Sensors, Volume 23, Issue 17 (September-1 2023) – 354 articles

There is a rapid increase in the number of edge devices in IoT solutions, generating vast amounts of data that need to be processed and analyzed efficiently. Traditional cloud-based architectures can face latency, bandwidth, and privacy challenges when dealing with this data flood. There is currently no unified approach to the creation of edge computing solutions. This work addresses this problem by exploring containerization for data processing solutions at the network’s edge. The current approach involves creating a specialized application compatible with the device used. Another approach involves using containerization for deployment and monitoring. The heterogeneity of edge environments would greatly benefit from a universal modular platform. Our proposed edge computing-based framework implements a streaming extract, transform, and load pipeline for data processing and analysis using ZeroMQ as the communication backbone and containerization for scalable deployment. Results demonstrate the effectiveness of the proposed framework, making it suitable for time-sensitive IoT applications. Full article

Effective mobility management is crucial for efficient operation of next-generation cellular systems in the millimeter wave (mmWave) band. Massive multiple-input–multiple-output (MIMO) systems are seen as necessary to overcome the significant path losses in this band, but the highly directional beam makes the channels more susceptible to radio link failures due to blockages. To meet stringent capacity and reliability requirements, multi-connectivity has attracted significant attention. This paper proposes a multiagent distributed Q learning-based mobility management scheme for multi-connectivity in mmWave cellular systems. A hierarchical structure is adopted to address the model complexity and speed up the learning process. The performance is assessed using a realistic measurement data set collected from Wireless Insite in an urban area and compared with independent Q learning and a heuristic scheme in terms of handover probability and spectral efficiency. Full article

Adolescent idiopathic scoliosis (AIS) is a common disease that, in many cases, can be conservatively treated through bracing. High adherence to brace prescription is fundamental to gaining the maximum benefit from this treatment approach. Wearable sensors are available that objectively monitor the brace-wearing time, but their use, combined with other interventions, is poorly investigated. The aims of the current review are as follows: (i) to summarize the real compliance with bracing reported by studies using sensors; (ii) to find out the real brace wearing rate through objective electronic monitoring; (iii) to verify if interventions made to increase adherence to bracing can be effective according to the published literature. We conducted a systematic review of the literature published on Medline, EMBASE, CINAHL, Scopus, CENTRAL, and Web of Science. We identified 466 articles and included examples articles, which had a low to good methodological quality. We found that compliance a greatly varied between 21.8 and 93.9% (weighted average: 58.8%), real brace wearing time varied between 5.7 and 21 h per day (weighted average 13.3), and specific interventions seemed to improve both outcomes, with compliance increasing from 58.5 to 66% and brace wearing increasing from 11.9 to 15.1 h per day. Two comparative studies showed positive effects of stand-alone counseling and information on the sensors’ presence when added to counseling. Sensors proved to be useful tools for objectively and continuously monitoring adherence to therapy in everyday clinical practice. Specific interventions, like the use of sensors, counseling, education, and exercises, could increase compliance. However, further studies using high-quality designs should be conducted in this field. Full article

Acquiring in situ water temperature data is an indispensable and important component for analyzing thermal dynamics in estuarine and coastal areas. However, the long-term and high-density monitoring of water temperature is costly and technically challenging. In this paper, we present the design, calibration, and application of the smart temperature sensor TS-V1, a low-power yet low-cost temperature sensor for monitoring the spatial–temporal variations of surface water temperatures and air temperatures in estuarine and coastal areas. The temperature output of the TS-V1 sensor was calibrated against the Fluke-1551A sensor developed in the United States and the CTD-Diver sensor developed in the Netherlands. The results show that the accuracy of the TS-V1 sensor is 0.08 °C, while sensitivity tests suggest that the TS-V1 sensor (comprising a titanium alloy shell with a thermal conductivity of 7.6 W/(m °C)) is approximately 0.31~0.54 s/°C slower than the CTD-Diver sensor (zirconia shell with thermal conductivity of 3 W/(m °C)) in measuring water temperatures but 6.92~10.12 s/°C faster than the CTD-Diver sensor in measuring air temperatures. In addition, the price of the proposed TS-V1 sensor is only approximately 1 and 0.3 times as much as the established commercial sensors, respectively. The TS-V1 sensor was used to collect surface water temperature and air temperature in the western part of the Pearl River Estuary from July 2022 to September 2022. These data wells captured water and air temperature changes, frequency distributions, and temperature characteristics. Our sensor is, thus, particularly useful for the study of thermal dynamics in estuarine and coastal areas. Full article

