Publikační činnost Katedry kybernetiky a biomedicínského inženýrství / Publications of Department of Cybernetics and Biomedical Engineering (450)

Permanent URI for this collectionhttp://hdl.handle.net/10084/64788

Kolekce obsahuje bibliografické záznamy publikační činnosti (článků) akademických pracovníků Katedry kybernetiky a biomedicínského inženýrství (450), dříve nazvané Katedra měřicí a řídicí techniky, v časopisech a v Lecture Notes in Computer Science registrovaných ve Web of Science od roku 2003 po současnost.
Do kolekce jsou zařazeny:
a) publikace, u nichž je v originálních dokumentech jako působiště autora (adresa) uvedena Vysoká škola báňská-Technická univerzita Ostrava (VŠB-TUO),
b) publikace, u nichž v originálních dokumentech není v adrese VŠB-TUO uvedena, ale autoři prokazatelně v době jejich zpracování a uveřejnění působili na VŠB-TUO.

Bibliografické záznamy byly původně vytvořeny v kolekci Publikační činnost akademických pracovníků VŠB-TUO, která sleduje publikování akademických pracovníků od roku 1990.

Browse

Now showing 1 - 20 out of 416 results

Efficient optimization-based trajectory planning for truck-trailer systems
(MDPI, 2024) Ožana,Štěpán; Krupa, Filip; Slanina, Zdeněk
This paper tackles the complex problem of trajectory planning for trucks with multiple trailers, with a specific focus on autonomous parking assistance applications. These systems aim to autonomously guide vehicles from a starting position to a target location while effectively navigating real-world obstacles. We propose a novel six-phase approach that combines global and local optimization techniques, enabling the efficient and accurate generation of reference trajectories. Our method is validated in a case study involving a truck with two trailers, illustrating its capability to handle intricate parking scenarios requiring precise obstacle avoidance and high maneuverability. Results demonstrate that the proposed strategy significantly improves trajectory planning efficiency and robustness in challenging environments.
Distributed sensing with nanoparticle-doped fibers in standard OTDR systems: Validation and performance analysis
(Elsevier, 2026) Leal-Junior, Arnaldo; Kepák, Stanislav; Pedruzzi, Eduarda; Nedoma, Jan; Martinek, Radek; Blanc, Wilfried
This paper presents the implementation of enhanced backscattering optical fibers in optical time-domain reflec tometry (OTDR). The increase in the backscattering is achieved by doping the fiber core with nanoparticles, resulting in the so-called nanoparticle-doped optical fibers (NPFs). The so-called NPF-OTDR sensor system takes advantage of the higher scattering of the NPFs to increase the spatial resolution and strain sensitivity of the distributed sensor system. The wavelength-dependent optical attenuation is measured over a wide range of wave lengths using the cutback method to evaluate the operational wavelength for the OTDR. In this case, the results indicated smaller attenuation in 1310 nm than at the other operational wavelengths of the OTDR system (i.e., 1550 nm and 1625 nm). For this reason, the OTDR is applied at the 1310 nm wavelength for different strains applied along the optical fiber (with 1 m separation between them). The results for strain estimation showed relative errors of around 3%, whereas the strain position estimation is around 0.24 m. These results indicate the feasibility of the proposed system with a potential spatial resolution of around 0.5 m (which can be even lower depending on the OTDR setup). Thus, the advantages of the OTDR using the NPF result in a better sensor system performance when compared with the standard optical fibers, significantly enhancing the conventional OTDR capabilities in terms of resolution with even comparable performance to specialty OTDR systems at a lower cost and with a simpler design.
