Next Issue
Volume 24, December-1
Previous Issue
Volume 24, November-1
 
 
sensors-logo

Journal Browser

Journal Browser

Sensors, Volume 24, Issue 22 (November-2 2024) – 315 articles

Cover Story (view full-size image): Ecological surveys of living things based on DNA gathered from environmental samples are attractive. To perform DNA-oriented surveys based on a simple protocol without any special training, we demonstrated, in this study, the detection of genes from cell-containing environmental waters using gene sensor arrays that require no DNA labeling and no external indicators. Cell-suspended PBS or river water were used as models of environmental waters containing living things. A sensor array was prepared by immobilizing ferrocene-terminated probes on an electrode array. The sensor array showed a large sequence-specific response to a target DNA. They also significantly detected DNA samples from the cells in river water at a general environmental concentration (38 cells mL−1) with 28-fold larger responses than those for 0 cells mL−1View this paper
  • Issues are regarded as officially published after their release is announced to the table of contents alert mailing list.
  • You may sign up for e-mail alerts to receive table of contents of newly released issues.
  • PDF is the official format for papers published in both, html and pdf forms. To view the papers in pdf format, click on the "PDF Full-text" link, and use the free Adobe Reader to open them.
Order results
Result details
Section
Select all
Export citation of selected articles as:
15 pages, 2414 KiB  
Article
Interaction of Sensitivity, Emotions, and Motivations During Visual Perception
by Sergey Lytaev
Sensors 2024, 24(22), 7414; https://doi.org/10.3390/s24227414 - 20 Nov 2024
Viewed by 874
Abstract
When an organism is exposed to environmental stimuli of varying intensity, the adaptive changes in the CNS can be explained by several conceptual provisions: the law of motivation–activation by Yerkes and Dodson, the laws of force and pessimal protective inhibition, and the theory [...] Read more.
When an organism is exposed to environmental stimuli of varying intensity, the adaptive changes in the CNS can be explained by several conceptual provisions: the law of motivation–activation by Yerkes and Dodson, the laws of force and pessimal protective inhibition, and the theory of emotion activation. Later, reinforcement sensitivity theory was developed in the fields of psychology and psychophysics. At the same time, cortical prepulse inhibition (PPI), the prepulse inhibition of perceived stimulus intensity (PPIPSI), and the augmentation/reduction phenomenon were proposed in sensory neurophysiology, which expanded our understanding of consciousness. The aim of this study was to identify markers of levels of activity of cognitive processes under normal and in psychopathological conditions while amplifying the information stimulus. For this purpose, we changed the contrast level of reversible checkerboard patterns and mapped the visual evoked potentials (VEPs) in 19 monopolar channels among 52 healthy subjects and 39 patients with a mental illness without an active productive pathology. Their cognitive functions were assessed by presenting visual tests to assess invariant pattern recognition, short-term visual memory, and Gestalt perception. The personalities of the healthy subjects were assessed according to Cattell’s 16-factor questionnaire, linking the data from neurophysiological and cognitive studies to factors Q4 (relaxation/tension) and C (emotional stability). According to the N70 and N150 VEP waves, the healthy subjects and the patients were divided into two groups. In some, there was a direct relationship between VEP amplitude and contrast intensity (21 patients and 29 healthy persons), while in the others, there was an inverse relationship, with a reduction in VEP amplitude (18 patients and 23 healthy persons). The relationship and mechanisms of subjects’ cognitive abilities and personality traits are discussed based on the data acquired from the responses to information stimuli of varied intensity. Full article
(This article belongs to the Section Intelligent Sensors)
Show Figures

Figure 1

18 pages, 6146 KiB  
Article
A Near-Infrared Imaging System for Robotic Venous Blood Collection
by Zhikang Yang, Mao Shi, Yassine Gharbi, Qian Qi, Huan Shen, Gaojian Tao, Wu Xu, Wenqi Lyu and Aihong Ji
Sensors 2024, 24(22), 7413; https://doi.org/10.3390/s24227413 - 20 Nov 2024
Viewed by 723
Abstract
Venous blood collection is a widely used medical diagnostic technique, and with rapid advancements in robotics, robotic venous blood collection has the potential to replace traditional manual methods. The success of this robotic approach is heavily dependent on the quality of vein imaging. [...] Read more.
Venous blood collection is a widely used medical diagnostic technique, and with rapid advancements in robotics, robotic venous blood collection has the potential to replace traditional manual methods. The success of this robotic approach is heavily dependent on the quality of vein imaging. In this paper, we develop a vein imaging device based on the simulation analysis of vein imaging parameters and propose a U-Net+ResNet18 neural network for vein image segmentation. The U-Net+ResNet18 neural network integrates the residual blocks from ResNet18 into the encoder of the U-Net to form a new neural network. ResNet18 is pre-trained using the Bootstrap Your Own Latent (BYOL) framework, and its encoder parameters are transferred to the U-Net+ResNet18 neural network, enhancing the segmentation performance of vein images with limited labelled data. Furthermore, we optimize the AD-Census stereo matching algorithm by developing a variable-weight version, which improves its adaptability to image variations across different regions. Results show that, compared to U-Net, the BYOL+U-Net+ResNet18 method achieves an 8.31% reduction in Binary Cross-Entropy (BCE), a 5.50% reduction in Hausdorff Distance (HD), a 15.95% increase in Intersection over Union (IoU), and a 9.20% increase in the Dice coefficient (Dice), indicating improved image segmentation quality. The average error of the optimized AD-Census stereo matching algorithm is reduced by 25.69%, and the improvement of the image stereo matching performance is more obvious. Future research will explore the application of the vein imaging system in robotic venous blood collection to facilitate real-time puncture guidance. Full article
(This article belongs to the Section Sensors and Robotics)
Show Figures

Figure 1

16 pages, 2963 KiB  
Article
An Entropy-Based Clustering Algorithm for Real-Time High-Dimensional IoT Data Streams
by Ibrahim Mutambik
Sensors 2024, 24(22), 7412; https://doi.org/10.3390/s24227412 - 20 Nov 2024
Viewed by 454
Abstract
The rapid growth of data streams, propelled by the proliferation of sensors and Internet of Things (IoT) devices, presents significant challenges for real-time clustering of high-dimensional data. Traditional clustering algorithms struggle with high dimensionality, memory and time constraints, and adapting to dynamically evolving [...] Read more.
The rapid growth of data streams, propelled by the proliferation of sensors and Internet of Things (IoT) devices, presents significant challenges for real-time clustering of high-dimensional data. Traditional clustering algorithms struggle with high dimensionality, memory and time constraints, and adapting to dynamically evolving data. Existing dimensionality reduction methods often neglect feature ranking, leading to suboptimal clustering performance. To address these issues, we introduce E-Stream, a novel entropy-based clustering algorithm for high-dimensional data streams. E-Stream performs real-time feature ranking based on entropy within a sliding time window to identify the most informative features, which are then utilized with the DenStream algorithm for efficient clustering. We evaluated E-Stream using the NSL-KDD dataset, comparing it against DenStream, CluStream, and MR-Stream. The evaluation metrics included the average F-Measure, Jaccard Index, Fowlkes–Mallows Index, Purity, and Rand Index. The results show that E-Stream outperformed the baseline algorithms in both clustering accuracy and computational efficiency while effectively reducing dimensionality. E-Stream also demonstrated significantly less memory consumption and fewer computational requirements, highlighting its suitability for real-time processing of high-dimensional data streams. Despite its strengths, E-Stream requires manual parameter adjustment and assumes a consistent number of active features, which may limit its adaptability to diverse datasets. Future work will focus on developing a fully autonomous, parameter-free version of the algorithm, incorporating mechanisms to handle missing features and improving the management of evolving clusters to enhance robustness and adaptability in dynamic IoT environments. Full article
(This article belongs to the Special Issue Advances in Big Data and Internet of Things)
Show Figures

