Detection Method Research Articles

An improved cross algorithm edge detection method and its application in coal gangue identification

ABSTRACT In order to solve the problems of edge detection error, image segmentation distortion and low recognition accuracy in coal gangue separation under complex background conditions, an improved cross algorithm edge detection method is proposed in this paper, and it is applied in an example of coal gangue recognition. The cross algorithm is used to detect the edges of coal and coal gangue images, and the detection results are processed by morphological technology to obtain the segmentation results of single coal and coal gangue images. The mean value, contrast and entropy value of coal and coal gangue are extracted as recognition features to construct a support vector machine (SVM) recognition model. The accuracy of SVM identification model is 96.67% for coal and 100% for coal gangue, and it is obviously superior to K-Nearest Neighbor (KNN) identification model, Probabilistic Neural Network (PNN) identification model and Back Propagation Neural Network (BP) identification model in time. This study provides a new method to solve the identification problem of coal and gangue.

Utilizing large language models for detecting hospital-acquired conditions: an empirical study on pulmonary embolism.

Utilizing large language models for detecting hospital-acquired conditions: an empirical study on pulmonary embolism.

Comparison of Immunomagnetic Bead Separation-Immunosensor Detection and Nested-PCR Methods for Detecting Mycobacterium avium Subspecies paratuberculosis in Cattle Feces.

Johne's disease, also known as paratuberculosis, is a chronic granulomatous enteritis disease that affects ruminants worldwide. The objective of this study was to assess the effectiveness of the immunomagnetic bead separation-immunosensor (IMB-IS) detection method compared to Nested-PCR for identifying Mycobacterium avium subspecies paratuberculosis (MAP) infection in cattle feces samples. Ninety rectal fecal samples were collected from selected cattle, comprising 59 serum-positive and 31 serum-negative cases based on serum ELISA. Following DNA extraction, nested-PCR was conducted using the IS900 primer sequence targeting the MAP-specific gene. Immunomagnetic bead (IMB) nanoparticles were synthesized by purifying hyperimmune donkey IgG through affinity chromatography and then conjugating it to Fe nanoparticles. Rhodamine-B hydrazone immunosensor (IS) was synthesized and conjugated to hyperimmune rabbit IgG. The synthesized IMB and IS were used to identify MAP in cattle fecal samples. The results of this study revealed that of the 90 stool samples tested using the nested-PCR method, 62 samples (68.88%) were positive, while 28 samples (31.12%) were negative. In the IMB-IS test based on optical density (OD), 64 samples were positive (71.1%), while 26 samples were negative (28.8%). This test exhibited a sensitivity of 100%, specificity of 92.85%, and an overall test accuracy of 97.77%. Given the considerations of cost, time, positive and negative predictive values, and acceptable accuracy of the IMB-IS test, it is recommended for evaluation in screening and epidemiological studies.

Electrochemical Sensor Based on Black Phosphorus for Antimony Detection Using Dip-Pen Nanolithography: The Role of Dwell Time.

Heavy metals, including Sb, are major pollutants with limits on their allowed concentration in drinking water. Therefore, there is a need for sensitive, simple, and portable detection methods for which electrochemical sensors are ideally suited. In this current study, Meta-chemical surfaces are developed for electrochemical sensing by patterning gold electrode surfaces with a mixture of black phosphorus (BP) and polymethyl methacrylate (PMMA) as nanoclusters using dip-pen nanolithography. It is found that the surface-to-volume ratio (S/V), fill factor, and ink composition affect the sensitivity of the sensor for Sb detection. The S/V ratio and fill factor can be altered by the dwell time, which has a complex effect on the limit of detection (varying from 14 to 24 ppb with the changes in the dwell time). Density functional theory calculations show that the binding between Sb(III) and BP is more exergonic in the presence of PMMA. These results are significant because they allow for the development of more sensitive Sb sensors, which can affect the wider field of the detection of heavy metals in drinking water sources and achieve higher efficiency than the commonly used instruments.