Unveiling faking in job interviews by examining facial thermal cues in deception detection
(Springer Nature, 2025) Hypšová, Petra; Seitl, Martin; Kubíček, Jan; Mimra, Tomáš
Detecting deceptive information in job interviews is a major challenge for improving personnel selection validity. Traditional interviewer-based methods for detecting applicant faking are limited in accuracy, highlighting the need for reliable techniques to enhance detection and minimize suboptimal hiring decisions. This study aimed to (a) investigate functional infrared thermal imaging (fITI), a non-invasive method for measuring facial temperature, to determine whether cognitive load and arousal during direct deception induce facial thermal changes indicative of faking; and to (b) develop an automated framework for processing infrared recordings, accounting for head movements and physiognomic differences. A within-person experimental design was conducted with 27 participants in simulated job interviews under five conditions: baseline, rehearsed truth, rehearsed lie, spontaneous truth, and spontaneous lie. A high-sensitivity infrared camera recorded thermal changes on the nose, forehead, and cheeks. Our dynamic feature extraction approach enabled robust temporal analysis despite naturalistic head movements. Linear mixed-effects models revealed no significant overall effect of faking; however, rehearsal and interaction effects were significant. Nasal temperature was higher in rehearsed than spontaneous responses and in spontaneous lies than spontaneous truths, while forehead and cheeks temperatures remained stable. Findings suggest that cognitive load from rehearsed and spontaneous questioning influences nasal thermal changes during faking.
Enhancing transparency and efficiency in blockchain harvest: Empowering farmers and consumers through transparent trading in agricultural applications
(Elsevier, 2025) Lakhan, Abdullah; Mohammed, Mazin Abed; Al-Budair, Lilian Qasim Alwan; Memon, Sajida; Slaný, Vlastimil; Deveci, Muhammet; Martinek, Radek
With the development of edge-cutting technologies, digital agriculture farming, product selling, and purchasing have been increasing progressively. On the other hand, transparency in digital agricultural agrochemicals (pesticides or herbicides) substances of products needs to be monitored carefully between production and selling to customers in a transparent form. Recently, blockchain has emerged as a decentralized technology, which is the most potent decentralized technology. It connects many nodes and validates their data transparency during application sharing. Therefore, it is a motivation to use blockchain technology for digital agriculture applications to meet data transparency, security, and privacy requirements. In this paper, we present enhancing transparency and efficiency in blockchain harvest: empowering farmers and consumers through transparent trading in agricultural application tasks. The application tasks are a mixture of agricultural things (IoT) sensors, products, monitoring, fertilizers, transport tracking, farmers, consumers, and institutional data for processing in digital healthcare applications. Our objective is to process the agricultural tasks during production and trading in an immutable, transparent, secure, and private form. To meet constraints such as processing time, blockchain validation, access control, and cyber-attacks, we suggest enhancing transparency and efficiency in blockchain harvesting and empowering farmers and consumers to trade securely and transparently. In proposed blockchain agriculture, all nodes are heterogeneous and connected; therefore, to avoid any time failure, cyberattacks, or block failure, we establish them in a stable form. This paper presents the agriculture blockchain harvest multi-tasking scheduling (ABHMTS) algorithm methodology, which amalgamates different methods: smart contracts, local offloading, multi-tasking, and blockchain validation with malware detection in blocks during data transfer among other nodes. The experimental results in simulation show that ABHMTS improved the transparency of agricultural task data by 98% and executed them with the minimum processing time by 20%, and reduced the risk of resource failure and resource consumption by 10% with higher malware detection in the framework as compared to baseline approaches.