Figure 1

18 pages, 13728 KiB  
Article
BG-YOLO: A Bidirectional-Guided Method for Underwater Object Detection
by Ruicheng Cao, Ruiteng Zhang, Xinyue Yan and Jian Zhang
Sensors 2024, 24(22), 7411; https://doi.org/10.3390/s24227411 - 20 Nov 2024
Viewed by 939
Abstract
Degraded underwater images decrease the accuracy of underwater object detection. Existing research uses image enhancement methods to improve the visual quality of images, which may not be beneficial in underwater image detection and lead to serious degradation in detector performance. To alleviate this [...] Read more.
Degraded underwater images decrease the accuracy of underwater object detection. Existing research uses image enhancement methods to improve the visual quality of images, which may not be beneficial in underwater image detection and lead to serious degradation in detector performance. To alleviate this problem, we proposed a bidirectional guided method for underwater object detection, referred to as BG-YOLO. In the proposed method, a network is organized by constructing an image enhancement branch and an object detection branch in a parallel manner. The image enhancement branch consists of a cascade of an image enhancement subnet and object detection subnet. The object detection branch only consists of a detection subnet. A feature-guided module connects the shallow convolution layers of the two branches. When training the image enhancement branch, the object detection subnet in the enhancement branch guides the image enhancement subnet to be optimized towards the direction that is most conducive to the detection task. The shallow feature map of the trained image enhancement branch is output to the feature-guided module, constraining the optimization of the object detection branch through consistency loss and prompting the object detection branch to learn more detailed information about the objects. This enhances the detection performance. During the detection tasks, only the object detection branch is reserved so that no additional computational cost is introduced. Extensive experiments demonstrate that the proposed method significantly improves the detection performance of the YOLOv5s object detection network (the mAP is increased by up to 2.9%) and maintains the same inference speed as YOLOv5s (132 fps). Full article
(This article belongs to the Special Issue Machine Learning in Image/Video Processing and Sensing)
Show Figures

Figure 1

15 pages, 6297 KiB  
Article
Numerical Simulation Study on Predicting the Critical Icing Conditions of Aircraft Pitot Tubes
by Qixi Chen, Lifen Zhang, Chengxin Zhou, Zhengang Liu and Yaguo Lyu
Sensors 2024, 24(22), 7410; https://doi.org/10.3390/s24227410 - 20 Nov 2024
Viewed by 467
Abstract
Aircraft pitot tubes are sophisticated instruments designed to detect airflow pressure and relay this information to onboard computers and flight instruments, enabling the calculation of airspeed through the measurement of total-static pressure differences. The formation of ice on aircraft pitot tubes can compromise [...] Read more.
Aircraft pitot tubes are sophisticated instruments designed to detect airflow pressure and relay this information to onboard computers and flight instruments, enabling the calculation of airspeed through the measurement of total-static pressure differences. The formation of ice on aircraft pitot tubes can compromise the acquisition of airspeed data, misguide pilots, and potentially cause catastrophic flight control failures. This article introduces a predictive methodology for identifying critical conditions that lead to icing on aircraft pitot tubes. Utilizing numerical simulation techniques, the methodology calculates the critical conditions for pitot tube icing across cruise flight regimes and atmospheric conditions, resulting in the generation of a critical condition envelope surface. By comparing these critical conditions against actual sensor data, a predictive danger zone can be established, offering an advanced warning system to ensure flight safety. Full article
(This article belongs to the Section Intelligent Sensors)
Show Figures

Figure 1

15 pages, 5384 KiB  
Article
Gradual Failure of a Rainfall-Induced Creep-Type Landslide and an Application of Improved Integrated Monitoring System: A Case Study
by Jun Guo, Fanxing Meng and Jingwei Guo
Sensors 2024, 24(22), 7409; https://doi.org/10.3390/s24227409 - 20 Nov 2024
Viewed by 425
Abstract
Landslides cause severe damage to life and property with a wide-ranging impact. Infiltration of rainfall is one of the significant factors leading to landslides. This paper reports on a phase creep landslide caused by long-term rainfall infiltration. A detailed geological survey of the [...] Read more.
Landslides cause severe damage to life and property with a wide-ranging impact. Infiltration of rainfall is one of the significant factors leading to landslides. This paper reports on a phase creep landslide caused by long-term rainfall infiltration. A detailed geological survey of the landslide was conducted, and the deformation development pattern and mechanism of the landslide were analyzed in conjunction with climatic characteristics. Furthermore, reinforcement measures specific to the landslide area were proposed. To monitor the stability of the reinforced slope, a Beidou intelligent monitoring and warning system suitable for remote mountainous areas was developed. The system utilizes LoRa Internet of Things (IoT) technology to connect various monitoring components, integrating surface displacement, deep deformation, structural internal forces, and rainfall monitoring devices into a local IoT network. A data processing unit was established on site to achieve preliminary processing and automatic handling of monitoring data. The monitoring results indicate that the reinforced slope has generally stabilized, and the improved intelligent monitoring system has been able to continuously and accurately reflect the real-time working conditions of the slope. Over the two-year monitoring period, 13 early warnings were issued, with more than 90% of the warnings accurately corresponding to actual conditions, significantly improving the accuracy of early warnings. The research findings provide valuable experience and reference for the monitoring and warning of high slopes in mountainous areas. Full article
(This article belongs to the Section Internet of Things)
Show Figures