Optical sensing offers several advantages owing to its non-invasiveness and high sensitivity. The miniaturization of optical sensors will mitigate spatial and weight constraints, expanding their applications and extending the principal advantages of optical sensing to different fields, such as healthcare, Internet of Things, artificial intelligence, and other aspects of society. In this study, we present the development of a miniature optical sensor for monitoring thrombi in extracorporeal membrane oxygenation (ECMO). The sensor, based on a complementary metal-oxide semiconductor integrated circuit (CMOS-IC), also serves as a photodiode, amplifier, and light-emitting diode (LED)-mounting substrate. It is sized 3.8 × 4.8 × 0.75 mm³ and provides reflectance spectroscopy at three wavelengths. Based on semiconductor and microelectromechanical system (MEMS) processes, the design of the sensor achieves ultra-compact millimeter size, customizability, prototyping, and scalability for mass production, facilitating the development of miniature optical sensors for a variety of applications. Full article

Low-cost Particulate Matter (PM) sensors offer an excellent opportunity to improve our knowledge about this type of pollution. Their size and cost, which support multi-node network deployment, along with their temporal resolution, enable them to report fine spatio-temporal resolution for a given area. These sensors have known issues across performance metrics. Generally, the literature focuses on the PM mass concentration reported by these sensors, but some models of sensors also report Particle Number Concentrations (PNCs) segregated into different PM size ranges. In this study, eight units each of Alphasense OPC-R1, Plantower PMS5003 and Sensirion SPS30 have been exposed, under controlled conditions, to short-lived peaks of PM generated using two different combustion sources of PM, exposing the sensors’ to different particle size distributions to quantify and better understand the low-cost sensors performance across a range of relevant environmental ranges. The PNCs reported by the sensors were analysed to characterise sensor-reported particle size distribution, to determine whether sensor-reported PNCs can follow the transient variations of PM observed by the reference instruments and to determine the relative impact of different variables on the performances of the sensors. This study shows that the Alphasense OPC-R1 reported at least five size ranges independently from each other, that the Sensirion SPS30 reported two size ranges independently from each other and that all the size ranges reported by the Plantower PMS5003 were not independent of each other. It demonstrates that all sensors tested here could track the fine temporal variation of PNCs, that the Alphasense OPC-R1 could closely follow the variations of size distribution between the two sources of PM, and it shows that particle size distribution and composition are more impactful on sensor measurements than relative humidity. Full article

Reliable and accurate full-field deformation measurements at elevated temperatures using digital image correlation (DIC) require stable and high-contrast high-temperature speckle patterns to be prepared on the sample surface. However, conventional high-temperature speckle patterns fabricated by the existing methods possess several limitations, e.g., easily fail to preserve original pattern features due to the harsh environment and heavily dependent on the operator’s experience. In this study, we propose a reliable and reproducible high-temperature speckle fabrication method based on air plasma spraying (APS) and flexible speckle templates. This method involves covering the sample surface with pre-designed speckle templates and then spraying the melted speckle powders onto the specimen surface using an air plasma spray technique to obtain customized speckle patterns. The validity of the proposed method was verified by the speckle fabrication on both planar and curved samples and heating tests with these samples. Experimental results demonstrate that the speckle patterns made by the proposed method adhere well to the sample surface, remain stable during the heating process, and exhibit excellent agreement with the reference values in terms of the thermal expansion coefficients. The proposed method provides a reliable and efficient way to create customized and stable speckle patterns for accurate high-temperature DIC measurements. Full article