Applied graph neural networks: Domain-driven insights from medicine to remote sensing
(Elsevier, 2026) Duc, Minh Ly; Kiet, Vo Thanh; Bilík, Petr; Martinek, Radek
Graph Neural Networks have emerged as a leading paradigm for processing data with irregular, graph-based structures. Their capacity to capture intricate interdependencies makes them particularly well-suited for complex domains such as medical image analysis, remote sensing, and privacy-centric computing. This paper presents a comprehensive review of 52 scholarly articles published between 2017 and 2025, focusing on the practical implementation of Graph Neural Networks across five key fields. Among the architectures examined, Graph Convolutional Networks dominate (60 %), followed by Graph Attention Networks (35 %), with the remainder comprising hybrid and domain-tailored models (22 %). These percentages are calculated based on the frequency of model usage across the reviewed 40 studies. Since several studies employed more than one Graph Neural Network architecture, the percentages may overlap and sum to more than 100 %. Although a total of 52 articles were reviewed during the mapping process, only 40 studies met all eligibility and quality criteria and were therefore included in the quantitative analysis. All percentage values reported in the abstract are calculated based solely on these 40 finalized studies. The remaining 12 articles were part of the broader literature survey but were excluded from statistical computation because they did not satisfy the inclusion criteria. To enable robust cross-domain assessment, we introduce an original evaluation framework composed of nine practical dimensions, including metrics like predictive accuracy, model interpretability, computational efficiency, resilience to noise, and support for real-time operation. The analysis highlights the superiority of Graph Convolutional Networks in hyperspectral imagery tasks, while Graph Attention Networks show growing success in detailed medical diagnostics due to their attention mechanisms. Unlike earlier reviews that focus on theoretical progress, this study emphasizes the effectiveness of real-world models under deployment conditions, with a focus on reproducibility, domain constraints, and scalability. We conclude by outlining future research priorities, such as the design of resource-efficient Graph Neural Networks for embedded systems and the creation of unified benchmarks to evaluate graph learning across multiple domains.
The role of photonic sensing technologies in Healthcare 5.0: A comprehensive review of future perspectives and applications
(IEEE, 2025) Leal-Junior, Arnaldo; Nedoma, Jan; Martinek, Radek
This article presents a review of Healthcare 5.0 and its historical perspective, considering the industrial revolutions. The main aspects and challenges of the next generation of healthcare systems are discussed. Due to the inherent advantages of optical fiber sensor (OFS) systems, such technology is comprehensively reviewed and discussed in terms of healthcare applications. Thereafter, an analytical correlation is performed between OFS technologies and the requirements of Healthcare 5.0, and discussed in detail. The key parameters in Healthcare 5.0 indicated that OFS technology aligns well with the requirements and challenges of Healthcare 5.0. Thus, photonic technologies play a key role in the future of healthcare, especially in innovative approaches for integrated sensing and communication (ISAC), as such approaches are readily available for use with optical fibers that combine excellent signal transmission and distributed sensing capabilities. Thus, the contributions of this work are related to the first review of Healthcare 5.0, focused on the role of photonic technologies in such new healthcare evolution, as well as the analysis concerning the key aspects of such applications.
A review of patient bed sensors for monitoring of vital signs
(MDPI, 2024) Recmanik, Michaela; Martinek, Radek; Nedoma, Jan; Jaroš, René; Pelc, Mariusz; Hájovský, Radovan; Velička, Jan; Pieš, Martin; Ševčáková, Marta; Kawala-Sterniuk, Aleksandra
The analysis of biomedical signals is a very challenging task. This review paper is focused on the presentation of various methods where biomedical data, in particular vital signs, could be monitored using sensors mounted to beds. The presented methods to monitor vital signs include those combined with optical fibers, camera systems, pressure sensors, or other sensors, which may provide more efficient patient bed monitoring results. This work also covers the aspects of interference occurrence in the above-mentioned signals and sleep quality monitoring, which play a very important role in the analysis of biomedical signals and the choice of appropriate signal-processing methods. The provided information will help various researchers to understand the importance of vital sign monitoring and will be a thorough and up-to-date summary of these methods. It will also be a foundation for further enhancement of these methods.
A double resistive-capacitive approach for the analysis of a hybrid battery-ultracapacitor integration study
(MDPI, 2025) Chmielewski, Adrian; Piórkowski, Piotr; Bogdziński, Krzysztof; Krawczyk, Paweł; Lorencki, Jakub; Kopczyński, Artur; Możaryn, Jakub; Costa-Castelló, Ramon; Ožana, Štěpán
The development of energy storage systems is significant for solving problems related to climate change. A hybrid energy storage system (HESS), combining batteries with ultracapacitors, may be a feasible way to improve the efficiency of electric vehicles and renewable energy applications. However, most existing research requires comprehensive modelling of HESS components under different operating conditions, hindering optimisation and real-world application. This study proposes a novel approach to analysing the set of differential equations of a substitute model of HESS and validates a model-based approach to investigate the performance of an HESS composed of a Valve-Regulated Lead Acid (VRLA) Absorbent Glass Mat (AGM) battery and a Maxwell ultracapacitor in a parallel configuration. Consequently, the set of differential equations describing the HESS dynamics is provided. The dynamics of this system are modelled with a double resistive-capacitive (2-RC) scheme using data from Hybrid Pulse Power Characterisation (HPPC) and pseudo-random cycles. Parameters are identified using the Levenberg-Marquardt algorithm. The model's accuracy is analysed, estimated and verified using Mean Square Errors (MSEs) and Normalised Root Mean Square Errors (NRMSEs) in the range of a State of Charge (SoC) from 0.1 to 0.9. Limitations of the proposed models are also discussed. Finally, the main advantages of HESSs are highlighted in terms of energy and open-circuit voltage (OCV) characteristics.