Figure 1

23 pages, 938 KiB  
Article
An Efficient Flow-Based Anomaly Detection System for Enhanced Security in IoT Networks
by Ibrahim Mutambik
Sensors 2024, 24(22), 7408; https://doi.org/10.3390/s24227408 - 20 Nov 2024
Viewed by 1140
Abstract
The growing integration of Internet of Things (IoT) devices into various sectors like healthcare, transportation, and agriculture has dramatically increased their presence in everyday life. However, this rapid expansion has exposed new vulnerabilities within computer networks, creating security challenges. These IoT devices, often [...] Read more.
The growing integration of Internet of Things (IoT) devices into various sectors like healthcare, transportation, and agriculture has dramatically increased their presence in everyday life. However, this rapid expansion has exposed new vulnerabilities within computer networks, creating security challenges. These IoT devices, often limited by their hardware constraints, lack advanced security features, making them easy targets for attackers and compromising overall network integrity. To counteract these security issues, Behavioral-based Intrusion Detection Systems (IDS) have been proposed as a potential solution for safeguarding IoT networks. While Behavioral-based IDS have demonstrated their ability to detect threats effectively, they encounter practical challenges due to their reliance on pre-labeled data and the heavy computational power they require, limiting their practical deployment. This research introduces the IoT-FIDS (Flow-based Intrusion Detection System for IoT), a lightweight and efficient anomaly detection framework tailored for IoT environments. Instead of employing traditional machine learning techniques, the IoT-FIDS focuses on identifying unusual behaviors by examining flow-based representations that capture standard device communication patterns, services used, and packet header details. By analyzing only benign traffic, this network-based IDS offers a streamlined and practical approach to securing IoT networks. Our experimental results reveal that the IoT-FIDS can accurately detect most abnormal traffic patterns with minimal false positives, making it a feasible security solution for real-world IoT implementations. Full article
(This article belongs to the Special Issue IoT Cybersecurity)
Show Figures

Figure 1

19 pages, 4058 KiB  
Article
Enhanced Fault Detection in Photovoltaic Panels Using CNN-Based Classification with PyQt5 Implementation
by Younes Ledmaoui, Adila El Maghraoui, Mohamed El Aroussi and Rachid Saadane
Sensors 2024, 24(22), 7407; https://doi.org/10.3390/s24227407 - 20 Nov 2024
Viewed by 668
Abstract
Solar photovoltaic systems have increasingly become essential for harvesting renewable energy. However, as these systems grow in prevalence, the issue of the end of life of modules is also increasing. Regular maintenance and inspection are vital to extend the lifespan of these systems, [...] Read more.
Solar photovoltaic systems have increasingly become essential for harvesting renewable energy. However, as these systems grow in prevalence, the issue of the end of life of modules is also increasing. Regular maintenance and inspection are vital to extend the lifespan of these systems, minimize energy losses, and protect the environment. This paper presents an innovative explainable AI model for detecting anomalies in solar photovoltaic panels using an enhanced convolutional neural network (CNN) and the VGG16 architecture. The model effectively identifies physical and electrical changes, such as dust and bird droppings, and is implemented using the PyQt5 Python tool to create a user-friendly interface that facilitates decision-making for users. Key processes included dataset balancing through oversampling and data augmentation to expand the dataset. The model achieved impressive performance metrics: 91.46% accuracy, 98.29% specificity, and an F1 score of 91.67%. Overall, it enhances power generation efficiency and prolongs the lifespan of photovoltaic systems, while minimizing environmental risks. Full article
(This article belongs to the Special Issue Sensor Enabled Smart Energy Solutions)
Show Figures

Figure 1

13 pages, 2588 KiB  
Article
Construction of Flexible Deterministic Sparse Measurement Matrix in Compressed Sensing Using Legendre Sequences
by Haiqiang Liu, Ming Li and Caiping Hu
Sensors 2024, 24(22), 7406; https://doi.org/10.3390/s24227406 - 20 Nov 2024
Viewed by 422
Abstract
Compressed sensing (CS) is an innovative signal acquisition and reconstruction technology that has broken through the limit of the Nyquist sampling theory. It is widely employed to optimize the measurement processes in various applications. One of the core challenges of CS is the [...] Read more.
Compressed sensing (CS) is an innovative signal acquisition and reconstruction technology that has broken through the limit of the Nyquist sampling theory. It is widely employed to optimize the measurement processes in various applications. One of the core challenges of CS is the construction of a measurement matrix. However, traditional random measurement matrices are often impractical. Additionally, many existing deterministic binary measurement matrices fail to provide the required flexibility for practical applications. In this study, inspired by the observation that pseudo-random sequences share similar properties with random sequences, we constructed a deterministic sparse measurement matrix with a flexible measurement number based on an pseudo-random sequence—the Legendre sequence. Empirical analysis of the phase transition and an assessment of the practical features of the proposed measurement matrix were conducted. We validated the effectiveness of the proposed measurement matrix on randomly synthesized signals and images. The results of our simulations reveal that our proposed measurement matrix performs better than several other measurement matrices, particularly in terms of accuracy and efficiency. Full article
(This article belongs to the Section Physical Sensors)
Show Figures

Figure 1

19 pages, 414 KiB  
Article
Quantum Privacy-Preserving Range Query Protocol for Encrypted Data in IoT Environments
by Chong-Qiang Ye, Jian Li and Xiao-Yu Chen
Sensors 2024, 24(22), 7405; https://doi.org/10.3390/s24227405 - 20 Nov 2024
Viewed by 467
Abstract
With the rapid development of IoT technology, securely querying sensitive data collected by devices within a specific range has become a focal concern for users. This paper proposes a privacy-preserving range query scheme based on quantum encryption, along with circuit simulations and performance [...] Read more.
With the rapid development of IoT technology, securely querying sensitive data collected by devices within a specific range has become a focal concern for users. This paper proposes a privacy-preserving range query scheme based on quantum encryption, along with circuit simulations and performance analysis. We first propose a quantum private set similarity comparison protocol and then construct a privacy-preserving range query scheme for IoT environments. By leveraging the properties of quantum homomorphic encryption, the proposed scheme enables encrypted data comparisons, effectively preventing the leakage of sensitive data. The correctness and security analysis demonstrates that the designed protocol guarantees users receive the correct query results while resisting both external and internal attacks. Moreover, the protocol requires only simple quantum states and operations, and does not require users to bear the cost of complex quantum resources, making it feasible under current technological conditions. Full article
(This article belongs to the Special Issue IoT Network Security (Second Edition))
Show Figures