Enhancing gasoline detergency is pivotal for enhancing fuel efficiency and mitigating exhaust emissions in gasoline vehicles. This study investigated gasoline vehicle emission characteristics with different gasoline detergency, explored synergistic emission reduction potentials, and developed versatile emission prediction models. The results indicate that improved fuel detergency leads to a reduction of 5.1% in fuel consumption, along with decreases of 3.2% in total CO₂, 55.4% in CO, and 15.4% in HC emissions. However, during low-speed driving, CO₂ and CO emissions reductions are limited, and HC emissions worsen. A synergistic emission reduction was observed, particularly with CO exhibiting a pronounced reduction compared to HC. The developed deep-learning-based vehicle emission model for different gasoline detergency (DPVEM-DGD) enables accurate emission predictions under various fuel detergency conditions. The Pearson correlation coefficients (Pearson’s r) between predicted and measured values of CO₂, CO, and HC emissions before and after adding detergency agents are 0.913 and 0.934, 0.895 and 0.915, and 0.931 and 0.969, respectively. The predictive performance improves due to reduced peak emissions resulting from improved fuel detergency. Elevated gasoline detergency not only reduces exhaust emissions but also facilitates more refined emission management to a certain extent. Full article

Resonance Raman spectroscopy (RRS) has been used as a reference method for measuring skin carotenoid levels (SCL), which indicate vegetable and fruit intake. However, RRS is not an easy-to-use method in SCL measurement due to its complicated implementation. In this study, a commercial spectrophotometer based on reflection spectroscopy (RS), which is relatively simple and inexpensive, was evaluated to confirm usability compared with RRS in measuring SCL. To investigate the agreement between RS and RRS, eighty participants were randomly assigned to a high-carotenoid diet group (21 mg/day of total carotenoids) or a control-carotenoid diet group (14 mg/day of total carotenoids) during a 6-week whole-diet intervention period and a 4-week tracking period. Strong correlations between the RS and RRS methods were observed at baseline (r = 0.944) and the entire period (r = 0.930). The rate of SCL increase was similar during the diet intervention; however, the initiation of the SCL decrease in RS was slower than in RRS during the tracking period. To confirm the agreement of RS and RRS from various perspectives, new visualization tools and indices were additionally applied and confirmed the similar response patterns of the two methods. The results indicate that the proposed RS method could be an alternative to RRS in SCL measurements. Full article

Multihop transmission is one of the important techniques to overcome the transmission coverage of each node in wireless sensor networks (WSNs). However, multihop transmission has a security issue due to the nature of a wireless medium. Additionally, the eavesdropper also attempts to interrupt the legitimate users’ transmission. Thus, in this paper, we study the secrecy performance of a multihop transmission under various eavesdropping attacks for WSNs. To improve the secrecy performance, we propose two node selection schemes in each cluster, namely, minimum node selection (MNS) and optimal node selection (ONS) schemes. To exploit the impact of the network parameters on the secrecy performance, we derive the closed-form expression of the secrecy outage probability (SOP) under different eavesdropping attacks. From the numerical results, the ONS scheme shows the most robust secrecy performance compared with the other schemes. However, the ONS scheme requires a lot of channel information to select the node in each cluster and transmit information. On the other side, the MNS scheme can reduce the amount of channel information compared with the ONS scheme, while the MNS scheme still provides secure transmission. In addition, the impact of the network parameters on the secrecy performance is also insightfully discussed in this paper. Moreover, we evaluate the trade-off of the proposed schemes between secrecy performance and computational complexity. Full article

One of the most interesting characteristics of collaborative robots is their ability to be used in close cooperation scenarios. In industry, this facilitates the implementation of human-in-loop workflows. However, this feature can also be exploited in different fields, such as healthcare. In this paper, a rehabilitation framework for the upper limbs of neurological patients is presented, consisting of a collaborative robot that helps users perform three-dimensional trajectories. Such a practice is aimed at improving the coordination of patients by guiding their motions in a preferred direction. We present the mechatronic setup, along with a preliminary experimental set of results from 19 volunteers (patients and control subjects) who provided positive feedback on the training experience (52% of the subjects would return and 44% enjoyed performing the exercise). Patients were able to execute the exercise, with a maximum deviation from the trajectory of 16 mm. The muscular effort required was limited, with average maximum forces recorded at around 50 N. Full article