A double resistive-capacitive approach for the analysis of a hybrid battery-ultracapacitor integration study
(MDPI, 2025) Chmielewski, Adrian; Piórkowski, Piotr; Bogdziński, Krzysztof; Krawczyk, Paweł; Lorencki, Jakub; Kopczyński, Artur; Możaryn, Jakub; Costa-Castelló, Ramon ; Ožana, Štěpán
The development of energy storage systems is significant for solving problems related to climate change. A hybrid energy storage system (HESS), combining batteries with ultracapacitors, may be a feasible way to improve the efficiency of electric vehicles and renewable energy applications. However, most existing research requires comprehensive modelling of HESS components under different operating conditions, hindering optimisation and real-world application. This study proposes a novel approach to analysing the set of differential equations of a substitute model of HESS and validates a model-based approach to investigate the performance of an HESS composed of a Valve-Regulated Lead Acid (VRLA) Absorbent Glass Mat (AGM) battery and a Maxwell ultracapacitor in a parallel configuration. Consequently, the set of differential equations describing the HESS dynamics is provided. The dynamics of this system are modelled with a double resistive–capacitive (2-RC) scheme using data from Hybrid Pulse Power Characterisation (HPPC) and pseudo-random cycles. Parameters are identified using the Levenberg–Marquardt algorithm. The model’s accuracy is analysed, estimated and verified using Mean Square Errors (MSEs) and Normalised Root Mean Square Errors (NRMSEs) in the range of a State of Charge (SoC) from 0.1 to 0.9. Limitations of the proposed models are also discussed. Finally, the main advantages of HESSs are highlighted in terms of energy and open-circuit voltage (OCV) characteristics.
Advancements in noninvasive blood pressure measurement: a comprehensive review of cuffed, cuffless, and contactless methods
(IEEE, 2026) Kauzlaričová, Terezie; Augustynek, Martin; Kubíček, Jan
Recent advances in noninvasive blood pressure (BP) monitoring are transforming the field, offering alternatives to traditional cuff-based methods that are often limited by discomfort and intermittent measurement. Following PRISMA guidelines, this review systematically analyses studies published between 2020 and 2024, categorizing current BP measurement techniques into cuff-based, cuffless, and contactless methods. Special attention is given to emerging technologies such as photoplethysmography (PPG), electrocardiography (ECG), bioimpedance (BIOZ), radar sensing, and remote PPG (rPPG), as well as the integration of artificial intelligence (AI) for signal processing and BP estimation. This review evaluates the comparative performance, clinical relevance, and readiness for real-world application across modalities. While cuffless and contactless approaches show significant promise for continuous and user-friendly monitoring, their accuracy often depends on calibration, multimodal data fusion, and robust algorithm design. Ethical considerations related to unobtrusive monitoring are also discussed. This work highlights current limitations, identifies future research opportunities, and supports the development of next-generation, AI-driven BP monitoring systems suitable for both clinical and consumer use.
