A comprehensive survey of research towards AI-enabled unmanned aerial systems in pre-, active-, and post-wildfire management

Information Fusion, Journal Year: 2024, Volume and Issue: 108, P. 102369 - 102369

Published: March 22, 2024

Wildfires have emerged as one of the most destructive natural disasters worldwide, causing catastrophic losses. These losses underscored urgent need to improve public knowledge and advance existing techniques in wildfire management. Recently, use Artificial Intelligence (AI) wildfires, propelled by integration Unmanned Aerial Vehicles (UAVs) deep learning models, has created an unprecedented momentum implement develop more effective Although survey papers explored learning-based approaches wildfire, drone disaster management, risk assessment, a comprehensive review emphasizing application AI-enabled UAV systems investigating role methods throughout overall workflow multi-stage including pre-fire (e.g., vision-based vegetation fuel measurement), active-fire fire growth modeling), post-fire tasks evacuation planning) is notably lacking. This synthesizes integrates state-of-the-science reviews research at nexus observations modeling, AI, UAVs - topics forefront advances elucidating AI performing monitoring actuation from pre-fire, through stage, To this aim, we provide extensive analysis remote sensing with particular focus on advancements, device specifications, sensor technologies relevant We also examine management approaches, monitoring, prevention strategies, well planning, damage operation strategies. Additionally, summarize wide range computer vision emphasis Machine Learning (ML), Reinforcement (RL), Deep (DL) algorithms for classification, segmentation, detection, tasks. Ultimately, underscore substantial advancement modeling cutting-edge UAV-based data, providing novel insights enhanced predictive capabilities understand dynamic behavior.

Language: Английский

---

SHAP-Driven Explainable Artificial Intelligence Framework for Wildfire Susceptibility Mapping Using MODIS Active Fire Pixels: An In-Depth Interpretation of Contributing Factors in Izmir, Türkiye

Remote Sensing, Journal Year: 2024, Volume and Issue: 16(15), P. 2842 - 2842

Published: Aug. 2, 2024

Wildfire susceptibility maps play a crucial role in preemptively identifying regions at risk of future fires and informing decisions related to wildfire management, thereby aiding mitigating the risks potential damage posed by wildfires. This study employs eXplainable Artificial Intelligence (XAI) techniques, particularly SHapley Additive exPlanations (SHAP), map Izmir Province, Türkiye. Incorporating fifteen conditioning factors spanning topography, climate, anthropogenic influences, vegetation characteristics, machine learning (ML) models (Random Forest, XGBoost, LightGBM) were used predict wildfire-prone areas using freely available active fire pixel data (MODIS Active Fire Collection 6 MCD14ML product). The evaluation trained ML showed that Random Forest (RF) model outperformed XGBoost LightGBM, achieving highest test accuracy (95.6%). All classifiers demonstrated strong predictive performance, but RF excelled sensitivity, specificity, precision, F-1 score, making it preferred for generating conducting SHAP analysis. Unlike prevailing approaches focusing solely on global feature importance, this fills critical gap employing summary dependence plots comprehensively assess each factor’s contribution, enhancing explainability reliability results. analysis reveals clear associations between such as wind speed, temperature, NDVI, slope, distance villages with increased susceptibility, while rainfall streams exhibit nuanced effects. spatial distribution classes highlights areas, flat coastal near settlements agricultural lands, emphasizing need enhanced awareness preventive measures. These insights inform targeted management strategies, highlighting importance tailored interventions like firebreaks management. However, challenges remain, including ensuring selected factors’ adequacy across diverse regions, addressing biases from resampling spatially varied data, refining broader applicability.

Language: Английский

---

Detection and Segmentation of Glioma Tumors Using an Improved Visual Geometry Group (IVGG) Deep Learning Structure

Brazilian Archives of Biology and Technology, Journal Year: 2025, Volume and Issue: 68

Published: Jan. 1, 2025

Abstract Glioma brain tumors have similar textural patterns to other tumors, making their detection and segmentation a challenging process. The approach of the Modified Tumor Detection System (MTDS) is presented in this study identify categorize images gliomas from healthy brains. Spatial Gabor Transform (SGT), feature calculations, deep learning structure comprise training work flow suggested MTDS technique. features are computed glioma image dataset normal these fed into classification architecture. In paper, proposed IVGG architecture derived existing Visual Geometry Group (VGG) improve rate system decrease computational time complexity. testing also consist SGT, computation produce result source either or glioma. Furthermore, Morphological Segmentation technique has been used find tumor locations image. Two separate imaging datasets evaluate validate MTDS's performance efficiency. BRATS Imaging 2020 (BI20) Kaggle Brain (KBI) datasets. Analysis efficiency done relation Jaccard index, recall, precision, rate.

Language: Английский

---

Wildfire Early Warning System Based on a Smart CO2 Sensors Network

Sensors, Journal Year: 2025, Volume and Issue: 25(7), P. 2012 - 2012

Published: March 23, 2025

Climate change exacerbates wildfire risks in regions like the Mediterranean, where rising temperatures and prolonged droughts create ideal fire conditions. Adapting to this scenario requires implementing advanced risk management strategies that leverage cutting-edge technologies. Wildfire early warning systems are crucial tools for detecting fires at an stage, helping prevent potential future damage. This paper proposes a smart CO2 sensor network-based system, relying on platform enables connection, management, processing of data from devices through cloud. The system was tested real controlled experiment, which 44 sensors were deployed strategically selected locations varying distances fire. To enhance detection, three Artificial Intelligence (AI) models developed using AutoEncoders (AEs) Long-Short-Term Memory (LSTM), these compared simple threshold-based (NO-AI) model. All AI models, especially LSTM-based model, able extract more valuable information records, activating up 56% than NO-AI model less time tracking front propagation based wind patterns. Therefore, not only improves detection but also effectively supports firefighting operations.