Gravity sensing is a valuable technique used for several applications, including fundamental physics, civil engineering, metrology, geology, and resource exploration. While classical gravimeters have proven useful, they face limitations, such as mechanical wear on the test masses, resulting in drift, and limited measurement speeds, hindering their use for long-term monitoring, as well as the need to average out microseismic vibrations, limiting their speed of data acquisition. Emerging sensors based on atom interferometry for gravity measurements could offer promising solutions to these limitations, and are currently advancing towards portable devices for real-world applications. This article provides a brief state-of-the-art review of portable atom interferometry-based quantum sensors and provides a perspective on routes towards improved sensors. Full article

The market for unmanned aerial systems (UASs) has grown considerably worldwide, but their ability to transmit sensitive information poses a threat to public safety. To counter these threats, authorities, and anti-drone organizations are ensuring that UASs comply with regulations, focusing on strategies to mitigate the risks associated with malicious drones. This study presents a technique for detecting drone models using identification (ID) tags in radio frequency (RF) signals, enabling the extraction of real-time telemetry data through the decoding of Drone ID packets. The system, implemented with a development board, facilitates efficient drone tracking. The results of a measurement campaign performance evaluation include maximum detection distances of 1.3 km for the Mavic Air, 1.5 km for the Mavic 3, and 3.7 km for the Mavic 2 Pro. The system accurately estimates a drone’s 2D position, altitude, and speed in real time. Thanks to the decoding of telemetry packets, the system demonstrates promising accuracy, with worst-case distances between estimated and actual drone positions of 35 m for the Mavic 2 Pro, 17 m for the Mavic Air, and 15 m for the Mavic 3. In addition, there is a relative error of 14% for altitude measurements and 7% for speed measurements. The reaction times calculated to secure a vulnerable site within a 200 m radius are 1.83 min (Mavic Air), 1.03 min (Mavic 3), and 2.92 min (Mavic 2 Pro). This system is proving effective in addressing emerging concerns about drone-related threats, helping to improve public safety and security. Full article

This paper presents a real-time underwater wireless optical communication (UWOC) system. The transmitter of our UWOC system is equipped with four blue LEDs, and we have implemented pre-emphasis technology to extend the modulation bandwidth of these LEDs. At the receiver end, a 3 mm diameter APD is utilized. Both the transmitter and receiver are housed in watertight chassis and are submerged in a water pool to conduct real-time underwater experiments. Through these experiments, we have obtained impressive results. The data rate achieved by our system reaches up to 135 Mbps, with a BER of 5.9 × 10⁻³, at a distance of 10 m. Additionally, we have developed a convenient method for measuring the underwater attenuation coefficient, using which we have found the attenuation coefficient of the water in experiments to be 0.289 dB/m. Furthermore, we propose a technique to estimate the maximum communication distance of an on–off keying UWOC system with intersymbol interference, based on the Q factor. By applying this method, we conclude that under the same water quality conditions, our system can achieve a maximum communication distance of 25.4 m at 80 Mbps. Overall, our research showcases the successful implementation of a real-time UWOC system, along with novel methods for measuring the underwater attenuation coefficient and estimating the maximum communication distance. Full article

Spaceborne Global Navigation Satellite Systems (GNSS) receivers have become ubiquitous sensors for spacecraft navigation, especially in Low Earth Orbits (LEOs), often also supporting science endeavors or as acting dedicated science payloads. Due to the large number of space-capable GNSS receiver models available, spacecraft designers, as well as scientists, may find it difficult to have or gain an overview of suitable state-of-the-art models for their purposes and constraints. Based on a literature review that included more than 90 different receiver models, this paper aims to provide an overview of space-capable GNSS receivers that have a heritage in space missions. It analyses trends from the collected data and provides an outlook on miniaturized GNSS receiver models, which have a high potential of being used in future space missions. Full article

Accurately forecasting electrical signals from three-phase Direct Torque Control (DTC) induction motors is crucial for achieving optimal motor performance and effective condition monitoring. However, the intricate nature of multiple DTC induction motors and the variability in operational conditions present significant challenges for conventional prediction methodologies. To address these obstacles, we propose an innovative solution that leverages the Fast Fourier Transform (FFT) to preprocess simulation data from electrical motors. A Bidirectional Long Short-Term Memory (Bi-LSTM) network then uses this altered data to forecast processed motor signals. Our proposed approach is thoroughly examined using a comparative examination of cutting-edge forecasting models such as the Recurrent Neural Network (RNN), Long Short-Term Memory (LSTM), and Gated Recurrent Unit (GRU). This rigorous comparison underscores the remarkable efficacy of our approach in elevating the precision and reliability of forecasts for induction motor signals. The results unequivocally establish the superiority of our method across stator and rotor current testing data, as evidenced by Mean Absolute Error (MAE) average results of 92.6864 and 93.8802 for stator and rotor current data, respectively. Additionally, compared to alternative forecasting models, the Root Mean Square Error (RMSE) average results of 105.0636 and 85.7820 underscore reduced prediction loss. Full article