Automated recognition of worker activity using machine learning and multi-sensor fusion
(Elsevier, 2026) Jaroš, René; Majidzadeh Gorjani, Ojan; Barnová, Kateřina; Lajdolf, Martin; Martinek, Radek; Bilík, Petr; Danys, Lukáš
Activity recognition and worker safety are critical parts of future industrial environments. Consecutive transitions to massive automatization and reductions in workforce could lead to more focused monitoring of the remaining workers. Effective recognition often relies on a number of sensors, such as electromyography leads, accelerome ters, magnetometers, or gyroscopes, with each method potentially impacting system cost. This study focuses on automatic worker activity recognition using multilayer perceptron and multi-sensor fusion. We present a testing methodology and analyze the influence of individual sensors on system accuracy. The experiment therefore aims to systematically quantify the trade-off between the type and number of sensors used and the resulting accuracy, with the goal of providing a quantified basis for designing future systems with an optimal price/performance ratio. The presented results demonstrate that different sensor fusion strategies lead to vastly different accuracies, ranging from 71 % to over 99 %. These early findings highlight the crucial role of optimal sensor selection in the design of cost-effective worker monitoring systems. The study demonstrated that using a combination of the accelerometer, magnetometer, gyroscope, and electromyography resulted in the highest test accuracy of 98.04 %. Among these, the gyroscope and electromyography contributed less significantly. When only the accelerometer, magnetometer, and gyroscope were used, the accuracy dropped by approximately 6 %.
Methods for magnetic signature comparison evaluation in vehicle re-identification context
(MDPI, 2024) Balamutas, Juozas; Navikas, Dangirutis; Markevičius, Vytautas; Čepėnas, Mindaugas; Valinevičius, Algimantas; Žilys, Mindaugas; Prauzek, Michal; Konečný, Jaromír; Frivaldský, Michal; Li, Zhixiong; Andriukaitis, Darius
Intelligent transportation systems represent innovative solutions for traffic congestion minimization, mobility improvements and safety enhancement. These systems require various inputs about vehicles and traffic state. Vehicle re-identification systems based on video cameras are most popular; however, more strict privacy policy necessitates depersonalized vehicle re-identification systems. Promising research for depersonalized vehicle re-identification systems involves leveraging the captured unique distortions induced in the Earth's magnetic field by passing vehicles. Employing anisotropic magneto-resistive sensors embedded in the road surface system captures vehicle magnetic signatures for similarity evaluation. A novel vehicle re-identification algorithm utilizing Euclidean distances and Pearson correlation coefficients is analyzed, and performance is evaluated. Initial processing is applied on registered magnetic signatures, useful features for decision making are extracted, different classification algorithms are applied and prediction accuracy is checked. The results demonstrate the effectiveness of our approach, achieving 97% accuracy in vehicle re-identification for a subset of 300 different vehicles passing the sensor a few times.
Enhancing myocardial infarction detection with vectorcardiography: fusion-based comparative analysis of machine learning methods
(Frontiers Media S.A., 2026) Vondrák, Jaroslav; Penhaker, Marek
Background: Early detection and diagnosis of myocardial infarction (MI) help physicians save lives through timely treatment. Vectorcardiography (VCG) is an alternative to the 12-lead electrocardiography, providing not only characteristic changes in cardiac electrical activity in MI patients but also unique spatial information often overlooked by traditional methods. Despite its potential, comprehensive comparative studies applying machine learning (ML) techniques specifically to VCG data remain limited. Methods: This study proposes a novel VCG processing methodology using a comparative analysis of machine learning-based algorithms for the automated detection of MI patients from VCG recordings, utilizing extracted domain knowledge VCG features that monitor morphological changes in cardiac activity. For this purpose, records from the PTB Diagnostic dataset were used. The extracted domain knowledge dataset of morphological features was then fed into a diverse set of 210 machine learning configurations, including K-nearest neighbor, Support Vector Machine, Discriminant Analysis, Artificial Neural Network, Decision Tree, Random Forest, Naive Bayes, Logistic Regression, and Ensemble Methods. To further improve classification performance, we combined analyzed high-performing models using a stacking ensemble strategy, which integrates multiple base classifiers into a meta-classifier. Results: The stacking-based decision-level fusion achieved high accuracy of 95.55%, sensitivity of 97.70%, specificity of 86.25%, positive predictive value of 96.86%, negative predictive value of 89.61% and f1-score of 97.27%. Conclusion: The results demonstrate that decision-level fusion via stacking improves classification performance and enhances the reliability of MI detection from VCG recordings, supporting cardiologists in decision-making.
