Disaster Detection Research Articles

Optimizing Deep Learning Models for Climate-Related Natural Disaster Detection from UAV Images and Remote Sensing Data

This research study utilized artificial intelligence (AI) to detect natural disasters from aerial images. Flooding and desertification were two natural disasters taken into consideration. The Climate Change Dataset was created by compiling various open-access data sources. This dataset contains 6334 aerial images from UAV (unmanned aerial vehicles) images and satellite images. The Climate Change Dataset was then used to train Deep Learning (DL) models to identify natural disasters. Four different Machine Learning (ML) models were used: convolutional neural network (CNN), DenseNet201, VGG16, and ResNet50. These ML models were trained on our Climate Change Dataset so that their performance could be compared. DenseNet201 was chosen for optimization. All four ML models performed well. DenseNet201 and ResNet50 achieved the highest testing accuracies of 99.37% and 99.21%, respectively. This research project demonstrates the potential of AI to address environmental challenges, such as climate change-related natural disasters. This study’s approach is novel by creating a new dataset, optimizing an ML model, cross-validating, and presenting desertification as one of our natural disasters for DL detection. Three categories were used (Flooded, Desert, Neither). Our study relates to AI for Climate Change and Environmental Sustainability. Drone emergency response would be a practical application for our research project.

Utilizing Remote Sensing and GIS to Study Natural Disasters "Volcanoes" and Their Impact on Climate Change

Multifarious regions around the world are exposed to natural hazards and disasters, each with unique characteristics. A higher frequency of extreme hydro-meteorological events, most probably related to climate change, and an increase in vulnerable population have been addressed as potential causes of such disasters. To mitigate the consequences of these disasters, Disaster Risk Management, including hazard assessment, elements-at-risk mapping, vulnerability and risk assessment of spatial components as well as Earth Observation (EO) products and Geographic Information Systems (GIS), should be considered. Multihazard assessment entails the evaluation of relationships between various hazards, including interconnected or cascading events, as well as focusing on various levels from global to local community levels, as each level manifests particular objectives and spatial data. This paper presents an overview of the diverse types of spatial data and explores the methods applied in hazard and risk assessments, with volcanic eruptions serving as a specific example. The rapid development of scientific research and the advancement of Earth Observation satellites in recent years have revolutionized the concepts of geologists and researchers. These satellites now play an indispensable role in supporting first responders during major disasters. The coordination of satellite deployment ensures a swift response along with allowing for the timely delivery of critical images. In tandem, remote sensing technologies and geographic information systems (GIS) have emerged as essential tools for geospatial analysis. The application of remote sensing and GIS for the detection of natural disasters was examined through a review of academic papers, offering an analysis of how remote sensing is utilized to assess natural hazards and their link to climate change.

Geological disaster detection in underground mining tunnels using a new electromagnetic method: Theoretical modeling and experimental evaluation

Geological disaster detection in underground mining tunnels using a new electromagnetic method: Theoretical modeling and experimental evaluation

Convolutional Neural Network-based Smart Disaster Management Framework for Real-time Detection and Management of Forest Fires

Introduction: Timely detection of catastrophic natural disasters, such as forest fires, is critical to minimizing losses and ensuring rapid response. Artificial intelligence is increasingly being recognized as a valuable tool in enhancing various stages of disaster management. Method: This paper presents the development of a smart framework utilizing machine learning techniques for real-time detection and monitoring of natural disasters, specifically forest fires. The proposed approach employs a 10-layer convolutional neural network (CNN) that classifies aerial images into Fire, Non-Fire, and Smoke categories with high precision and speed. In addition to this, a CNN-based feature extraction process is performed and integrated with various ML classifiers, including support vector machine, k-nearest neighbor, decision tree, random forest, and extra trees. Results: Extensive performance analysis reveals that the proposed 10-layer CNN model outperforms other classifiers, achieving an accuracy of 97.64% in the binary classification of fire vs. nonfire and 95.61% in the three-class classification of Fire, Non-Fire and Smoke classes. Furthermore, a comparative study with existing state-of-the-art methods demonstrates the proposed model's superior performance in both accuracy and computational complexity. Conclusion: These results demonstrate the potential of the proposed CNN-based framework to serve as a reliable and effective tool for real-time disaster management across various applications, providing valuable support to emergency response teams in mitigating the impact of natural disasters.