Quantitative Lipoprotein Subclass Analysis in Pig Plasma by 1H NMR Spectroscopy and Stability Assessment.

This study explores, for the first time, the accurate prediction of lipoprotein subclasses in pig plasma by a partial least-squares regression (PLSR) model based on the optimization of the 1H NMR detection method. The 1H NMR-based detection method of plasma metabolites first was optimized and evaluated. The coefficients of variation for intraday and interday detection were less than 5%, and there were no obvious metabolic differences among repeated tests of plasma samples. The variability of plasma metabolite detection based on 1H NMR revealed the consistent and stable performance of NMR methods, and the prediction models for a total of 116 subclasses in four lipoprotein classes including plasma very-low-density lipoprotein, low-density lipoprotein, intermediate-density lipoprotein, and high-density lipoprotein were established combining 1H NMR spectra and PLSR models using the data from the ultracentrifugation method. The cross-validation results showed that the PLSR prediction models for 107 lipoproteins' main and subfractions performed excellently (R2 > 0.5), which met the method requirements. The PLSR models for the remaining nine lipoprotein major components and subcomponents performed well (0.2 < R2 < 0.5), which basically met the method requirements. According to the PLSR models based on 1H NMR, the concentrations of lipoprotein subclasses were predicted, including APO A1 (277.84-731.4 μg/mL), APO B (11.97-431.5 μg/mL), PL (142.36-1100.76 ng/mL), TG (70.21-915.35 μmol/L), CH (1313.56-6761.79 μmol/L), FC (6.37-93.06 μmol/L), and CE (1319.93-6854.85 μmol/L). Therefore, the 1H NMR-based method for the detection of lipoprotein subclasses in pig plasma was successfully established and could provide the methodological basis for the research on molecular mechanism, function, and application of lipoprotein subclasses.

A Review: One-Shot Object Detection Methods for Conditional Detection of Retail and Warehouse Products

Facilitating the rapid dispatch and replenishment of products is a critical task in most large warehouses. Automated systems often rely on Deep Learning based object detection methods to monitor operations. A major challenge is the significant requirement for annotated data, and the system’s difficulty in adapting to new products. In contrast, human operators can quickly learn to recognize and adapt to new products with just a single example. This survey focuses on methods that enable conditional detection using a single support example per class. We first introduce common feature fusion techniques and discuss datasets suitable for warehouse and retail products. Next, we provide a comprehensive overview of the current State-Of-The-Art in One-Shot Object Detection. We categorize these approaches based on their detectors, which identify the object’s bounding boxes and classes. Then, we delve into detailed implementations of these methods, analyzing how they leverage this innovative vision approach to improve performance. Finally, we identify promising current trends in this emerging field.

Relation-based self-distillation method for 2D object detection

The challenge of enhancing the detection accuracy of widely adopted and stable object detectors while maintaining cost-effectiveness has long been a topic of significant interest and concern within the industry. To address this challenge, this paper proposes a general relation-based self-distillation framework suitable for object detection to help existing detectors achieve a better balance between accuracy and overhead. Compared to existing self-distillation methods, the framework we propose focuses on integrating relation-based knowledge into the self-distillation framework. To achieve this goal, we propose a relation-based self-distillation method within the framework and design a targeted optimization strategy in the form of an adaptive filtering strategy. The relation-based self-distillation method constrains the detector from focusing on the differences in the representation of the same type of object in different scenarios; and the adaptive filtering strategy filters the low-confidence results predicted by the detector before calling the matching mechanism, thereby ensuring the efficiency of the training process. A large number of experimental results show that our method can significantly improve the accuracy of existing detectors and reduce their redundant prediction results without increasing the computational resource overhead of existing detectors.

Emerging Approaches for Studying Lipid Dynamics, Metabolism, and Interactions in Cells.