Adopting artificial intelligence algorithms for remote fetal heart rate monitoring and classification using wearable fetal phonocardiography
(Elsevier, 2024) Abburi, Radha; Hatai, Indranil; Jaroš, René; Martinek, Radek; Babu, Thirunavukkarasu Arun; Babu, Sharmila Arun; Samanta, Sibendu
Fetal phonocardiography (FPCG) is a non-invasive Fetal Heart Rate (FHR) monitoring technique that can detect vibrations and murmurs in heart sounds. However, acquiring fetal heart sounds from a wearable FPCG device is challenging due to noise and artefacts. This research contributes a resilient solution to overcome the conventional issues by adopting Artificial Intelligence (AI) with FPCG for automated FHR monitoring in an end-to-end manner, named (AI-FHR). Four sequential methodologies were used to ensure reliable and accurate FHR monitoring. The proposed method removes low-frequency noises and high-frequency noises by using Chebyshev II high-pass filters and Enhanced Complete Ensemble Empirical Mode Decomposition with Adaptive Noise (ECEEMDAN) in combination with Phase Shifted Maximal Overlap Discrete Wavelet Transform (PS-MODWT) filters, respectively. The denoised signals are segmented to reduce complexity, and the segmentation is performed using multi-agent deep Q-learning (MA-DQL). The segmented signal is provided to reduce the redundancies in cardiac cycles using the Artificial Hummingbird Optimization (AHBO) algorithm. The segmented and non-redundant signals are converted into 3D spectrograms using a machine learning algorithm called variational auto-encoder-general adversarial networks (VAE-GAN). The feature extraction and classification are carried out by adopting a hybrid of the bidirectional gated recurrent unit (BiGRU) and the multi-boosted capsule network (MBCapsNet). The proposed method was implemented and simulated using MATLAB R2020a and validated by adopting effective validation metrics. The results demonstrate that the proposed method performed better than the current method with accuracy (81.34%), sensitivity (72%), F1- score (83%), Energy (0.808 J), and complexity index (13.34). Like other optimization methods, AHO needs precise parameter adjustment in order to function well. Its performance may be greatly impacted by the selection of parameters, including population size, exploration rate, and learning rate.
Hybrid six sigma based on recursive kalman filter and weibull distribution to estimate the lifespan of Bulb LEDs
(Elsevier, 2024) Duc, Minh Ly; Bilík, Petr; Martinek, Radek
The durability of LED bulbs ensures that the lights can operate for a long time without needing to be replaced or maintained too often. The brightness of high-power LED bulbs provides a strong and efficient light source, saves energy, and has a long lifespan. The traditional method of testing LED bulb life span has limitations such as long testing time, limited accuracy, high cost, inability to predict accurately, and failure to respond quickly. Therefore, the development of new and improved testing methods is necessary to effectively and accurately evaluate and ensure the durability of LED bulbs. This research paper proposes a hybrid Six Sigma method based on the recursive Kalman filter and Weibull distribution to estimate the lifespan of Bulb LEDs. The goal of the new method for evaluating LED bulb life is to provide accurate, reliable, and multi-dimensional information about the life of LED lamps, meeting the requirements of accuracy, diversity, time-saving, resources, credibility, and the ability to evaluate from a variety of perspectives. As a result of the research, the recursive Kalman filter method has strengths such as high accuracy, flexibility, saving computational resources, the ability to handle inaccurate data, and scalability. When applied to LED bulb life assessment, it provides an accurate and reliable estimate of LED lamp life. Research results on longevity testing were reduced from 1650 h to 710 h. The end-of-life test time is 710 h, and the aging time per life test cycle is 135 h. The cost of life testing is reduced from 2100.17 USD to 410.12 USD. This LED bulb lifespan testing method can be applied to similar types of LED products to improve the company's productivity and business efficiency.