Figure 1

25 pages, 3415 KiB  
Article
An Innovative Neighbor Attention Mechanism Based on Coordinates for the Recognition of Facial Expressions
by Cheng Peng, Bohao Li, Kun Zou, Bowen Zhang, Genan Dai and Ah Chung Tsoi
Sensors 2024, 24(22), 7404; https://doi.org/10.3390/s24227404 - 20 Nov 2024
Viewed by 585
Abstract
For solving the facial expression recognition (FER) problem, we introduce a novel feature extractor called the coordinate-based neighborhood attention mechanism (CNAM), which uses the coordinate attention (CA) method to capture the directional relationships in separate horizontal and vertical directions, the input features from [...] Read more.
For solving the facial expression recognition (FER) problem, we introduce a novel feature extractor called the coordinate-based neighborhood attention mechanism (CNAM), which uses the coordinate attention (CA) method to capture the directional relationships in separate horizontal and vertical directions, the input features from a preprocessing unit, and then passes this to two residual blocks, one consisting of the neighborhood attention (NA) mechanism, which captures the local interaction of features within the neighborhood of a feature vector, while the other one contains a channel attention implemented by a multilayer perceptron (MLP). We apply the feature extractor, the CNAM module, to four FER benchmark datasets, namely, RAF-DB, AffectNet(7cls), AffectNet(8cls), and CK+, and through qualitative and quantitative analysis techniques, we conclude that the insertion of the CNAM module could decrease the intra-cluster distances and increase the inter-cluster distances among the high-dimensional feature vectors. The CNAM compares well with other state-of-the-art (SOTA) methods, being the best-performing method for the AffectNet(7cls) and CK+ datasets, while for the RAF-DB and AffectNet(8cls) datasets, its performance is among the top-performing SOTA methods. Full article
(This article belongs to the Section Sensing and Imaging)
Show Figures

Figure 1

16 pages, 9667 KiB  
Article
Development of a Conceptual Scheme for Controlling Tool Wear During Cutting, Based on the Interaction of Virtual Models of a Digital Twin and a Vibration Monitoring System
by Lapshin Viktor, Turkin Ilya, Gvindzhiliya Valeriya, Dudinov Ilya and Gamaleev Denis
Sensors 2024, 24(22), 7403; https://doi.org/10.3390/s24227403 - 20 Nov 2024
Viewed by 431
Abstract
This article discusses the issue of the joint use of neural network algorithms for data processing and deterministic mathematical models. The use of a new approach is proposed, to determine the discrepancy between data from a vibration monitoring system of the cutting process [...] Read more.
This article discusses the issue of the joint use of neural network algorithms for data processing and deterministic mathematical models. The use of a new approach is proposed, to determine the discrepancy between data from a vibration monitoring system of the cutting process and the calculated data obtained by modeling mathematical models of the digital twin system of the cutting process. This approach is justified by the fact that some coordinates for the state of the cutting process cannot be measured, and the vibration signals measured by the vibration monitoring system (the vibration acceleration of the tip of the cutting tool) are subject to external disturbing influences. Both the experimental method and the Matlab 2022b simulation method were used as research methods. The experimental research method is based on the widespread use of modern analog vibration transducers, the signals from which undergo the process of digitalization and further processing in order to identify arrays of additional information required for virtual digital twin models. The results obtained allow us to formulate a new conceptual approach to the construction of systems for determining the degree of cutting tool wear, based on the joint use of computational virtual models of the digital twin system and data obtained from the vibration monitoring system of the cutting process. Full article
(This article belongs to the Section Physical Sensors)
Show Figures

Figure 1

20 pages, 1628 KiB  
Review
Energy Efficiency for 5G and Beyond 5G: Potential, Limitations, and Future Directions
by Adrian Ichimescu, Nirvana Popescu, Eduard C. Popovici and Antonela Toma
Sensors 2024, 24(22), 7402; https://doi.org/10.3390/s24227402 - 20 Nov 2024
Viewed by 675
Abstract
Energy efficiency constitutes a pivotal performance indicator for 5G New Radio (NR) networks and beyond, and achieving optimal efficiency necessitates the meticulous consideration of trade-offs against other performance parameters, including latency, throughput, connection densities, and reliability. Energy efficiency assumes it is of paramount [...] Read more.
Energy efficiency constitutes a pivotal performance indicator for 5G New Radio (NR) networks and beyond, and achieving optimal efficiency necessitates the meticulous consideration of trade-offs against other performance parameters, including latency, throughput, connection densities, and reliability. Energy efficiency assumes it is of paramount importance for both User Equipment (UE) to achieve battery prologue and base stations to achieve savings in power and operation cost. This paper presents an exhaustive review of power-saving research conducted for 5G and beyond 5G networks in recent years, elucidating the advantages, disadvantages, and key characteristics of each technique. Reinforcement learning, heuristic algorithms, genetic algorithms, Markov Decision Processes, and the hybridization of various standard algorithms inherent to 5G and 5G NR represent a subset of the available solutions that shall undergo scrutiny. In the final chapters, this work identifies key limitations, namely, computational expense, deployment complexity, and scalability constraints, and proposes a future research direction by theoretically exploring online learning, the clustering of the network base station, and hard HO to lower the consumption of networks like 2G or 4G. In lowering carbon emissions and lowering OPEX, these three additional features could help mobile network operators achieve their targets. Full article
Show Figures

Figure 1

17 pages, 88951 KiB  
Article
Sec-CLOCs: Multimodal Back-End Fusion-Based Object Detection Algorithm in Snowy Scenes
by Rui Gong, Xiangsuo Fan, Dengsheng Cai and You Lu
Sensors 2024, 24(22), 7401; https://doi.org/10.3390/s24227401 - 20 Nov 2024
Viewed by 597
Abstract
LiDAR and cameras, often regarded as the “eyes” of intelligent driving vehicles, are vulnerable to adverse weather conditions like haze, rain, and snow, compromising driving safety. In order to solve this problem and enhance the environmental sensing capability under severe weather conditions, this [...] Read more.
LiDAR and cameras, often regarded as the “eyes” of intelligent driving vehicles, are vulnerable to adverse weather conditions like haze, rain, and snow, compromising driving safety. In order to solve this problem and enhance the environmental sensing capability under severe weather conditions, this paper proposes a multimodal back-end fusion object detection method, Sec-CLOCs, which is specifically optimized for vehicle detection under heavy snow. This method achieves object detection by integrating an improved YOLOv8s 2D detector with a SECOND 3D detector. First, the quality of image data is enhanced through the Two-stage Knowledge Learning and Multi-contrastive Regularization (TKLMR) image processing algorithm. Additionally, the DyHead detection head and Wise-IOU loss function are introduced to optimize YOLOv8s and improve 2D detection performance.The LIDROR algorithm preprocesses point cloud data for the SECOND detector, yielding 3D object detection results. The CLOCs back-end fusion algorithm is then employed to merge the 2D and 3D detection outcomes, thereby enhancing overall object detection capabilities. The experimental results show that the Sec-CLOCs algorithm achieves a vehicle detection accuracy of 82.34% in moderate mode (30–100 m) and 81.76% in hard mode (more than 100 m) under heavy snowfall, which demonstrates the algorithm’s high detection performance and robustness. Full article
(This article belongs to the Section Sensing and Imaging)
Show Figures