Lipids are a major class of biological molecules, the primary components of cellular membranes, and critical signaling molecules that regulate cell biology and physiology. Due to their dynamic behavior within membranes, rapid transport between organelles, and complex and often redundant metabolic pathways, lipids have traditionally been considered among the most challenging biological molecules to study. In recent years, a plethora of tools bridging the chemistry-biology interface has emerged for studying different aspects of lipid biology. Here, we provide an overview of these approaches. We discuss methods for lipid detection, including genetically encoded biosensors, synthetic lipid analogs, and metabolic labeling probes. For targeted manipulation of lipids, we describe pharmacological agents and controllable enzymes, termed membrane editors, that harness optogenetics and chemogenetics. To conclude, we survey techniques for elucidating lipid-protein interactions, including photoaffinity labeling and proximity labeling. Collectively, these strategies are revealing new insights into the regulation, dynamics, and functions of lipids in cell biology.

Evaluation of how US women react to a decision aid informing them of the harms and benefits of mammography: a qualitative study

ObjectivesTo evaluate how women react to a decision aid (DA) informing them of the harms and benefits of mammography, including overdiagnosis.BackgroundBreast cancer is the most common cancer affecting women. In...

Broiler Behavior Detection and Tracking Method Based on Lightweight Transformer

Broiler Behavior Detection and Tracking Method Based on Lightweight Transformer

Rapid diagnosis of Diaporthe helianthi in sunflower using RPA/CRISPR-Cas12 and lateral flow assay

Diaporthe helianthi is one of the main fungal pathogens responsible for causing phomopsis stem canker and significant yield losses of sunflowers. In this study, the calmodulin (Cal) gene of D. helianthi was selected to develop a rapid detection method involving recombinase polymerase amplification (RPA) combined with CRISPR/Cas12a-based detection at 37 °C. The developed detection system could complete the specific detection of D. helianthi in 45 min, including 25 min for RPA and 20 min for CRISPR/Cas12 reaction. The detection system could be coupled with both lateral flow test strips and fluorescence signal reading modes. The detection limit for lateral flow assay was 1 pg/μL genomic D. helianthi DNA (14 copies/µL); The detection limit for fluorescence signal was 0.1 pg/μL genomic DNA (1.4 copies/µL), approximately 100 times higher than that of the real-time PCR. Thus, the developed RPA/CRISPR-Cas12a system meets the need for portable detection of D. helianthi on-site at ports and in the field.

CRISPR/Cas12a-powered nanoconfined biosensing platform with hybrid chain reaction cascading guanine nanowire amplification for ultrasensitive dual-mode detection of lipopolysaccharide.

Traditional endotoxin detection methods face challenges in sensitivity, interference resistance, and reliability. This study develops a CRISPR/Cas12a-powered nanoconfined biosensing system that integrates mesoporous nanoengineering with a hybrid chain reaction (HCR) cascading guanine nanowire (G-wire) dual amplification strategy for ultrasensitive dual-mode detection of lipopolysaccharide (LPS). By leveraging a vertically ordered mesoporous silica membrane (VMSM) as a molecular sieve and CRISPR trans-cleavage activity modulator, the system achieves precise regulation of Ru(bpy)32+ adsorption via LPS-suppressed HCR assembly. This architecture enables physical confinement-mediated electrochemiluminescence (ECL) and fluorescence (FL) signal transduction, with dual-mode outputs providing mutual validation for enhanced reliability. The biosensor exhibits superior sensitivity with detection limits of 3.4 pg mL-1 for ECL and 1.4 pg mL-1 for FL, while also offering a broad dynamic range (0.005-100 ng mL-1), significantly outperforming conventional LPS assays. The CRISPR-triggered HCR cascading G-wire dual amplification synergizes with nanoconfinement of VMSM to ensure robust anti-interference performance in complex matrices, validated by recovery rates of 97.8-102.5% in real samples. By integrating CRISPR programmability with nanoengineered signal amplification, this work establishes a transformative paradigm for portable, high-precision endotoxin detection in clinical diagnostics, industrial monitoring, and environmental safety applications.