In low-voltage distribution systems, the load types are complex, so traditional detection methods cannot effectively identify series arc faults. To address this problem, this paper proposes an arc fault detection method based on multimodal feature fusion. Firstly, the different mode features of the current signal are extracted by mathematical statistics, Fourier transform, wavelet packet transform, and continuous wavelet transform. The different modal features include one-dimensional features, such as time-domain features, frequency-domain features, and wavelet packet energy features, and two-dimensional features of time-spectrum images. Secondly, the extracted features are preprocessed and prioritized for importance based on different machine learning algorithms to improve the feature data quality. The features of higher importance are input into an arc fault detection model. Finally, an arc fault detection model is constructed based on a one-dimensional convolutional network and a deep residual shrinkage network to achieve high accuracy. The proposed detection method has higher detection accuracy and better performance compared with the arc fault detection method based on single-mode features. Full article

In electrical impedance tomography (EIT) detection of industrial two-phase flows, the Gauss-Newton algorithm is often used for imaging. In complex cases with multiple bubbles, this method has poor imaging accuracy. To address this issue, a new algorithm called the artificial bee colony–optimized radial basis function neural network (ABC-RBFNN) is applied to industrial two-phase flow EIT for the first time. This algorithm aims to enhance the accuracy of image reconstruction in electrical impedance tomography (EIT) technology. The EIDORS-v3.10 software platform is utilized to generate electrode data for a 16-electrode EIT system with varying numbers of bubbles. This generated data is then employed as training data to effectively train the ABC-RBFNN model. The reconstructed electrical impedance image produced from this process is evaluated using the image correlation coefficient (ICC) and root mean square error (RMSE) criteria. Tests conducted on both noisy and noiseless test set data demonstrate that the ABC-RBFNN algorithm achieves a higher ICC value and a lower RMSE value compared to the Gauss–Newton algorithm and the radial basis function neural network (RBFNN) algorithm. These results validate that the ABC-RBFNN algorithm exhibits superior noise immunity. Tests conducted on bubble models of various sizes and quantities, as well as circular bubble models, demonstrate the ABC-RBFNN algorithm’s capability to accurately determine the size and shape of bubbles. This outcome confirms the algorithm’s generalization ability. Moreover, when experimental data collected from a 16-electrode EIT experimental device is employed as test data, the ABC-RBFNN algorithm consistently and accurately identifies the size and position of the target. This achievement establishes a solid foundation for the practical application of the algorithm. Full article

A single-fiber photoacoustic (PA) sensor with a silicon cantilever beam for trace acetylene (C₂H₂) gas analysis was proposed. The miniature gas sensor mainly consisted of a microcantilever and a non-resonant PA cell for the real-time detection of acetylene gas. The gas diffused into the photoacoustic cell through the silicon cantilever beam gap. The volume of the PA cell in the sensor was about 14 μL. By using a 1 × 2 fiber optical coupler, a 1532.8 nm distributed feedback (DFB) laser and a white light interference demodulation module were connected to the single-fiber photoacoustic sensor. A silicon cantilever was utilized to improve the performance when detecting the PA signal. To eliminate the interference of the laser-reflected light, a part of the Fabry–Perot (F-P) interference spectrum was used for phase demodulation to achieve the highly sensitive detection of acetylene gas. The minimum detection limit (MDL) achieved was 0.2 ppm with 100 s averaging time. In addition, the calculated normalized noise equivalent absorption (NNEA) coefficient was 4.4 × 10⁻⁹ W·cm⁻¹·Hz^−1/2. The single-fiber photoacoustic sensor designed has great application prospects in the early warning of transformer faults. Full article