Graphical abstract

20 pages, 1947 KiB  
Article
Pressure Control of Multi-Mode Variable Structure Electro–Hydraulic Load Simulation System
by He Hao, Hao Yan, Qi Zhang and Haoyu Li
Sensors 2024, 24(22), 7400; https://doi.org/10.3390/s24227400 - 20 Nov 2024
Viewed by 553
Abstract
During the loading process, significant external position disturbances occur in the electro–hydraulic load simulation system. To address these position disturbances and effectively mitigate the impact of uncertainty on system performance, this paper first treats model parameter uncertainty and external disturbances as lumped disturbances. [...] Read more.
During the loading process, significant external position disturbances occur in the electro–hydraulic load simulation system. To address these position disturbances and effectively mitigate the impact of uncertainty on system performance, this paper first treats model parameter uncertainty and external disturbances as lumped disturbances. The various states of the servo valve and the pressures within the hydraulic cylinder chambers are then examined. Building on this foundation, the paper proposes a nonlinear multi-mode variable structure independent load port electro–hydraulic load simulation system that is tailored for specific loading conditions. Secondly, in light of the significant motion disturbances present, this paper proposes an integral sliding mode active disturbance rejection composite control strategy that is based on fixed-time convergence. Based on the structure of the active disturbance rejection control framework, the fixed-time integral sliding mode and active disturbance rejection control algorithms are integrated. An extended state observer is designed to accurately estimate the lumped disturbance, effectively compensating for it to achieve precise loading of the independent load port electro–hydraulic load simulation system. The stability of the designed controller is also demonstrated. The results of the simulation research indicate that when the command input is a step signal, the pressure control accuracy under the composite control strategy is 99.94%, 99.86%, and 99.76% for disturbance frequencies of 1 Hz, 3 Hz, and 5 Hz, respectively. Conversely, when the command input is a sinusoidal signal, the pressure control accuracy remains high, measuring 99.94%, 99.8%, and 99.6% under the same disturbance frequencies. Furthermore, the simulation demonstrates that the influence of sensor random noise on the system remains within acceptable limits, highlighting the effective filtering capabilities of the extended state observer. This research establishes a solid foundation for the collaborative control of load ports and the engineering application of the independent load port electro–hydraulic load simulation system. Full article
(This article belongs to the Section Electronic Sensors)
Show Figures

Figure 1

18 pages, 6536 KiB  
Article
Strong Interference Elimination in Seismic Data Using Multivariate Variational Mode Extraction
by Zhichao Yu, Yuyang Tan and Yiran Lv
Sensors 2024, 24(22), 7399; https://doi.org/10.3390/s24227399 - 20 Nov 2024
Viewed by 445
Abstract
Seismic data acquired in the presence of mechanical vibrations or power facilities may be contaminated by strong interferences, significantly decreasing the data signal-to-noise ratio (S/N). Conventional methods, such as the notch filter and time-frequency transform method, are usually inadequate for suppressing non-stationary interference [...] Read more.
Seismic data acquired in the presence of mechanical vibrations or power facilities may be contaminated by strong interferences, significantly decreasing the data signal-to-noise ratio (S/N). Conventional methods, such as the notch filter and time-frequency transform method, are usually inadequate for suppressing non-stationary interference noises, and may distort effective signals if overprocessing. In this study, we propose a method for eliminating mechanical vibration interferences in seismic data. In our method, we extended the variational mode extraction (VME) technique to a multivariate form, called multivariate variational mode extraction (MVME), for synchronous analysis of multitrace seismic data. The interference frequencies are determined via synchrosqueezing-based time-frequency analysis of process recordings; their corresponding modes are extracted and removed from seismic data using MVME with optimal balancing factors. We used synthetic data to investigate the effectiveness of the method and the influence of tuning parameters on processing results, and then applied the method to field datasets. The results have demonstrated that, compared with the conventional methods, the proposed method could effectively suppress the mechanical vibration interferences, improve the S/Ns and enhance polarization analysis of seismic signals. Full article
(This article belongs to the Special Issue Sensor Technologies for Seismic Monitoring)
Show Figures

Figure 1

17 pages, 4619 KiB  
Article
Efficient Video Compression Using Afterimage Representation
by Minseong Jeon and Kyungjoo Cheoi
Sensors 2024, 24(22), 7398; https://doi.org/10.3390/s24227398 - 20 Nov 2024
Viewed by 423
Abstract
Recent advancements in large-scale video data have highlighted the growing need for efficient data compression techniques to enhance video processing performance. In this paper, we propose an afterimage-based video compression method that significantly reduces video data volume while maintaining analytical performance. The proposed [...] Read more.
Recent advancements in large-scale video data have highlighted the growing need for efficient data compression techniques to enhance video processing performance. In this paper, we propose an afterimage-based video compression method that significantly reduces video data volume while maintaining analytical performance. The proposed approach utilizes optical flow to adaptively select the number of keyframes based on scene complexity, optimizing compression efficiency. Additionally, object movement masks extracted from keyframes are accumulated over time using alpha blending to generate the final afterimage. Experiments on the UCF-Crime dataset demonstrated that the proposed method achieved a 95.97% compression ratio. In binary classification experiments on normal/abnormal behaviors, the compressed videos maintained performance comparable to the original videos, while in multi-class classification, they outperformed the originals. Notably, classification experiments focused exclusively on abnormal behaviors exhibited a significant 4.25% improvement in performance. Moreover, further experiments showed that large language models (LLMs) can interpret the temporal context of original videos from single afterimages. These findings confirm that the proposed afterimage-based compression technique effectively preserves spatiotemporal information while significantly reducing data size. Full article
Show Figures

Figure 1

20 pages, 7387 KiB  
Article
Skin Phototype Classification with Machine Learning Based on Broadband Optical Measurements
by Xun Yu, Keat Ghee Ong and Michael Aaron McGeehan
Sensors 2024, 24(22), 7397; https://doi.org/10.3390/s24227397 - 20 Nov 2024
Viewed by 518
Abstract
The Fitzpatrick Skin Phototype Classification (FSPC) scale is widely used to categorize skin types but has limitations such as the underrepresentation of darker skin phototypes, low classification resolution, and subjectivity. These limitations may contribute to dermatological care disparities in patients with darker skin [...] Read more.
The Fitzpatrick Skin Phototype Classification (FSPC) scale is widely used to categorize skin types but has limitations such as the underrepresentation of darker skin phototypes, low classification resolution, and subjectivity. These limitations may contribute to dermatological care disparities in patients with darker skin phototypes, including the misdiagnosis of wound healing progression and escalated dermatological disease severity. This study introduces (1) an optical sensor measuring reflected light across 410–940 nm, (2) an unsupervised K-means algorithm for skin phototype classification using broadband optical data, and (3) methods to optimize classification across the Near-ultraviolet-A, Visible, and Near-infrared spectra. The differentiation capability of the algorithm was compared to human assessment based on FSPC in a diverse participant population (n = 30) spanning an even distribution of the full FSPC scale. The FSPC assessment distinguished between light and dark skin phototypes (e.g., FSPC I vs. VI) at 560, 585, and 645 nm but struggled with more similar phototypes (e.g., I vs. II). The K-means algorithm demonstrated stronger differentiation across a broader range of wavelengths, resulting in better classification resolution and supporting its use as a quantifiable and reproducible method for skin type classification. We also demonstrate the optimization of this method for specific bandwidths of interest and their associated clinical implications. Full article
(This article belongs to the Special Issue Novel Optical Sensors for Biomedical Applications—2nd Edition)
Show Figures