Ship pipeline defect detection method based on deep learning and transfer fusion of ultrasonic guided wave signals

Ship pipeline defect detection method based on deep learning and transfer fusion of ultrasonic guided wave signals

A Ratiometric Fluorescent Probe Based on Carbon Dots and Quantum Dots for Glucose Detection.

Glucose is a critical energy source for human cells, and maintaining stable glucose levels is essential for normal physiological functions. Abnormal glucose levels can lead to health issues such as diabetes, obesity, and hypoglycemia, highlighting the need for precise and rapid glucose detection methods for clinical diagnostics and health management. In this study, a novel ratiometric fluorescent probe was constructed for glucose detection. Using glucose oxidase (GOx) and horseradish peroxidase (HRP) as enzymatic mediators, the probe employs orange-emitting carbon dots (oCDs) as the detection signal and red-emitting quantum dots (rQDs) as the reference signal. Glucose is oxidized by GOx to produce hydrogen peroxide (H2O2), which subsequently generates hydroxyl radicals (·OH) under HRP catalysis. The ·OH interacts electrostatically with oCDs, forming non-fluorescent complexes and quenching the oCDs fluorescence. Glucose concentration is quantified by monitoring the fluorescence intensity ratio (I590/I715) between oCDs and rQDs. The probe has a limit of detection (LOD) of 0.47 µM and a limit of quantification (LOQ) of 1.58 µM. After methodological validation, it has been successfully applied to the detection of glucose in human serum. This study provides a solid basis for accurate glucose monitoring and holds potential for the early diagnosis and treatment of metabolic diseases such as diabetes.

Vibration signal analysis for rolling bearings faults diagnosis based on deep-shallow features fusion

In engineering applications, the bearing faults diagnosis is essential for maintaining reliability and extending the lifespan of rotating machinery, thereby preventing unexpected industrial production downtime. Prompt fault diagnosis using vibration signals is vital to ensure seamless operation of industrial system avert catastrophic breakdowns, reduce maintenance costs, and ensure continuous productivity. As industries evolve and machines operate under diverse conditions, traditional fault detection methods often fall short. In spite of significant research in recent years, there remains a pressing need for improve existing methods of fault diagnosis. To fill this research gap, this research work aims to propose an efficient and robust system for diagnosing bearing faults, using deep and Shallow features. Through the evaluated experiments, our proposed model Multi-Block Histograms of Local Phase Quantization (MBH-LPQ) showed excellent performance in classification accuracy, and the audio-trained VGGish model showed the best performance in all tasks. Contributions of this work include: Combine the proposed Shallow descriptor, derived from a novel hand-crafted discriminative features MBH-LPQ, with deep features obtained from VGGish pre-trained of Convolutional Neural Network (CNN) using audio spectrograms, by merging at the score level using Weighted Sum (WS). This approach is designed to take advantage of the complementary strengths of both feature models, thus enhancing overall bearing fault diagnostic performance. Furthermore, experiments conducted to verify the approach’s performance is assessed based on fault classification accuracy demonstrated a significant accuracy rate on two different noisy datasets, with an accuracy rate of 98.95% and 100% being reached on the CWRU and PU datasets benchmark, respectively.