DT-LSMAS: Digital Twin-Assisted Large-Scale Multiagent System for Healthcare Workflows
(IEEE, 2024) Lakhan, Abdullah; Mohammed, Mazin Abed; Zebar, Dilovan Asaad; Abdulkareem, Karrar Hameed; Deveci, Muhammet; Marhoon, Haydar Abdulameer; Nedoma, Jan; Martinek, Radek
Digital healthcare has garnered much attention from academia and industry for health and well-being. Many digital healthcare architectures based on large-scale edge and cloud multiagent systems (LSMASs) have recently been presented. The LSMAS allows agents from different institutions to work together to achieve healthcare processing goals for users. This article presents a digital twin large-scale multiagent strategy (DT-LSMAS) comprising mobile, edge, and cloud agents. The DT-LSMAS comprised different schemes for healthcare workflows, such as added healthcare workflows, application partitioning, and scheduling. We consider healthcare workflows with different biosensor data such as heartbeat, blood pressure, glucose monitoring, and other healthcare tasks. We partitioned workflows into mobile, edge, and cloud agents to meet the deadline, total time, and security of workflows in large-scale edge and cloud nodes. To handle the large-scale resource for real-time sensor data, we suggested digital twin-enabled edge nodes, where delay-sensitive workflow tasks are scheduled and executed under their quality of service requirements. Simulation results show that the DT-LSMAS outperformed in terms of total time by 50%, minimizing the risk of resource leakage and deadline missing during scheduling on heterogeneous nodes. In conclusion, the DT-LSMAS obtained optimal results for workflow applications.
Edge-cloud remote sensing data-based plant disease detection using deep neural networks with transfer learning
(IEEE, 2024) Mohammed, Mazin Abed; Lakhan, Abdullah; Abdulkareem, Karrar Hameed; Almujally, Nouf Abdullah; Al-Attar, Bourair Bourair Sadiq Mohammed Taqi; Memon, Sajida; Marhoon, Haydar Abdulameer; Martinek, Radek
These days, the disease among different plants has been increasing day by day. It is a very hard task for government institutions and farmers to collect data on plant diseases from different distributed lands among regions. Therefore, data collection, disease detection, and processing are the key issues for plants when they are suffering from healthy and unhealthy issues in different lands. This article presents edge-cloud remote sensing data-based plant disease detection by exploiting deep neural networks with transfer learning. The objective is to solve the aforementioned issues, such as data collection at a wide range, disease detection, and processing them with higher accuracy and time on different machines. We suggest transfer learning commutative fuzzy deep convolutional neural network (FCDCNN) schemes based on combinatorial optimization problems. The convex function optimizes the processing time and learning rate of data training on different edge and cloud nodes to collect more and more data from different plants from distributed lands. In the concave function, we predict the diseases among different plants, such as sugarcane, blueberry, cotton, and cherry with images, videos, and numeric values. The plant disease detection app uses edge nodes and remote satellite point cloud nodes to gather and train data using transfer learning and make predictions using fuzzy DCNN schemes that are more accurate and take less time to process. Simulation results show that FCDCNN obtained higher accuracy by 98% with less processing time 25% and trained with a higher ratio of data than existing schemes.
Multi-agent reinforcement learning framework based on information fusion biometric ticketing data in different public transport modes
(Elsevier, 2024) Lakhan, Abdullah; Rashid, Ahmed N.; Mohammed, Mazin Abed; Zebari, Dilovan Asaad; Deveci, Muhammet; Wang, Limin, limin; Abdulkareem, Karrar Hameed; Nedoma, Jan; Martinek, Radek
In smart cities, biometric technologies have become extensively used for ticket authentication on public transport. Information fusion plays a key role in biometric ticketing, allowing ticket validation with more data source validation in different public transport modes. This paper proposes a novel biometric technology -based mobile ticket application -based system. We formulate the problem as a multi -agent reinforcement learning framework for biometric ticketing in multi -transport environments. Specifically, we propose the Asynchronous Advantage Critic Biometric Ticketing Framework (A3CBTF) algorithm, which consists of different schemes based on the proposed system. The proposed algorithm framework operates in hybrid transport modes using a parallel reinforcement learning scheme. A key advantage of A3CBTF is that it enables passengers to use a single ticket for various public transport modes. Additionally, even when a passenger's mobile device is stolen, lost, or has a dead battery, they can still validate their tickets through different information fusion sources, such as fingerprint and face recognition. A3CBTF is a multi -agent system that integrates mobile, transport, edge, and cloud servers to facilitate ticket validation in a distributed environment. By optimizing both convex and concave optimizations, A3CBTF ensures efficient ticket validation with minimal processing time and maximizes validation rewards across different biometric technologies. Experimental results demonstrate that A3CBTF outperforms mobile off with other options such as fingerprint and face recognition in public transport as compared to other ticketing systems.