Figure 1

16 pages, 9195 KiB  
Article
Simulating and Verifying a 2D/3D Laser Line Sensor Measurement Algorithm on CAD Models and Real Objects
by Rok Belšak, Janez Gotlih and Timi Karner
Sensors 2024, 24(22), 7396; https://doi.org/10.3390/s24227396 - 20 Nov 2024
Viewed by 461
Abstract
The increasing adoption of 2D/3D laser line sensors in industrial and research applications necessitates accurate and efficient simulation tools for tasks such as surface inspection, dimensional verification, and quality control. This paper presents a novel algorithm developed in MATLAB for simulating the measurements [...] Read more.
The increasing adoption of 2D/3D laser line sensors in industrial and research applications necessitates accurate and efficient simulation tools for tasks such as surface inspection, dimensional verification, and quality control. This paper presents a novel algorithm developed in MATLAB for simulating the measurements of any 2D/3D laser line sensor on STL CAD models. The algorithm uses a modified fast-ray triangular intersection method, addressing challenges such as overlapping triangles in assembly models and incorporating sensor resolution to ensure realistic simulations. Quantitative analysis shows a significant reduction in computation time, enhancing the practical utility of the algorithm. The simulation results exhibit a mean deviation of 0.42 mm when compared to real-world measurements. Notably, the algorithm effectively handles complex geometric features, such as holes and grooves, and offers flexibility in generating point cloud data in both local and global coordinate systems. This work not only reduces the need for physical prototyping, thereby contributing to sustainability, but also supports AI training by generating accurate synthetic data. Future work should aim to further optimize the simulation speed and explore noise modeling to enhance the realism of simulated measurements. Full article
(This article belongs to the Special Issue Advances in Optical Sensing, Instrumentation and Systems: 2nd Edition)
Show Figures

Figure 1

27 pages, 6231 KiB  
Review
A Review of Unmanned Aerial Vehicle Based Antenna and Propagation Measurements
by Venkat R. Kandregula, Zaharias D. Zaharis, Qasim Z. Ahmed, Faheem A. Khan, Tian Hong Loh, Jason Schreiber, Alexandre Jean René Serres and Pavlos I. Lazaridis
Sensors 2024, 24(22), 7395; https://doi.org/10.3390/s24227395 - 20 Nov 2024
Viewed by 780
Abstract
This paper presents a comprehensive survey of state-of-the-art UAV–based antennas and propagation measurements. Unmanned aerial vehicles (UAVs) have emerged as powerful tools for in situ electromagnetic field assessments due to their flexibility, cost-effectiveness, and ability to operate in challenging environments. This paper highlights [...] Read more.
This paper presents a comprehensive survey of state-of-the-art UAV–based antennas and propagation measurements. Unmanned aerial vehicles (UAVs) have emerged as powerful tools for in situ electromagnetic field assessments due to their flexibility, cost-effectiveness, and ability to operate in challenging environments. This paper highlights various UAV applications, from testing large–scale antenna arrays, such as those used in the square kilometer array (SKA), to evaluating channel models for 5G/6G networks. Additionally, the review discusses technical challenges, such as positioning accuracy and antenna alignment, and it provides insights into the latest advancements in portable measurement systems and antenna designs tailored for UAV use. During the UAV–based antenna measurements, key contributors to the relatively small inaccuracies of around 0.5 to 1 dB are identified. In addition to factors such as GPS positioning errors and UAV vibrations, ground reflections can significantly contribute to inaccuracies, leading to variations in the measured radiation patterns of the antenna. By minimizing ground reflections during UAV–based antenna measurements, errors in key measured antenna parameters, such as HPBW, realized gain, and the front-to-back ratio, can be effectively mitigated. To understand the source of propagation losses in a UAV to ground link, simulations were conducted in CST. These simulations identified scattering effects caused by surrounding buildings. Additionally, by simulating a UAV with a horn antenna, potential sources of electromagnetic coupling between the antenna and the UAV body were detected. The survey concludes by identifying key areas for future research and emphasizing the potential of UAVs to revolutionize antenna and propagation measurement practices to avoid the inaccuracies of the antenna parameters measured by the UAV. Full article
(This article belongs to the Special Issue New Methods and Applications for UAVs)
Show Figures

Figure 1

12 pages, 5713 KiB  
Article
Temperature and Frequency Dependence of Human Cerebrospinal Fluid Dielectric Parameters
by Weice Wang, Mingxu Zhu, Benyuan Liu, Weichen Li, Yu Wang, Junyao Li, Qingdong Guo, Fang Du, Canhua Xu and Xuetao Shi
Sensors 2024, 24(22), 7394; https://doi.org/10.3390/s24227394 - 20 Nov 2024
Viewed by 439
Abstract
Accurate human cerebrospinal fluid (CSF) dielectric parameters are critical for biological electromagnetic applications such as the electromagnetic field modelling of the human brain, the localization and intensity assessment of electrical generators in the brain, and electromagnetic protection. To detect brain damage signals during [...] Read more.
Accurate human cerebrospinal fluid (CSF) dielectric parameters are critical for biological electromagnetic applications such as the electromagnetic field modelling of the human brain, the localization and intensity assessment of electrical generators in the brain, and electromagnetic protection. To detect brain damage signals during temperature changes by electrical impedance tomography (EIT), the change in CSF dielectric parameters with frequency (10 Hz–100 MHz) and temperature (17–39 °C) was investigated. A Debye model was first established to capture the complex impedance frequency and temperature characteristics. Furthermore, the receiver operating characteristic (ROC) analysis based on the dielectric parameters of normal and diseased CSF was carried out to identify lesions. The Debye model’s characteristic fc parameters linearly increased with increasing temperature (R2 = 0.989), and R0 and R1 linearly decreased (R2 = 0.990). The final established formula can calculate the complex impedivity of CSF with a maximum fitting error of 3.79%. Furthermore, the ROC based on the real part of impedivity at 10 Hz and 17 °C yielded an area under the curve (AUC) of 0.898 with a specificity of 0.889 and a sensitivity of 0.944. These findings are expected to facilitate the application of electromagnetic technology, such as disease diagnosis, specific absorption rate calculation, and biosensor design. Full article
(This article belongs to the Special Issue Electrical Impedance Spectroscopy Technology)
Show Figures

