Efficient and Lightweight Time Synchronization Technique for IoT Devices Using Randomized Synchronization Tokens DOI
Ismail Kertiou, Mostefa Kara, Abdelkader Laouid

et al.

Published: Dec. 21, 2023

Clock synchronization is an important challenge in the Internet of Things (IoT), especially for wireless sensor networks (WSNs). It refers to aligning timekeeping multiple devices or systems a common reference time. ensures accuracy and reliability collected data, which are often used critical applications, such as environmental monitoring security. However, clock techniques can significantly impact resources, including energy consumption computing power. In this paper, we present novel approach WSNs, low-resource networks. To minimize address constraint limited resources nodes, randomly select subset neighbors, contrast existing methods consider all neighbors. Subsequently, each node selects half values from chosen nodes compute average. Finally, adjusts its computed Experimental evaluation results show that system achieves after 4 iterations requires only one additional iteration when 10% malicious. The proposed technique reduces influence malicious storage space, computation time, consumption, IoT applications.

Language: Английский

State-level multidimensional agricultural drought susceptibility and risk assessment for agriculturally prominent areas DOI
S M Samiul Islam, Serhan Yeşilköy, Özlem Baydaroğlu

et al.

International Journal of River Basin Management, Journal Year: 2024, Volume and Issue: unknown, P. 1 - 18

Published: Feb. 13, 2024

Given the growing climate variability, quantifying droughts has gained significant importance, particularly in agriculturally concentrated areas such as Iowa. This study presents a novel approach for evaluating risk of agricultural drought, which combines geospatial methods with fuzzy logic algorithm. The integrates diverse array meteorological, physical, and social factors, yielding more comprehensive nuanced understanding impacts drought. covered sector within Corn Belt region Iowa formulated maps illustrating vulnerability drought timeframe spanning from 2015 to 2021. illustrate progress analysis, fully representing spatial temporal dimensions uniqueness this is ascribed its methodological framework, thorough assessment prior research inform assignment weights parameters logic-based index. findings demonstrate notable increase proportion Iowa's land area classified at a'very high' risk, rising 0.66% 5.39% 2018. upward trend suggests an escalating susceptibility conditions. Mid-Iowa western portion state exhibited increased 'high' 'extremely threats during period. accuracy our was validated using Kappa coefficient 75%. indicator potential be utilized context mitigation program monitoring. Moreover, methodology can modified implementation geographical across globe.

Language: Английский

Citations

11
Artificial Intelligence Algorithms in Flood Prediction: A General Overview DOI
Manish Pandey

Springer eBooks, Journal Year: 2024, Volume and Issue: unknown, P. 243 - 296

Published: Jan. 1, 2024

Language: Английский

Citations

5
Integrating vision‐based AI and large language models for real‐time water pollution surveillance DOI
R. Dinesh Jackson Samuel, Yusuf Sermet, David M. Cwiertny

et al.

Water Environment Research, Journal Year: 2024, Volume and Issue: 96(8)

Published: Aug. 1, 2024

Water pollution has become a major concern in recent years, affecting over 2 billion people worldwide, according to UNESCO. This can occur by either naturally, such as algal blooms, or man-made when toxic substances are released into water bodies like lakes, rivers, springs, and oceans. To address this issue monitor surface-level local bodies, an informative real-time vision-based surveillance system been developed conjunction with large language models (LLMs). integrated camera connected Raspberry Pi for processing input frames is further linked LLMs generating contextual information regarding the type, causes, impact of pollutants on both human health environment. multi-model setup enables authorities take necessary steps mitigate it. train vision model, seven types found bloom, synthetic foams, dead fishes, oil spills, wooden logs, industrial waste run-offs, trashes were used achieving accurate detection. ChatGPT API model generate about detected. Thus, conduct autonomously alert immediate action, eliminating need intervention. PRACTITIONER POINTS: Combines cameras pollutant information. Uses YOLOv5 detect fish, waste. Supports various modules environments, including drones mobile apps broad monitoring. Educates environmental healthand alerts pollution.

Language: Английский

Citations

3
Refraction-based waterlogging depth measurement using solely traffic cameras for transparent flood monitoring DOI
Jintao Qin, Ping Shen

Journal of Hydrology, Journal Year: 2025, Volume and Issue: unknown, P. 132917 - 132917

Published: Feb. 1, 2025

Language: Английский

Citations

0
HydroSignal: open-source internet of things information communication platform for hydrological education and outreach DOI
Baran Kaynak, Omer Mermer, Yusuf Sermet

et al.

Environmental Monitoring and Assessment, Journal Year: 2025, Volume and Issue: 197(4)

Published: March 6, 2025

Language: Английский

Citations

0
State-Level Multidimensional Agricultural Drought Susceptibility and Risk Assessment for Agriculturally Prominent Areas DOI Creative Commons
S M Samiul Islam, Serhan Yeşilköy, Özlem Baydaroğlu

et al.

EarthArXiv (California Digital Library), Journal Year: 2023, Volume and Issue: unknown

Published: June 16, 2023

Due to the shifting climate, extreme events are being observed more frequently globally. Drought is one of most common natural hazards that severely impacts communities in terms economic losses and agricultural production disruption. Considering global trade, drought an region affects food security other regions because disrupted supply. Decision-makers often consult susceptibility maps when preparing mitigation plans so adverse a event can be reduced. Creating demanding, requiring lot data (i.e., hydrological land use), expertise, thorough assessment accurately picture vulnerable region’s condition. The process also relies on complex hydrometeorological models. objective this investigation examine vulnerability impact formulate susceptibility, exposure, risk by considering multitude atmospheric, physical social indicators. Subsequent notion, fuzzy logic algorithm has been devised assigning comprehensive array weights each parameter derived from exhaustive literature review used for preliminary state Iowa. This located Corn Belt region, its primary activity agriculture. Iowa have generated period spanning 2015 2021 validated using Kappa coefficient. produced support decisions

Language: Английский

Citations

4
Visual blockage assessment at culverts using synthetic images to mitigate blockage-originated floods DOI Creative Commons
Umair Iqbal, Johan Barthélemy, Pascal Perez

et al.