Renewable energy resource management using an integrated robust decision making model under entropy and similarity measures of fuzzy hypersoft set
(Elsevier, 2024) Saeed, Muhammad Haris; Saeed, Muhammad; Rahman, Atiqe Ur; Ahsan, Muhammad; Mohammed, Mazin Abed; Marhoon, Haydar Abdulameer; Nedoma, Jan; Martinek, Radek
The demand for renewable energy has significantly increased over the last decade with increased attention to the preservation of the environment and sustainable, optimal resource management. As traditional sources of energy production are depleting at an alarming rate and causing longlasting environmental damage, it is essential to explore green and cost-effective methodologies for meeting energy demand. With each country having different geographical, political, social, and natural factors, the problem arises of which renewable energy should be utilized for optimal resource management. This multi -criteria decision making (MCDM) challenge is tackled by applying a dynamic fuzzy hypersoft set -based Method for the evaluation of currently deployed Renewable Energy systems and providing a decision support system for the installation of new ones based on the factors mentioned above for Turkey. As the installation of new renewable energy projects and the evaluation of old ones is significantly influenced by human judgment, it leaves great room for uncertainty primarily because of the psychological factors of the expert. The novel concept of Fuzzy Hypersoft Sets (FHSs) and their Entropy (EN) and TOPSIS-based operations are first discussed with reference to the problem at hand. The presented structure is superior to the ones in the literature by allowing access to data parameters as sub -parametric values while utilizing the versatility of Fuzzy structures to deal with uncertainty. The technique has great potential to serve as a potential decision support system in any setting. For now, hypothetical expert ratings are used to illustrate the working of the dynamic structure along with a sensitivity analysis to investigate the primary criterion weights in sorting. The evaluation of currently deployed renewable energy systems using our methodology revealed an average improvement in system performance compared to traditional methods. Furthermore, the decision support system for the installation of new projects based on geographical, political, social, and natural factors exhibited a potential increase in overall system efficiency. These numeric outcomes highlight the effectiveness and practical applicability of our approach in optimizing resource management and fostering sustainable energy practices.
Retinal image dataset of infants and retinopathy of prematurity
(Springer Nature, 2024) Timkovič, Juraj; Nowaková, Jana; Kubíček, Jan; Hasal, Martin; Varyšová, Alice; Kolarčík, Lukáš; Maršolková, Kristýna; Augustynek, Martin; Snášel, Václav
Retinopathy of prematurity (ROP) represents a vasoproliferative disease, especially in newborns and infants, which can potentially affect and damage the vision. Despite recent advances in neonatal care and medical guidelines, ROP still remains one of the leading causes of worldwide childhood blindness. The paper presents a unique dataset of 6,004 retinal images of 188 newborns, most of whom are premature infants. The dataset is accompanied by the anonymized patients' information from the ROP screening acquired at the University Hospital Ostrava, Czech Republic. Three digital retinal imaging camera systems are used in the study: Clarity RetCam 3, Natus RetCam Envision, and Phoenix ICON. The study is enriched by the software tool ReLeSeT which is aimed at automatic retinal lesion segmentation and extraction from retinal images. Consequently, this tool enables computing geometric and intensity features of retinal lesions. Also, we publish a set of pre-processing tools for feature boosting of retinal lesions and retinal blood vessels for building classification and segmentation models in ROP analysis.

Browse

Recent Submissions