A Wildfire Smoke Detection System Using Unmanned Aerial Vehicle Images Based on the Optimized YOLOv5

Sensors, Journal Year: 2022, Volume and Issue: 22(23), P. 9384 - 9384

Published: Dec. 1, 2022

Wildfire is one of the most significant dangers and serious natural catastrophe, endangering forest resources, animal life, human economy. Recent years have witnessed a rise in wildfire incidents. The two main factors are persistent interference with environment global warming. Early detection fire ignition from initial smoke can help firefighters react to such blazes before they become difficult handle. Previous deep-learning approaches for been hampered by small or untrustworthy datasets, making it challenging extrapolate performances real-world scenarios. In this study, we propose an early system using unmanned aerial vehicle (UAV) images based on improved YOLOv5. First, curated 6000-wildfire image dataset existing UAV images. Second, optimized anchor box clustering K-mean++ technique reduce classification errors. Then, network's backbone spatial pyramid pooling fast-plus layer concentrate small-sized regions. Third, bidirectional feature network was applied obtain more accessible faster multi-scale fusion. Finally, pruning transfer learning were implemented refine architecture speed, correctly identify small-scale areas. experimental results proved that proposed method achieved average precision 73.6% outperformed other one- two-stage object detectors custom dataset.

Language: Английский

---

Morphological Characteristics and Sustainable Adaptive Reuse Strategies of Regional Cultural Architecture: A Case Study of Fenghuang Ancient Town, Xiangxi, China

Buildings, Journal Year: 2025, Volume and Issue: 15(1), P. 119 - 119

Published: Jan. 2, 2025

This study explores the application of sustainable adaptive reuse strategies in preserving historical and cultural buildings, with goal establishing a scientifically robust framework indicators for systematic evaluation practical application. Focusing on diaojiaolou (stilted buildings) along Tuojiang River ancient town Fenghuang, research integrates findings from an extensive literature review theoretical to propose multidimensional indicator system encompassing preservation, rehabilitation, urban renewal, reuse, revitalization, restoration. Through detailed field investigations comprehensive case studies, examines significance stilted including their spatial organization, structural design, material composition, decorative features. It also critically assesses effectiveness promoting regional development. The reveal that these not only preserve integrity buildings but enhance functionality meet modern needs, achieving balance among heritage conservation, environmental sustainability, socio-economic progress. validates scientific rigor utility proposed six-dimensional as evaluative tool, offering model development preservation similar assets.

Language: Английский

---

Towards a wildfire vulnerability index using expert judgement

International Journal of Wildland Fire, Journal Year: 2025, Volume and Issue: 34(1)

Published: Jan. 16, 2025

Background Climate change is expected to significantly wildfire frequency and magnitude, which poses particular challenges for countries with limited experience in managing wildfires, such as Austria. Aims To develop a vulnerability index, this study aims at weighting the characteristics of buildings their surroundings (vulnerability indicators) by involving different national stakeholders (Austria) international experts. Methods Expert judgement analytic hierarchy process (AHP) were used indicators identified previous through literature review. Key results The two expert groups regarding compared. A index that combines all into single assigned each building introduced. Conclusions tool decision-makers other end users sets foundations ongoing research field assessment based on trans-disciplinary approaches both academia stakeholders. Implications can be support decision-making, risk reduction climate adaptation strategies but it also guide local level.

Language: Английский

---

Quantifying wildfire risk to the built environment in rural rangelands of the US Interior West

Philosophical Transactions of the Royal Society B Biological Sciences, Journal Year: 2025, Volume and Issue: 380(1924)

Published: April 1, 2025

Fire increasingly conflicts with the built environment. The wildland–urban interface (WUI) describes areas where vegetation near environment increases wildfire hazard. In United States, attention concentrates on WUI in forested areas, but human populations are extending into rangelands. combination of expansion and woody plant encroachment might present novel challenges to management, especially given rural nature rangelands US, which extends response time emergency services. We use publicly available data describe abundance, distribution, type overall risk Most US Interior West (54%) occurs rangeland: majority is rangeland 4.3% that—over 1 million km 2 —is WUI. rural: 59% further than 10 from town tribal even more remote. Rangeland approximately twice as likely be degraded by non-WUI rangeland, suggesting that conventional fire suppression tactics for fuels insufficient or unsafe. Greater awareness help leverage community-level adaptive capacity against protecting lives property beyond urban/peri-urban zones. This article part theme issue ‘Novel regimes under climate changes influences: impacts, ecosystem responses feedbacks’.

Language: Английский

---

Wildfire susceptibility mapping by incorporating damage proxy maps, differenced normalized burn Ratio, and deep learning algorithms based on sentinel-1/2 data: a case study on Maui Island, Hawaii

GIScience & Remote Sensing, Journal Year: 2024, Volume and Issue: 61(1)