Figure 1

14 pages, 5176 KiB  
Article
Study on Fatigue Allowance Formulation Based on Physiological Measurements
by Li Qu, Juntong Zhang, Di Wang, Lin Zhang and Zhunan Wu
Sensors 2024, 24(22), 7393; https://doi.org/10.3390/s24227393 - 20 Nov 2024
Viewed by 306
Abstract
The fatigue allowance effectively mitigates declines in physiological state due to work fatigue. An appropriate allowance rate facilitates timely recovery for employees and serves as a crucial basis for labor quota formulation. In this paper, the action mode in mechanical processing was extracted [...] Read more.
The fatigue allowance effectively mitigates declines in physiological state due to work fatigue. An appropriate allowance rate facilitates timely recovery for employees and serves as a crucial basis for labor quota formulation. In this paper, the action mode in mechanical processing was extracted and disassembled into six action units. The study conducted fatigue measurement experiments based on physiological measurement methods, including exercise fatigue tests at different frequencies and work fatigue tests over varying durations. As the frequency of actions increased, the rate monotonic scheduling index showed a linear increasing trend and the degree of fatigue caused by the action was different. The fatigue coefficient of different action units and the fatigue index of the fatigue instability period were obtained by fitting. Hazard ratio indicators showed significant differences, and the corresponding fatigue recovery rest time was obtained for different continuous operation hours. By further fitting the above data, a fatigue relaxation rate model suitable for simulating operation methods was obtained (the fatigue coefficient for the simulated operations in this study is 0.076152) which could provide a reasonable basis for the formulation of fatigue allowance rates for machining methods. Full article
(This article belongs to the Section Biomedical Sensors)
Show Figures

Figure 1

14 pages, 4501 KiB  
Article
Moisture Distribution and Ice Front Identification in Freezing Soil Using an Optimized Circular Capacitance Sensor
by Xing Hu, Qiao Dong, Bin Shi, Kang Yao, Xueqin Chen and Xin Yuan
Sensors 2024, 24(22), 7392; https://doi.org/10.3390/s24227392 - 20 Nov 2024
Viewed by 357
Abstract
As the interface between frozen and unfrozen soil, the ice front is not only a spatial location concept, but also a potentially dangerous interface where the mechanical properties of soil could change abruptly. Accurately identifying its spatial position is essential for the safe [...] Read more.
As the interface between frozen and unfrozen soil, the ice front is not only a spatial location concept, but also a potentially dangerous interface where the mechanical properties of soil could change abruptly. Accurately identifying its spatial position is essential for the safe and efficient execution of large-scale frozen soil engineering projects. Electrical capacitance tomography (ECT) is a promising method for the visualization of frozen soil due to its non-invasive nature, low cast, and rapid response. This paper presents the design and optimization of a mobile circular capacitance sensor (MCCS). The MCCS was used to measure frozen soil samples along the depth direction to obtain moisture distribution and three-dimensional images of the ice front. Finally, the experimental results were compared with the simulation results from COMSOL Multiphysics to analyze the deviations. It was found that the fuzzy optimization design based on multi-criteria orthogonal experiments makes the MCCS meet various performance requirements. The average permittivity distribution was proposed to reflect moisture distribution along the depth direction and showed good correlation. Three-dimensional reconstructed images could provide the precise position of the ice front. The simulation results indicate that the MCCS has a low deviation margin in identifying the position of the ice front. Full article
Show Figures

Figure 1

3 pages, 179 KiB  
Editorial
Special Issue: Artificial Intelligence and Smart Sensor-Based Industrial Advanced Technology
by Luyu Jia, Bairong Sun, Weilin Tan, Shurong Zhang, Bin Zhang and Jianxiong Zhu
Sensors 2024, 24(22), 7391; https://doi.org/10.3390/s24227391 - 20 Nov 2024
Viewed by 565
Abstract
With the rapid growth of smart sensors and industrial data, artificial intelligence (AI) technology (such as machine learning, machine vision, multi-sensor fusion, cloud computing, edge computing, digital twins, etc [...] Full article
17 pages, 812 KiB  
Article
Enhancing Direction-of-Arrival Estimation with Multi-Task Learning
by Simone Bianco, Luigi Celona, Paolo Crotti, Paolo Napoletano, Giovanni Petraglia and Pietro Vinetti
Sensors 2024, 24(22), 7390; https://doi.org/10.3390/s24227390 - 20 Nov 2024
Viewed by 515
Abstract
There are numerous methods in the literature for Direction-of-Arrival (DOA) estimation, including both classical and machine learning-based approaches that jointly estimate the Number of Sources (NOS) and DOA. However, most of these methods do not fully leverage the potential synergies between these two [...] Read more.
There are numerous methods in the literature for Direction-of-Arrival (DOA) estimation, including both classical and machine learning-based approaches that jointly estimate the Number of Sources (NOS) and DOA. However, most of these methods do not fully leverage the potential synergies between these two tasks, which could yield valuable shared information. To address this limitation, in this article, we present a multi-task Convolutional Neural Network (CNN) capable of simultaneously estimating both the NOS and the DOA of the signal. Through experiments on simulated data, we demonstrate that our proposed model surpasses the performance of state-of-the-art methods, especially in challenging environments characterized by high noise levels and dynamic conditions. Full article
Show Figures

Figure 1

18 pages, 4163 KiB  
Article
Privacy-Preserving Synthetic Data Generation Method for IoT-Sensor Network IDS Using CTGAN
by Saleh Alabdulwahab, Young-Tak Kim and Yunsik Son
Sensors 2024, 24(22), 7389; https://doi.org/10.3390/s24227389 - 20 Nov 2024
Viewed by 543
Abstract
The increased usage of IoT networks brings about new privacy risks, especially when intrusion detection systems (IDSs) rely on large datasets for machine learning (ML) tasks and depend on third parties for storing and training the ML-based IDS. This study proposes a privacy-preserving [...] Read more.
The increased usage of IoT networks brings about new privacy risks, especially when intrusion detection systems (IDSs) rely on large datasets for machine learning (ML) tasks and depend on third parties for storing and training the ML-based IDS. This study proposes a privacy-preserving synthetic data generation method using a conditional tabular generative adversarial network (CTGAN) aimed at maintaining the utility of IoT sensor network data for IDS while safeguarding privacy. We integrate differential privacy (DP) with CTGAN by employing controlled noise injection to mitigate privacy risks. The technique involves dynamic distribution adjustment and quantile matching to balance the utility–privacy tradeoff. The results indicate a significant improvement in data utility compared to the standard DP method, achieving a KS test score of 0.80 while minimizing privacy risks such as singling out, linkability, and inference attacks. This approach ensures that synthetic datasets can support intrusion detection without exposing sensitive information. Full article
(This article belongs to the Special Issue Recent Trends and Advances in Sensors Cybersecurity)
Show Figures