Retrieval of cloud fraction using machine learning algorithms based on FY-4A AGRI observations

Abstract. Cloud fraction as a vital component of meteorological satellite products plays an essential role in environmental monitoring, disaster detection, climate analysis and other research areas. Random forest (RF) and multilayer perceptron (MLP) algorithms were used in this paper to retrieve the cloud fraction of AGRI (Advanced Geosynchronous Radiation Imager) on board the Fengyun-4A (FY-4A) satellite based on its full-disk level-1 radiance observation. Corrections have been made subsequently to the retrieved cloud fraction in areas where solar glint occurs using a correction curve fitted with sunglint angle as weight. The algorithm includes two steps: the cloud detection is conducted firstly for each AGRI field of view to identify whether it is clear sky, partly cloudy or overcast within the observation field. Then, the cloud fraction is retrieved for the scene identified as partly cloudy. The 2B-CLDCLASS-lidar cloud fraction product from the CloudSat and CALIPSO active remote sensing satellite is employed as the truth to assess the accuracy of the retrieval algorithm. Comparison with the operational AGRI level-2 cloud fraction product is also conducted at the same time. The results indicate that both the RF and MLP cloud detection models achieved high accuracy, surpassing that of operational products. However, both algorithms demonstrated weaker discrimination capabilities for partly cloudy conditions compared to clear-sky and overcast situations. Specifically, they tended to misclassify fields of view with low cloud fractions (e.g., cloud fraction = 0.16) as clear sky and those with higher cloud fractions (e.g., cloud fraction = 0.83) as overcast. Between the two models, RF exhibited higher overall accuracy. Both RF and MLP models performed well in cloud fraction retrieval, showing lower mean error (ME), mean absolute error (MAE) and root mean square error (RMSE) compared to operational products. The ME for both RF and MLP cloud fraction retrieval models was close to zero, while RF had slightly lower MAE and RMSE than MLP. During daytime, the high reflectance in sunglint areas led to larger retrieval errors for both RF and MLP algorithms. However, after correction, the retrieval accuracy in these regions improved significantly. At night, the absence of visible light observations from the AGRI instrument resulted in lower classification accuracy compared to daytime, leading to higher cloud fraction retrieval errors during nighttime.

Performance evaluation of NLP and CNN models for disaster detection using social media data

Performance evaluation of NLP and CNN models for disaster detection using social media data

A Comprehensive Review of Remote Sensing Technologies for Improved Geological Disaster Management

ABSTRACTOver the past two decades, the frequency of natural hazard incidents has steadily risen, leading to substantial human casualties, infrastructure destruction, societal and economic disruption. The occurrence of disasters, both of natural and human origin, has exhibited an upward trend in frequency over the past few decades on a global scale, posing a significant threat to diverse populations. Over time, remote sensing technologies have proven to be effective in analysing and monitoring diverse natural disasters, including but not limited to droughts, earthquakes, tsunamis, landslides and cyclones. The significant extent of its coverage capacity and the ability to repeat observations make its application economically efficient. This paper aims to elucidate the fundamental contributions and role of remote sensing in disaster management applications. In a comprehensive analysis, this study explores recent practical applications in the context of disaster events. The utilisation of diverse methodologies and functions of remote sensing in disaster monitoring and control is further expanded to encompass the domain of disaster risk management, employing cutting‐edge sensors and satellites from emerging technological advancements. This paper also addresses challenges related to disaster monitoring, detection and management. Emergencies, particularly during typical catastrophe scenarios, often witness partial disruptions in communication networks. Consequently, the role of alternative networks becomes paramount in enabling effective disaster detection and management strategies. Emerging issues are highlighted, and potential directions for future research are outlined and also support sustainable development goal 13 (climate action).