Recently, cardiovascular disease has become the leading cause of death worldwide. Abnormal heart rate signals are an important indicator of cardiovascular disease. At present, the ECG signal acquisition instruments on the market are not portable and manual analysis is applied in data processing, which cannot address the above problems. To solve these problems, this study proposes an ECG acquisition and analysis system based on machine learning. The ECG analysis system responsible for ECG signal classification includes two parts: data preprocessing and machine learning models. Multiple types of models were built for overall classification, and model fusion was conducted. Firstly, traditional models such as logistic regression, support vector machines, and XGBoost were employed, along with feature engineering that primarily included morphological features and wavelet coefficient features. Subsequently, deep learning models, including convolutional neural networks and long short-term memory networks, were introduced and utilized for model fusion classification. The system’s classification accuracy for ECG signals reached 99.13%. Future work will focus on optimizing the model and developing a more portable instrument that can be utilized in the field. Full article

In recent years, parking lot management systems have garnered significant research attention, particularly concerning the application of deep learning techniques. Numerous approaches have emerged for tackling parking lot occupancy challenges using deep learning models. This study contributes to the field by addressing a critical aspect of parking lot management systems: accurate vehicle occupancy determination in specific parking spaces. We propose an advanced solution by harnessing an optimized MobileNetV3 model with custom architectural enhancements, trained on the CNRPark-EXT and PKLOT datasets. The model processes individual parking space patches from real-time video feeds, providing occupancy classification for each patch, identifying occupied or available spaces. Our architectural modifications include the integration of a convolutional block attention mechanism in place of the native attention module and the adoption of blueprint separable convolutions instead of the traditional depth-wise separable convolutions. In terms of performance, our proposed model exhibits superior results when benchmarked against state-of-the-art methods. Achieving an exceptional area under the ROC curve (AUC) value of 0.99 for most experiments with the PKLot dataset, our enhanced MobileNetV3 showcases its exceptional discriminatory power in binary classification. Benchmarked against the CarNet and mAlexNet models, representative of previous state-of-the-art solutions, our proposed model showcases exceptional performance. During evaluations using the combined CNRPark-EXT and PKLot datasets, the proposed model attains an impressive average accuracy of 98.01%, while CarNet achieves 97.03%. Beyond achieving high accuracy and precision comparable to previous models, the proposed model exhibits promise for real-time applications. This work contributes to the advancement of parking lot occupancy detection by offering a robust and efficient solution with implications for urban mobility enhancement and resource optimization. Full article

The advancements in ship detection technology using convolutional neural networks (CNNs) regarding synthetic aperture radar (SAR) images have been significant. Yet, there are still some limitations in the existing detection algorithms. First, the backbones cannot generate high-quality multiscale feature maps. Second, there is a lack of suitable attention mechanisms to suppress false alarms. Third, the current feature intensification algorithms are unable to effectively enhance the shallow feature’s semantic information, which hinders the detection of small ships. Fourth, top-level feature maps have rich semantic information; however, as a result of the reduction of channels, the semantic information is weakened. These four problems lead to poor performance in SAR ship detection and recognition. To address the mentioned issues, we put forward a new approach that has the following characteristics. First, we use Convnext as the backbone to generate high-quality multiscale feature maps. Second, to suppress false alarms, the multi-pooling channel attention (MPCA) is designed to generate a corresponding weight for each channel, suppressing redundant feature maps, and further optimizing the feature maps generated by Convnext. Third, a feature intensification pyramid network (FIPN) is specifically designed to intensify the feature maps, especially the shallow feature maps. Fourth, a top-level feature intensification (TLFI) is also proposed to compensate for semantic information loss within the top-level feature maps by utilizing semantic information from different spaces. The experimental dataset employed is the SAR Ship Detection Dataset (SSDD), and the experimental findings display that our approach exhibits superiority compared to other advanced approaches. The overall Average Precision (AP) reaches up to 95.6% on the SSDD, which improves the accuracy by at least 1.7% compared to the current excellent methods. Full article