Figure 1

11 pages, 6594 KiB  
Article
Simultaneous Structural Monitoring over Optical Ground Wire and Optical Phase Conductor via Chirped-Pulse Phase-Sensitive Optical Time-Domain Reflectometry
by Jorge Canudo, Pascual Sevillano, Andrea Iranzo, Sacha Kwik, Javier Preciado-Garbayo and Jesús Subías
Sensors 2024, 24(22), 7388; https://doi.org/10.3390/s24227388 - 20 Nov 2024
Viewed by 461
Abstract
Optimizing the use of existing high-voltage transmission lines demands real-time condition monitoring to ensure structural integrity and continuous service. Operating these lines at the current capacity is limited by safety margins based on worst-case weather scenarios, as exceeding these margins risks bringing conductors [...] Read more.
Optimizing the use of existing high-voltage transmission lines demands real-time condition monitoring to ensure structural integrity and continuous service. Operating these lines at the current capacity is limited by safety margins based on worst-case weather scenarios, as exceeding these margins risks bringing conductors dangerously close to the ground. The integration of optical fibers within modern transmission lines enables the use of Distributed Fiber Optic Sensing (DFOS) technology, with Chirped-Pulse Phase-Sensitive Optical Time-Domain Reflectometry (CP-ΦOTDR) proving especially effective for this purpose. CP-ΦOTDR measures wind-induced vibrations along the conductor, allowing for an analysis of frequency-domain vibration modes that correlate with conductor length and sag across spans. This monitoring system, capable of covering distances up to 40 km from a single endpoint, enables dynamic capacity adjustments for optimized line efficiency. Beyond sag monitoring, CP-ΦOTDR provides robust detection of external threats, including environmental interference and mechanical intrusions, which could compromise cable stability. By simultaneously monitoring the Optical Phase Conductor (OPPC) and Optical Ground Wire (OPGW), this study offers the first comprehensive, real-time evaluation of both structural integrity and potential external aggressions on overhead transmission lines. The findings demonstrate that high-frequency data offer valuable insights for classifying mechanical intrusions and environmental interferences based on spectral content, while low-frequency data reveal the diurnal temperature-induced sag evolution, with distinct amplitude responses for each cable. These results affirm CP-ΦOTDR’s unique capacity to enhance line safety, operational efficiency, and proactive maintenance by delivering precise, real-time assessments of both structural integrity and external threats. Full article
(This article belongs to the Section Optical Sensors)
Show Figures

Figure 1

18 pages, 9613 KiB  
Article
Toward Versatile Small Object Detection with Temporal-YOLOv8
by Martin C. van Leeuwen, Ella P. Fokkinga, Wyke Huizinga, Jan Baan and Friso G. Heslinga
Sensors 2024, 24(22), 7387; https://doi.org/10.3390/s24227387 - 20 Nov 2024
Viewed by 652
Abstract
Deep learning has become the preferred method for automated object detection, but the accurate detection of small objects remains a challenge due to the lack of distinctive appearance features. Most deep learning-based detectors do not exploit the temporal information that is available in [...] Read more.
Deep learning has become the preferred method for automated object detection, but the accurate detection of small objects remains a challenge due to the lack of distinctive appearance features. Most deep learning-based detectors do not exploit the temporal information that is available in video, even though this context is often essential when the signal-to-noise ratio is low. In addition, model development choices, such as the loss function, are typically designed around medium-sized objects. Moreover, most datasets that are acquired for the development of small object detectors are task-specific and lack diversity, and the smallest objects are often not well annotated. In this study, we address the aforementioned challenges and create a deep learning-based pipeline for versatile small object detection. With an in-house dataset consisting of civilian and military objects, we achieve a substantial improvement in YOLOv8 (baseline mAP = 0.465) by leveraging the temporal context in video and data augmentations specifically tailored to small objects (mAP = 0.839). We also show the benefit of having a carefully curated dataset in comparison with public datasets and find that a model trained on a diverse dataset outperforms environment-specific models. Our findings indicate that small objects can be detected accurately in a wide range of environments while leveraging the speed of the YOLO architecture. Full article
(This article belongs to the Section Sensing and Imaging)
Show Figures

Figure 1

16 pages, 4359 KiB  
Article
Adaptive Kernel Convolutional Stereo Matching Recurrent Network
by Jiamian Wang, Haijiang Sun and Ping Jia
Sensors 2024, 24(22), 7386; https://doi.org/10.3390/s24227386 - 20 Nov 2024
Viewed by 384
Abstract
For binocular stereo matching techniques, the most advanced method currently is using an iterative structure based on GRUs. Methods in this class have shown high performance on both high-resolution images and standard benchmarks. However, simply replacing cost aggregation with a GRU iterative method [...] Read more.
For binocular stereo matching techniques, the most advanced method currently is using an iterative structure based on GRUs. Methods in this class have shown high performance on both high-resolution images and standard benchmarks. However, simply replacing cost aggregation with a GRU iterative method leads to the original cost volume for disparity calculation lacking non-local geometric and contextual information. Based on this, this paper proposes a new GRU iteration-based adaptive kernel convolution deep recurrent network architecture for stereo matching. This paper proposes a kernel convolution-based adaptive multi-scale pyramid pooling (KAP) module that fully considers the spatial correlation between pixels and adds new matching attention (MAR) to refine the matching cost volume before inputting it into the iterative network for iterative updates, enhancing the pixel-level representation ability of the image and improving the overall generalization ability of the network. At present, the AKC-Stereo network proposed in this paper has a higher improvement than the basic network. On the Sceneflow dataset, the EPE of AKC-Stereo reaches 0.45, which is 0.02 higher than the basic network. On the KITTI 2015 dataset, the AKC-Stereo network outperforms the base network by 5.6% on the D1-all metric. Full article
(This article belongs to the Section Sensor Networks)
Show Figures

Figure 1

16 pages, 10376 KiB  
Article
Machine Vision-Based Real-Time Monitoring of Bridge Incremental Launching Method
by Haibo Xie, Qianyu Liao, Lei Liao and Yanghang Qiu
Sensors 2024, 24(22), 7385; https://doi.org/10.3390/s24227385 - 20 Nov 2024
Viewed by 463
Abstract
With the wide application of the incremental launching method in bridges, the demand for real-time monitoring of launching displacement during bridge incremental launching construction has emerged. In this paper, we propose a machine vision-based real-time monitoring method for the forward displacement and lateral [...] Read more.
With the wide application of the incremental launching method in bridges, the demand for real-time monitoring of launching displacement during bridge incremental launching construction has emerged. In this paper, we propose a machine vision-based real-time monitoring method for the forward displacement and lateral offset of bridge incremental launching in which the linear shape of the bottom surface of the girder is a straight line. The method designs a kind of cross target, and realizes efficient detection, recognition, and tracking of multiple targets during the dynamic process of beam incremental launching by training a YOLOv5 target detection model and a DeepSORT multi-target tracking model. Then, based on the convex packet detection and K-means clustering algorithm, the pixel coordinates of the center point of each target are calculated, and the position change of the beam is monitored according to the change in the center-point coordinates of the targets. The feasibility and effectiveness of the proposed method are verified by comparing the accuracy of the total station and the method through laboratory simulation tests and on-site real-bridge testing. Full article
(This article belongs to the Section Intelligent Sensors)
Show Figures

Figure 1

Previous Issue
Back to TopTop