Published: May 14, 2024

Climate change has contributed to the recent increase in wildfire occurrences, vegetation failures, human health risks, physical damage, and economic losses. Wildfire susceptibility mapping is an essential technique for assessing areas prone wildfires. In this study, we proposed combination of damage proxy map (DPM) differenced normalized burn ratio (dNBR) method generate a precise inventory used it predict susceptible wildfire. The maps were produced using frequency (FR), convolutional neural network (CNN), long short-term memory (LSTM)-based deep learning their performances compared. We implemented on Maui Island, Hawaii, where wildfires frequently occur. started process by generating from 2019 2023 based DPM applied Sentinel-1 synthetic aperture radar (SAR) data combined with dNBR retrieved Sentinel-2 data. was randomly divided into training dataset (70%) testing (30%). Fifteen wildfire-related factors, including topographical, meteorological, land use, environmental, anthropological selected factors conducting study literature considering spatial correlation analysis FR method, information gain (IGR), multicollinearity assessment tolerance (TOL) variance inflation factor (VIF) metrics. level area five, namely very high, moderate, low, low. FR, CNN, LSTM similar patterns, significantly influenced use rainfall factors. highly are located gentle slopes covered agricultural unhealthy vegetation, these have low intensity but receive high levels solar radiation. Meanwhile, relatively occur forests wet canopy evaporation. prediction results evaluated under receiver operating characteristic (ROC) curve (AUC), CNN performed slightly better than LSTM, AUC values 0.879, 0.877, 0.870, respectively. Hence, algorithm appropriate, specifically area. addition, provides public awareness.

Language: Английский

---

Fire Hazards: Socio-economic and Regional Issues

Springer eBooks, Journal Year: 2024, Volume and Issue: unknown

Published: Jan. 1, 2024

Language: Английский

---

Disaster Risk Management and Spatial Planning: Evidence from the Fire-Stricken Area of Mati, Greece

Sustainability, Journal Year: 2023, Volume and Issue: 15(12), P. 9776 - 9776

Published: June 19, 2023

The debate over spatial planning highlights the need for more interdisciplinary, strategic, and collaborative methods to achieve broad policy goals such as resilience sustainability. Risk-based is gaining importance due rising vulnerability of urban infrastructure. Incorporating disaster risk management into requires a geographically based strategy reducing catastrophe risk. This article outlines role in reconstruction Mati settlement Attica, Greece, that was devastated by forest fire 2018. It presents set proposals relate reorganization area considers reduction management, well sustainability issues relating mobility, natural environment, recovery coastline public resource. basis this contribution Urban Planning Research Laboratory National Technical University Athens/School Architecture preparation Special Plan fire-stricken Mati, on behalf Chamber Greece.

Language: Английский

---

Ignition Locations and Simplified Design Guidelines for Enhancing the Resilience of Dwellings against Wildland Fires

Fire, Journal Year: 2024, Volume and Issue: 7(2), P. 40 - 40

Published: Jan. 28, 2024

This paper presents a study based on new fireproof design guidelines for dwellings against the impact of wildfires. The main objective is to present results from surveys large wildfires 2017 in Portugal, identifying vulnerabilities that may result spot ignitions when exposed Utilizing information gathered these surveys, it possible recommend fire resistance and reaction class requirements using European indoor standards adapting them suit wildfire conditions. focuses classical predominantly located high-risk zones within wildland–urban interface. These assessments have potential generate construction recommendations employing traditional materials commonly found industry.

Language: Английский

---

Wildland–Urban Interface: Definition and Physical Fire Risk Mitigation Measures, a Systematic Review

Fire, Journal Year: 2023, Volume and Issue: 6(9), P. 343 - 343

Published: Sept. 1, 2023

Due to the associated fire risk, wildland–urban interface (WUI) has drawn attention of researchers and managers from a range backgrounds. From land management point view, it is important identify WUI determine areas prioritise for risk prevention. It also know mitigation measures available select most appropriate each specific context. In this systematic review, definitions were investigated physical reducing examined perspective. The PRISMA 2020 Statement was applied records published until 31 December 2022 retrieved Web Science, Scopus, other research engines. A total 162 publications scientific journals grey literature scrutinised selected analysis. Only providing an original definition or proposing reduce at retained, while those relating emergency social perception not considered. bias reduced by internal cross-assessment team. Definitions (n = 40 publications) changed according objective, varying broadly in identification anthropogenic wildland components interface. Terminology varied definition, term wildland–human (WHI) found be more comprehensive than WUI. Methodological ranged using aggregated data through buildings with considerable precision. Five categories 128 identified: clearance distances, landscaping, fuel management, planning, design materials. effective early stages urban development, maintenance assets vegetation crucial preparedness. This review represents analysis evidence on which can base their actions hazard number studies investigating considerable, but experimental quantitative results are scarce, better communication coordination among groups agencies advisable. registered.

Language: Английский

---

Modelling the vulnerability of urban settings to wildland–urban interface fires in Chile

Natural hazards and earth system sciences, Journal Year: 2024, Volume and Issue: 24(4), P. 1521 - 1537

Published: May 2, 2024

Abstract. Wildland–urban interface (WUI) regions are particularly vulnerable to wildfires due their proximity both nature and urban developments, posing significant risks lives property. To enhance our understanding of the risk profiles in WUI areas, we analysed seven fire case studies central Chile. We developed a mixed-method approach for conducting local-scale analyses, which involved field surveys, remote-sensing through satellite drone imagery, GIS-based analysis collected data. The methodology led generation georeferenced dataset damaged undamaged dwellings, including 16 variables representing physical characteristics, spatial arrangement, availability suppression resources. A binary classification model was then used assess relative importance these attributes as indicators vulnerability. revealed that arrangement factors have greater impact on damage prediction than structural conditions preparedness individual units. Specifically, such dwelling neighbours, distance vegetation, border groups, from origin substantially contribute damage. Other associated with less affluent homes may also increase likelihood damage, although further data required confirmation. This study provides insights design, planning, governance areas Chile, aiding development mitigation strategies built structures broader territorial area.

Language: Английский

---