Asymmetric Network Combining CNN and Transformer for Building Extraction from Remote Sensing Images.

The accurate extraction of buildings from remote sensing images is crucial in fields such as 3D urban planning, disaster detection, and military reconnaissance. In recent years, models based on Transformer have performed well in global information processing and contextual relationship modeling, but suffer from high computational costs and insufficient ability to capture local information. In contrast, convolutional neural networks (CNNs) are very effective in extracting local features, but have a limited ability to process global information. In this paper, an asymmetric network (CTANet), which combines the advantages of CNN and Transformer, is proposed to achieve efficient extraction of buildings. Specifically, CTANet employs ConvNeXt as an encoder to extract features and combines it with an efficient bilateral hybrid attention transformer (BHAFormer) which is designed as a decoder. The BHAFormer establishes global dependencies from both texture edge features and background information perspectives to extract buildings more accurately while maintaining a low computational cost. Additionally, the multiscale mixed attention mechanism module (MSM-AMM) is introduced to learn the multiscale semantic information and channel representations of the encoder features to reduce noise interference and compensate for the loss of information in the downsampling process. Experimental results show that the proposed model achieves the best F1-score (86.7%, 95.74%, and 90.52%) and IoU (76.52%, 91.84%, and 82.68%) compared to other state-of-the-art methods on the Massachusetts building dataset, the WHU building dataset, and the Inria aerial image labeling dataset.

A dataset of semantic segmentation in earthquake disaster detection based on social media images

A dataset of semantic segmentation in earthquake disaster detection based on social media images

DDS-P: Stochastic models based performance of IoT disaster detection systems across multiple geographic areas

Effective management of catastrophic events in high-risk zones necessitates a holistic technological approach to protect ecosystems, biodiversity, and native populations. Limitations in sensor range and connectivity hamper real-time data gathering in secluded areas, while financial and technical hurdles hinder the creation of cost-effective, automated systems. This study presents stochastic models, the LoRaW protocol, and cloud technology to enhance sensor deployment simulations. Wireless Sensor Networks and LoRa technology are crucial for extensive monitoring and communication infrastructures. Stochastic Petri Net models optimize system components by assessing crucial performance indicators, such as average response time and system utilization, thus improving disaster response and supporting research hypotheses.

Natural disasters detection using explainable deep learning

Deep learning applications have far-reaching implications in people’s daily lives. Disaster management professionals are becoming increasingly interested in applying deep learning to prepare for and respond to natural disasters. In this paper, we aim to assist natural disaster management professionals in preparing for disasters by developing a framework that can accurately classify natural disasters and interpret the results using a combination of a deep learning model and an XAI method to ensure reliability and ease of interpretation without a technical background. Two main aspects categorize the novelty of our work. The first is utilizing pre-trained Models such as VGGNet19, ResNet50, and ViT for accurate classification of natural disaster images. The second is implementing three explainable AI techniques-Gradient-weighted Class Activation Mapping (Grad-CAM), Grad CAM++, and Local Interpretable Model-agnostic Explanations (LIME) to ensure the interpretability of the model’s predictions, making the decision-making process transparent and reliable. Experiments on the Natural disaster datasets (Niloy et al. 2021) and MEDIC with a ViT-B-32 model achieved a high accuracy of 95.23%. Additionally, explainable artificial intelligence techniques such as LIME, Grad-CAM, and Grad-CAM++ are used to evaluate model performance and visualize decision-making. Our code is available at.11https://github.com/tariqshaban/disaster-classification-with-xai.