Language: Английский

---

Enhancement of Forest Fire Assessment by KDCPMNN Approach in Sikkim, India Using Remote Sensing Images

Communications in computer and information science, Journal Year: 2025, Volume and Issue: unknown, P. 379 - 394

Published: Jan. 1, 2025

Language: Английский

---

A Comparative Exploration of Time Series Models for Wild Fire Prediction

Published: Jan. 11, 2024

Wildfires pose a great threat to human safety and property arising from both natural causes. According technical assessment by the Forest Survey of India more than 95% fires are anthropogenic origin. erupt due burning fossils local communities for crop rotation, camp without proper supervision etc. Climate change further elevates risk fostering dry conditions. Traditionally, wildfire prediction relied on statistical models expert judgment. However, emergence Machine Learning (ML) Deep (DL) techniques has significantly improved accuracy forest fire prediction. The objective this work is prevent wildfires save ecosystem. In work, LightGBM(Light Gradient Boosting Machine) LSTM(Long Short-Term Memory) machine learning utilized predict fire. Both exhibit high F1 scores 97% in prediction, enabling development reliable systems. results these ML-based may aid identifying highrisk areas, optimizing prevention measures, refining evacuation plans, guiding firefighting efforts.

Language: Английский

---

FireDetXplainer: Decoding Wildfire Detection With Transparency and Explainable AI Insights

IEEE Access, Journal Year: 2024, Volume and Issue: 12, P. 52378 - 52389

Published: Jan. 1, 2024

Recent analyses by leading national wildfire and emergency monitoring agencies have highlighted an alarming trend: the impact of devastation has escalated to nearly three times that a decade ago. To address this challenge, we propose FireDetXplainer (FDX), robust deep-learning model enhances interpretability often lacking in current solutions. FDX employs innovative approach, combining transfer learning fine-tuning methodologies with Learning without Forgetting (LwF) framework. A key aspect our methodology is utilization pre-trained MobileNetV3 model, renowned for its efficiency image classification tasks. Through strategic adaptation augmentation, achieved exceptional accuracy 99.91%. The further refined convolutional blocks advanced pre-processing techniques, contributing high level precision. Leveraging diverse datasets from Kaggle Mendeley, incorporates Explainable AI (XAI) tools such as Gradient Weighted Class Activation Map (Grad-CAM) Local Interpretable Model-Agnostic Explanations (LIME) comprehensive result interpretation. Our extensive experimental results demonstrate not only outperforms existing state-of-the-art models but does so remarkable accuracy, making it highly effective solution interpretable management.

Language: Английский

---

Lightweight Fire Detection Algorithm Based on Improved YOLOv5

International Journal of Advanced Computer Science and Applications, Journal Year: 2024, Volume and Issue: 15(6)

Published: Jan. 1, 2024

Among all kinds of disasters, fire is one the most frequent and common major disasters that threaten public safety social development. At present, widely used smoke sensor method to detect susceptible factors such as distance, resulting in untimely detection. With development computer vision technology, image detection technology based on machine learning has been superior traditional methods terms accuracy speed, gradually become emerging mainstream field this stage, proposed related studies are high-performance hardware devices, which limits practical application relevant results. This paper proposes an improved algorithm YOLOv5 model address issues high memory usage, slow operating costs current algorithms. The introduces FasterNet network into backbone reduce usage improve speed. Using Ghost-Shuffle Convolution (GSConv) neck reduces number parameters computational costs. Introducing a one-time aggregation cross-stage partial module (VoV-GSCSP) enhance feature extraction capability model. experimental results show compared with original model, achieves better recognition performance, average 98.3%, 31.4% reduction 13% increase decreased by 33%, workload 35%. can achieve fast accurate identification fires, lightweight more suitable for deployment implementation general embedded hardware.

Language: Английский

---

An Efficient Technique for Cloud Resource Management Using Machine Learning Model

2022 International Conference on Innovative Computing, Intelligent Communication and Smart Electrical Systems (ICSES), Journal Year: 2023, Volume and Issue: unknown, P. 1 - 4

Published: Dec. 14, 2023

Resource management is a vital process in the cloud for satisfying customer requirement. Resources get task from user and perform necessary action. The needs data which are collected various environments. existing systems not concentrating on but it maps request to resources. level provides service response user. It suffers related issues, so solved by incorporate efficient machine learning over cloud. proposed model classifies dataset based workload condition namely GPU CPU level. also developed workload. resources provisioning allocate specific region with parameters. Dynamic load assignment helps keeping cost their acceptable Various deep models have been analysed achieves high resource monitoring order handle faulty make then active

Language: Английский

---

Wildfire Smoke Detection Based on Enhanced Yolov7 and Mountain Range Line

Published: Jan. 1, 2024

Download This Paper Open PDF in Browser Add to My Library Share: Permalink Using these links will ensure access this page indefinitely Copy URL DOI

Language: Английский

---