturing multi-angle, high-resolution images and real-time environmental data without disturbing the rabbits. Key environmental parameters are collected accurately and compared with welfare standards. After training on an original dataset, which contains a total of 2325 sets of images (each set includes RGB, NIR, TIR, and depth image), the system is able to detect dead rabbits using a fusion-based object detection model during inspections. LIVEMOS-G offers a scalable, non-intrusive solution for intelligent livestock inspection, contributing to enhanced biosecurity, animal welfare, and data-driven management in high-density, modern rabbit farms. It also shows the potential to be extended to other species, contributing to the sustainable development of the animal farming industry as a whole. Full article

Small intestine tumors are rare. The four main groups include adenocarcinomas, neuroendocrine neoplasms (NEN), lymphomas, and mesenchymal tumors. The jejunum and ileum can only be examined endoscopically with device-assisted enteroscopy techniques (DAET), which are indicated only when specific clinical or imaging findings are present. The initial diagnosis of tumors of the small intestine is mostly made using computed tomography (CT). Video capsule endoscopy (VCE), computed tomography (CT) enterography, and magnetic resonance (MR) enterography are also time-consuming and costly modalities. Modern transabdominal gastrointestinal ultrasound (US) with high-resolution transducers is a dynamic examination method that is underrepresented in the diagnosis of small intestine tumors. US can visualize wall thickening, loss of wall stratification, luminal stenosis, and dilatation of proximal small-intestinal segments, as well as associated lymphadenopathy. This review aims to highlight the role and imaging features of ultrasound in the diagnosis of small-intestinal tumors. Full article

Road infrastructure elements like guardrails, bollards, delineators, and traffic signs are critical for traffic safety but are significantly underrepresented in existing driving datasets, which primarily focus on vehicles and pedestrians. To address this crucial gap, we introduce DORIE (Dataset of Road Infrastructure Elements), a novel, high-resolution dataset specifically curated for real-time patrol vehicle monitoring along the A2 motorway in Spain. DORIE features 938 manually annotated images containing over 6800 object instances across ten safety-critical categories, including both static infrastructure and dynamic traffic participants. To establish a robust performance benchmark, we conducted an extensive evaluation of the YOLO family of detectors (versions 8, 11, and 12) across multiple scales and input resolutions. The results show that larger YOLO models and higher-resolution inputs yield up to 40% improvement in mean Average Precision (mAP) compared to smaller architectures, particularly for small and visually diverse classes such as traffic signs and bollards. The inference latency ranged between 5.7 and 245.2 ms per frame, illustrating the trade-off between detection accuracy and processing speed relevant to real-time operation. By releasing DORIE with detailed annotations and quantitative YOLO-based baselines, we provide a verifiable and reproducible resource to advance research in infrastructure monitoring and support the development of intelligent road safety and maintenance systems. Full article

The precise prediction of high-speed railway bridge pier settlement plays a crucial role in construction, maintenance, and long-term operation; however, current mainstream prediction methods mostly rely on independent analyses based on traditional or hybrid models, neglecting the impact of geological and environmental factors on subsidence. To address this issue, this paper proposes a multi-factor settlement prediction model for high-speed railway bridge piers named the Reversible Instance Normalization Multi-Scale Adaptive Resolution Stream CMamba, abbreviated as RMCMamba. During the data preprocessing process, the Enhanced PS-InSAR technology is adopted to obtain the time series data of land settlement in the study region. Utilizing the cubic improved Hermite interpolation method to fill the missing values of monitoring and considering the environmental parameters such as groundwater level, temperature, precipitation, etc., a multi-factor high-speed railway bridge pier settlement dataset is constructed. RMCMamba fuses the reversible instance normalization (RevIN) and the multiresolution forecasting head (MARSHead), enhancing the model’s long-range dependence capture capability and solving the time series data distribution drift problem. Experimental results demonstrate that in the multi-factor prediction scenario, RMCMamba achieves an MAE of 0.049 mm and an RMSE of 0.077 mm; in the single-factor prediction scenario, the proposed method reduces errors compared to traditional prediction approaches and other deep learning-based methods, with MAE values improving by 4.8% and 4.4% over the suboptimal method in multi-factor and single-factor scenarios, respectively. Ablation experiments further verify the collaborative advantages of combining reversible instance normalization and the multi-resolution forecasting head, as RMCMamba’s MAE values improve by 5.8% and 4.4% compared to the original model in multi-factor and single-factor scenarios. Hence, the proposed method effectively enhances the prediction accuracy of high-speed railway bridge pier settlement, and the constructed multi-source data fusion framework, along with the model improvement strategy, provides technological and experiential references for relevant fields. Full article