The detection and classification of bone marrow (BM) cells is a critical cornerstone for hematology diagnosis. However, the low accuracy caused by few BM-cell data samples, subtle difference between classes, and small target size, pathologists still need to perform thousands of manual identifications daily. To address the above issues, we propose an improved BM-cell-detection algorithm in this paper, called YOLOv7-CTA. Firstly, to enhance the model’s sensitivity to fine-grained features, we design a new module called CoTLAN in the backbone network to enable the model to perform long-term modeling between target feature information. Then, in order to cooperate with the CoTLAN module to pay more attention to the features in the area to be detected, we integrate the coordinate attention (CoordAtt) module between the CoTLAN modules to improve the model’s attention to small target features. Finally, we cluster the target boxes of the BM cell dataset based on K-means++ to generate more suitable anchor boxes, which accelerates the convergence of the improved model. In addition, in order to solve the imbalance between positive and negative samples in BM-cell pictures, we use the Focal loss function to replace the multi-class cross entropy. Experimental results demonstrate that the best mean average precision (mAP) of the proposed model reaches 88.6%, which is an improvement of 12.9%, 8.3%, and 6.7% compared with that of the Faster R-CNN model, YOLOv5l model, and YOLOv7 model, respectively. This verifies the effectiveness and superiority of the YOLOv7-CTA model in BM-cell-detection tasks. Full article

Currently, strawberry harvesting relies heavily on human labour and subjective assessments of ripeness, resulting in inconsistent post-harvest quality. Therefore, the aim of this work is to automate this process and provide a more accurate and efficient way of assessing ripeness. We explored a unique combination of YOLOv7 object detection and augmented reality technology to detect and visualise the ripeness of strawberries. Our results showed that the proposed YOLOv7 object detection model, which employed transfer learning, fine-tuning and multi-scale training, accurately identified the level of ripeness of each strawberry with an mAP of 0.89 and an F1 score of 0.92. The tiny models have an average detection time of 18 ms per frame at a resolution of 1280 × 720 using a high-performance computer, thereby enabling real-time detection in the field. Our findings distinctly establish the superior performance of YOLOv7 when compared to other cutting-edge methodologies. We also suggest using Microsoft HoloLens 2 to overlay predicted ripeness labels onto each strawberry in the real world, providing a visual representation of the ripeness level. Despite some challenges, this work highlights the potential of augmented reality to assist farmers in harvesting support, which could have significant implications for current agricultural practices. Full article

Nowadays, unmanned aerial vehicle (UAV) communication systems are commonly considered as one of the key enabling technologies for 6G. The hybrid free space optical (FSO)/radio frequency (RF) system has the advantages of both FSO and RF links to improve communication system performance, and the relay-assisted system adopts multi-hop transmission and cooperative diversity methods to extend communication coverage. Thus, a joint consideration of UAV-assistedUAV assisted relay in hybrid FSO/RF transmission is meaningful. In this paper, we aim to analyze the performance of UAV-assisted multi-hop parallel hybrid FSO/RF communication systems with and without pointing errors (PE) in terms of Bit Error Rate (BER) and outage probability. In our considered system, the FSO sub-link adopts the Exponential Weibull turbulence model and the RF sub-link suffers the Nakagami fading model. With these, new mathematical formulas of both BER and outage probability are derived under the UAV-assisted hybrid FSO/RF with different modulation methods. Through numerical evaluationnumerical simulations, the performances of UAV-assisted hybrid FSO/RF systems are analyzed under different weather conditions, modulation methods, optical receiver aperture, RF fading parameters, pointing errors, and relay structures. The results demonstrate that (1) compared to hybrid FSO/RF direct links, UAV-assisted hybrid FSO/RF systems can further improve system performance; (2) the performance of UAV-assisted hybrid FSO/RF systems varies with different relay structures; (3) large receiver aperture and RF fading parameters can further improve the communication performance of hybrid FSO/RF direct links and UAV-assisted hybrid FSO/RF systems. Full article

Optical camera communication (OCC) is one of the most promising optical wireless technology communication systems. This technology has a number of benefits compared to radio frequency, including unlimited spectrum, no congestion due to high usage, and low operating costs. OCC operates in order to transmit an optical signal from a light-emitting diode (LED) and receive the signal with a camera. However, identifying, detecting, and extracting data in a complex area with very high mobility is the main challenge in operating the OCC. In this paper, we design and implement a real-time OCC system that can communicate in high mobility conditions, based on You Only Look Once version 8 (YOLOv8). We utilized an LED array that can be identified accurately and has an enhanced data transmission rate due to a greater number of source lights. Our system is validated in a highly mobile environment with camera movement speeds of up to 10 m/s at 2 m, achieving a bit error rate of ${10}^{-2}$. In addition, this system achieves high accuracy of the LED detection algorithm with mAP0.5 and mAP0.5:0.95 values of 0.995 and 0.8604, respectively. The proposed method has been tested in real time and achieves processing speeds up to 1.25 ms. Full article