Establishment of an RAA-CRISPR/Cas12a-based diagnostic method for the detection of fowl adenovirus serotype 4 virus in chickens and wild birds

Fowl adenovirus (FAdV) serotype 4, recognized as the causative agent of hydropericardium syndrome (HPS) in chickens, causes substantial economic losses in poultry farming. To develop a simple, rapid, and reliable diagnostic method for the timely detection of FAdV-4 nucleic acid, we integrated the CRISPR/Cas12a system with recombinase-aided amplification (RAA). This approach enables visual detection of FAdV-4 with a sensitivity of one genome copy. The results can be obtained within 40 to 50 min without the need for complex instrumentation, making it ideal for remote field applications. Using this method, we investigated the prevalence of FAdV-4 in both common farm poultry and wild birds. Our results indicated that the FAdV-4-positive rate in wild birds was 51.19%, suggesting that wild birds may serve as specific reservoirs for this virus. In summary, we present a sensitive, swift, accurate, and inexpensive detection method for FAdV-4, along with an investigation of its epidemic situation in birds. Our study advances the detection and epidemiological understanding of FAdV-4 transmission among farm poultry and wild birds.

Generative adversarial message passing-based anomaly detection

Generative adversarial message passing-based anomaly detection

CLRNetV2: A Faster and Stronger Lane Detector.

Lane is critical in the vision navigation system of intelligent vehicles. Naturally, the lane is a traffic sign with high-level semantics, whereas it owns the specific local pattern which needs detailed low-level features to localize accurately. Using different feature levels is of great importance for accurate lane detection, but it is still under-explored. On the other hand, current lane detection methods still struggle to detect complex dense lanes, such as Y-shape or fork-shape. In this work, we present Cross Layer Refinement Network aiming at fully utilizing both high-level and low-level features in lane detection. In particular, it first detects lanes with high-level semantic features and then performs refinement based on low-level features. In this way, we can exploit more contextual information to detect lanes while leveraging local-detailed features to improve localization accuracy. We present Fast-ROIGatherto gather global context, which further enhances the representation of lane features. To detect dense lanes accurately, we propose Correlation Discrimination Module (CDM) to discriminate the correlation of dense lanes, enabling nearly cost-free high-quality dense lane prediction. In addition to our novel network design, we introduce LineIoU loss which regresses lanes as a whole unit to improve localization accuracy. Experiments demonstrate our approach significantly outperforms the state-of-the-art lane detection methods.

Online high speed detection method for end face quality of cored filter rods under complex conditions

Aiming at the current high-speed real-time inspection demand faced in the production of filter rods with cores and the limitations of traditional quality control methods, this study proposes a high-speed online inspection method based on machine vision. As the production speed of filter rods increases, the traditional inspection methods cannot meet the real-time and accuracy requirements in a high-speed environment, while the restricted production environment and more disturbing factors lead to an increase in the detection error. Therefore, this paper proposes an efficient visual inspection method, which combines the ellipse fitting algorithm to accurately obtain the coordinates of the filter rods and their mandrel centre points, and evaluates the mandrel deviation through the Euclidean distance calculation to judge the product qualification. The experimental results show that the system is capable of inspecting at a rate of 4200 pcs per minute in a complex environment with a defect detection rate of more than 95%. The technology significantly improves the inspection efficiency and product quality of the filter rod production line, providing a reliable intelligent inspection solution for the filter rod manufacturing industry.

Fault detection for high-speed train traction system using autoencoder-Fréchet inception distance

Abstract The detection of the process state in the traction system plays a crucial role in ensuring the safe operation of high-speed trains. However, the running data in the traction system exhibits nonlinear and dynamic nature, which leads to the failure of traditional detection methods. To address the above problems, a novel fault detection method based on autoencoder-Fréchet inception distance (AE-FID) is proposed in this paper. Firstly, autoencoder (AE) is used for feature extraction to capture the inherent nonlinear structure in the data. Secondly, a novel fault detection statistic is developed by combining sliding window technology and Fréchet inception distance (FID) to eliminate the influence of data dynamics. Finally, the proposed statistic is utilized to detect faults in the traction system, and the occurrence of faults is measured in terms of structural differences to mitigate the influence of the masking effects. The proposed method is evaluated through a numerical example that simulates dynamic nonlinear systems and the Traction Drive Control System-Fault Injection Benchmark simulation platform of a high-speed train. The experimental results demonstrate the effectiveness of the proposed method in fault detection for traction systems.