Journal of Hydroinformatics, Journal Year: 2023, Volume and Issue: 25(4), P. 1531 - 1545

Published: July 1, 2023

Abstract The assessment of visual blockages in cross-drainage hydraulic structures, such as culverts and bridges, is crucial for ensuring their efficient functioning preventing flash flooding incidents. extraction blockage-related information through computer vision algorithms can provide valuable insights into the blockage. However, absence comprehensive datasets has posed a significant challenge effectively training models. In this study, we explore use synthetic data, images culvert (SIC) hydraulics lab dataset (VHD), combination with limited real-world dataset, openings blockage (ICOB), to evaluate performance opening detector. Faster Region-based Convolutional Neural Network (Faster R-CNN) model ResNet50 backbone was used impact data evaluated two experiments. first involved different combinations while second reduced images. results experiment revealed that structured training, where were initial ICOB fine-tuning, resulted slightly improved detection performance. showed conjunction number images, significantly degradation rates.

Language: Английский

Citations

2
Toward Transferable Water Level Trend Monitoring Using River Cameras: Integrating Domain-Specific Models and Segment Anything Model (SAM) for Water Segmentation DOI Open Access
Ze Wang, Heng Lyu, Shunan Zhou

et al.

Authorea (Authorea), Journal Year: 2024, Volume and Issue: unknown

Published: March 21, 2024

Water level data is critical for hydrologic model calibration. The extensive river camera networks, in conjunction with advanced deep learning techniques, form the foundation imaging-based monitoring of water trends. However, limited annotated and tedious local deployment restricts applicability models new scenarios. This study proposes a novel transferable framework by combining General AI domain-specific segmentation, uses static observer flooding index (SOFI) as proxy variations. Segment Anything Model (SAM), generic computer vision Meta AI, segmenting images into discrete while semantically unknown objects. A ResUnet pre-trained on non-local dataset simultaneously identifies pixels highest probability being water, which are then overlaid onto segmented to specify object. was applied image sequences acquired from cameras stationed at four locations Tewkesbury, UK, segmentation trend monitoring. SOFI time series were calculated based masks underwent quality control using an unsupervised clustering method. obtained signal showed average correlation 0.83 real fluctuations, significantly surpassing single model's 0.54. provided qualified calibration referring both error magnitudes distribution patterns. Our has thus moved toward ease-of-use implementation

Language: Английский

Citations

0
Toward Transferable Water Level Trend Monitoring Using River Cameras: Integrating Domain-Specific Models and Segment Anything Model (SAM) for Water Segmentation DOI Open Access
Ze Wang, Heng Lyu, Shunan Zhou

et al.

Authorea (Authorea), Journal Year: 2024, Volume and Issue: unknown

Published: March 25, 2024

Water level variations influence the biochemical and hydrological processes within rivers. Through extensive river camera networks, obtaining reliable water segmentations from image data can practically support monitoring of levels. However, limited annotated tedious local deployment restrict applicability segmentation models in new scenarios. To pursue transferability, this study proposes a novel framework that combines domain-specific with General AI for segmentation. The utilizes ResUnet model pretrained on non-local dataset to identify pixels highest probability being water. Segment Anything Model (SAM), promptable foundational computer vision developed by Meta AI, is then employed use these as prompts generating masks. Different prompt modes SAM are compared. We applied sequences acquired cameras stationed at four locations Tewkesbury, UK. significantly improved performance, an increase over 15% Intersection Union (IoU) ResUnet. Meanwhile, results substantiated point more optimal mode feeding prior knowledge SAM. static observer flooding index (SOFI) time series calculated based framework’s segmented masks under exhibit average correlation 0.90 real fluctuations, surpassing single model’s 0.54. Our thus represents step toward implementing robust trend monitoring.

Language: Английский

Citations

0
Monitoring Water Level Trend Using River Cameras: Integrating Domain-Specific Models and Segment Anything Model (SAM) for Transferable Water Segmentation DOI Open Access
Ze Wang, Heng Lyu, Shunan Zhou

et al.

Authorea (Authorea), Journal Year: 2024, Volume and Issue: unknown

Published: April 1, 2024

Water level variations influence the biochemical and hydrological processes within river networks. Through cameras, obtaining reliable water segmentation from image data can practically support monitoring of level. However, limited annotated tedious local deployment restrict applicability current deep learning models in new scenarios. To pursue transferability, this study proposes a novel framework that combines domain-specific with General AI for segmentation. The utilizes ResUnet model pre-trained on non-local dataset to identify pixel highest probability being image. Segment Anything Model (SAM), promptable foundational computer vision developed by Meta AI, is then adopted use as prompt generating masks. When prompted, different modes SAM are used comparison. We applied sequences acquired cameras stationed at four locations Tewkesbury, UK. significantly improved performance, an increase over 15% Intersection Union (IoU) single model. Meanwhile, results substantiated point optimal mode feeding prior knowledge SAM. static observer flooding index (SOFI) time series calculated based framework’s segmented masks under exhibit average correlation 0.90 real fluctuations, surpassing 0.54 attained ResUnet. Our thus represents step toward implementing robust trend monitoring.

Language: Английский

Citations

0