The increasing network speeds of today’s Internet require high-performance, high-throughput network devices. However, the lack of affordable, flexible, and readily available devices poses a challenge for packet classification and filtering. This problem is exacerbated by the increase in volumetric Distributed Denial-of-Service (DDoS) attacks, which require efficient packet processing and filtering. To meet the demands of high-speed networks and configurable network processing devices, this paper investigates a hybrid hardware/software packet filter prototype that combines reconfigurable FPGA technology and high-speed software filtering on commodity hardware. It uses a novel approach that offloads filtering rules to the hardware and employs a Longest Prefix Matching (LPM) algorithm and allowlists/blocklists based on millions of IP prefixes. The hybrid filter demonstrates improvements over software-only filtering, achieving performance gains of nearly 30%, depending on the rulesets, offloading methods, and traffic types. The significance of this research lies in developing a cost-effective alternative to more-expensive or less-effective filters, providing high-speed DDoS packet filtering for IPv4 traffic, as it still dominates over IPv6. Deploying these filters on commodity hardware at the edge of the network can mitigate the impact of DDoS attacks on protected networks, enhancing the security of all devices on the network, including Internet of Things (IoT) devices. Full article

Although measuring worker productivity is crucial, the measurement of the productivity of each worker is challenging due to their dispersion across various construction jobsites. This paper presents a framework for measuring productivity based on an inertial measurement unit (IMU) and activity classification. Two deep learning algorithms and three sensor combinations were utilized to identify and analyze the feasibility of the framework in masonry work. Using the proposed method, worker activity classification could be performed with a maximum accuracy of 96.70% using the convolutional neural network model with multiple sensors, and a minimum accuracy of 72.11% using the long short-term memory (LSTM) model with a single sensor. Productivity could be measured with an accuracy of up to 96.47%. The main contributions of this study are the proposal of a method for classifying detailed activities and an exploration of the effect of the number of IMU sensors used in measuring worker productivity. Full article

Enhancing recovery is a fundamental component of high-performance sports training since it enables practitioners to potentiate physical performance and minimise the risk of injuries. Using a new sports legging embedded with an intelligent system for electrostimulation, localised heating and compression (completely embodied into the textile structures), we aimed to analyse acute recovery following a fatigue protocol. Surface electromyography- and torque-related variables were recorded on eight recreational athletes. A fatigue protocol conducted in an isokinetic dynamometer allowed us to examine isometric torque and consequent post-exercise acute recovery after using the sports legging. Regarding peak torque, no differences were found between post-fatigue and post-recovery assessments in any variable; however, pre-fatigue registered a 16% greater peak torque when compared with post-fatigue for localised heating and compression recovery methods. Our data are supported by recent meta-analyses indicating that individual recovery methods, such as localised heating, electrostimulation and compression, are not effective to recover from a fatiguing exercise. In fact, none of the recovery methods available through the sports legging tested was effective in acutely recovering the torque values produced isometrically. Full article

Most advanced autonomous driving systems (ADS) today rely on the prior creation of high-definition maps (HD maps). This process is expensive and needs to be performed frequently to keep up with the changing conditions of the road environment. Creating accurate navigation maps online is an alternative to reduce the cost and broaden the current operational design domains (ODD) of modern ADS. This paper offers a snapshot of the state of the art in drivable area estimation, which is an essential technology to deploy ADS in ODDs where HD maps are limited or unavailable. The proposed review introduces a novel architecture breakdown that fits learning-based and non-learning-based techniques and allows the analysis of a set of impactful and recent drivable area algorithms. In addition to that, complimentary information for practitioners is provided: (i) an assessment of the influence of modern sensing technologies on the task under study and (ii) a selection of relevant datasets for evaluation and benchmarking purposes. Full article