MANET based Integrated Sensor System for Disaster Detection and Communication in Hazardous Environments

MANET based Integrated Sensor System for Disaster Detection and Communication in Hazardous Environments

Research on Tunnel Boring Machine Tunnel Water Disaster Detection and Radar Echo Signal Processing

This study focused on the detection of water inrush in tunnels excavated by full-section hard rock tunnel boring machines (TBMs) and employed ground penetrating radar methods for conducting research on radar signal processing algorithms. The research demonstrates that conventional techniques are inadequate for eliminating the interference of TBM equipment on radar signal propagation. This study employs a radar antenna array method for signal transmission, utilizing a wavelet double-threshold filtering algorithm and wave propagation theory to suppress clutter. These methods exhibit strong signal reception capabilities and are effective in eliminating 13.1% of the direct wave components. The adoption of a novel, efficient radar signal imaging algorithm simplifies the imaging process. Results of verification indicate that the synthetic aperture algorithm, enhanced with cross-correlation calculation, yields the optimal imaging effect. This investigation, which was conducted in conjunction with the construction of a diversion tunnel in a specific region, has confirmed the applicability of the ground penetrating radar method for the detection of water inrush in TBM tunnels by conducting a comparative analysis of the direct wave removal algorithm and the integration of the optimal imaging algorithm. The innovative application of ground penetrating radar within TBM tunnels, along with a targeted technology to mitigate signal interference from metal equipment, has led to the selection of an appropriate algorithm for both signal processing and imaging. This approach offers a novel solution for the detection of water source disasters in TBM tunnels.

Deteksi Pola Kejadian Bencana Menggunakan Algoritma Naïve Bayes di Kabupaten Boyolali

Boyolali Regency, which is geographically located between Mount Merbabu and the still active Mount Merapi, does have high potential as a disaster-prone area. According to data from the Boyolali Regional Disaster Management Agency, during the period from 2022 to 2023, 936 disaster events were recorded in the region. These disasters have a devastating impact on ecosystems and the environment, causing significant material losses, causing psychological disorders, and even threatening human safety. In this context, detecting patterns of disaster events is an important step in disaster mitigation and management efforts. This research was conducted with the aim of understanding patterns of disaster events based on data collected by BPBD Boyolali. The method used is the Naive Bayes algorithm, which has proven to be a powerful tool in data analysis. The results of this research show a high level of accuracy of 92.52%, precision of 92.68%, and recall of 91.64% using the Naive Bayes algorithm. This indicates that the algorithm is effective in recognizing and classifying patterns of disaster events based on existing data. Thus, the results of this research can provide a valuable contribution to the development of an early disaster detection system, as well as assist the authorities in designing more effective mitigation strategies to protect the community and environment in Boyolali Regency.

Prospects of utilizing the Korean satellite program for geological disaster detection and analysis

Prospects of utilizing the Korean satellite program for geological disaster detection and analysis

Agricultural Disaster Prevention System: Insights from Taiwan’s Adaptation Strategies

In response to the adverse effects of climate change-induced frequent extreme disasters on agricultural production and supply stability, this study develops a comprehensive agricultural disaster prevention system based on current adaptation strategies for mitigating agricultural meteorological disasters. The primary goal is to enhance disaster preparedness and recovery through three core platforms: a fine-scale weather forecast service system, a crop disaster early warning system, and an agricultural information service platform for disasters. The results show that every major agricultural production township in Taiwan now has dedicated agricultural weather stations and access to refined weather forecasts. Additionally, a disaster prevention calendar for 76 important crops is established, integrating cultivation management practices and critical disaster thresholds for different growth periods. Utilizing this calendar, the crop disaster early warning system can provide timely disaster-related information and pre-disaster prevention assistance to farmers through various information dissemination tools. As a disaster approaches, the agricultural information service platform for disasters provides updates on current crop growth conditions. This service not only pinpoints areas at higher risk of disasters and vulnerable crop types but also offers mitigation suggestions to prevent potential damage. Administrative efficiency is then improved with a response mechanism incorporating drones and image analysis for early disaster detection and rapid response. In summary, the collaborative efforts outlined in this study demonstrate a proactive approach to agricultural disaster prevention. By leveraging technological advancements and interdisciplinary cooperation, the aim is to safeguard agricultural livelihoods and ensure food security in the face of climate-induced challenges.

A practicable geophysical resistivity imaging sensing approach for urban geological water disaster detection: A real-world case study

In the urban environment, the transient electromagnetic method often suffers from severe electromagnetic interference because it employs coils to perceive the electromagnetic field. To address this issue, the DC method employs electrodes to shield the electromagnetic noise in the urban environment. In this paper, a new geophysical resistivity imaging sensing approach based on a DC inversion model is proposed to explore and identify the path of underground water to solve the water disaster problem. A real-world case study is carried out in Xinjiang Region. In order to investigate the underground water conducted path, an electrode network is designed and installed for the buildings of interest to collect the DC electrical current and DC electrical potential data. Then, the proposed method establishes a DC resistivity inversion model based on the forward theory and finite-volume technique to extract the geological information from the collected data and visualize the underground water conducted path. Different from existing popular numerical techniques, such as the finite element and finite difference, the finite-volume technique has clear theoretical explanations to make the obtained geological images understandable. As a result, the underground water path can be directly observed from the geological images calculated by the proposed approach to successfully identify and locate the underground water disaster in real-world applications.

Анализ и перспективы применения методов машинного обучения для чрезвычайных ситуаций

This article focuses on fire research, disaster and hazard forecasting, risk and vulnerability assessment, early detection of disasters, early warning systems, disaster monitoring, damage assessment and post-disaster response, and case studies.

Implementation of Lightweight Convolutional Neural Networks with an Early Exit Mechanism Utilizing 40 nm CMOS Process for Fire Detection in Unmanned Aerial Vehicles.

The advancement of unmanned aerial vehicles (UAVs) enables early detection of numerous disasters. Efforts have been made to automate the monitoring of data from UAVs, with machine learning methods recently attracting significant interest. These solutions often face challenges with high computational costs and energy usage. Conventionally, data from UAVs are processed using cloud computing, where they are sent to the cloud for analysis. However, this method might not meet the real-time needs of disaster relief scenarios. In contrast, edge computing provides real-time processing at the site but still struggles with computational and energy efficiency issues. To overcome these obstacles and enhance resource utilization, this paper presents a convolutional neural network (CNN) model with an early exit mechanism designed for fire detection in UAVs. This model is implemented using TSMC 40 nm CMOS technology, which aids in hardware acceleration. Notably, the neural network has a modest parameter count of 11.2 k. In the hardware computation part, the CNN circuit completes fire detection in approximately 230,000 cycles. Power-gating techniques are also used to turn off inactive memory, contributing to reduced power consumption. The experimental results show that this neural network reaches a maximum accuracy of 81.49% in the hardware implementation stage. After automatic layout and routing, the CNN hardware accelerator can operate at 300 MHz, consuming 117 mW of power.

Deep Learning-Based Landslide Recognition Incorporating Deformation Characteristics

Landslide disasters pose a significant threat, with their highly destructive nature underscoring the critical importance of timely and accurate recognition for effective early warning systems and emergency response efforts. In recent years, substantial advancements have been made in the realm of landslide recognition (LR) based on remote sensing data, leveraging deep learning techniques. However, the intricate and varied environments in which landslides occur often present challenges in detecting subtle changes, especially when relying solely on optical remote sensing images. InSAR (Interferometric Synthetic Aperture Radar) technology emerges as a valuable tool for LR, providing more detailed ground deformation data and enhancing the theoretical foundation. To harness the slow deformation characteristics of landslides, we developed the FCADenseNet model. This model is designed to learn features and patterns within ground deformation data, with a specific focus on improving LR. A noteworthy aspect of our model is the integration of an attention mechanism, which considers various monitoring factors. This holistic approach enables the comprehensive detection of landslide disasters across entire watersheds, providing valuable information on landslide hazards. Our experimental results demonstrate the effectiveness of the FCADenseNet model, with an F1-score of 0.7611, which is 9.53% higher than that of FC_DenseNet. This study substantiates the feasibility and efficacy of combining InSAR with deep learning methods for LR. The insights gained from this research contribute to the advancement of regional landslide geological hazard monitoring, identification, and prevention strategies.