Search Results (49)

Abnormal operating modes in the iron ore sintering process often lead to reduced productivity and inferior sinter quality. The timely early warning of such modes is therefore essential in maintaining stable production and ensuring product quality. To this end, we develop an early warning approach that integrates cross-sectional image features from the discharge end. First, an edge detection-based scheme is designed to isolate and analyze the red fire layer in the image. Second, a random forest feature importance ranking is employed to select process variables. Third, a Bayesian neural network is trained on both selected process variables and visual features extracted from the red fire layer to construct the early warning model. Finally, the burn-through point is adopted as the classification criterion, and experiments are carried out on raw data collected from an industrial plant. The results demonstrate that the proposed method enables the accurate early detection of abnormal operating modes, achieving accuracy of 94.07%, and thus holds strong potential for industrial application. Full article

Accurate real-time tracking of forest fires using UAV platforms is crucial for timely early warning, reliable spread prediction, and effective autonomous suppression. Existing detection-based multi-object tracking methods face challenges in accurately associating targets and maintaining smooth tracking trajectories in complex forest environments. These difficulties stem from the highly nonlinear movement of flames relative to the observing UAV and the lack of robust fire-specific feature modeling. To address these challenges, we introduce AO-OCSORT, an association-optimized observation-centric tracking framework designed to enhance robustness in dynamic fire scenarios. AO-OCSORT builds on the YOLOX detector. To associate detection results across frames and form smooth trajectories, we propose a temporal–physical similarity metric that utilizes temporal information from the short-term motion of targets and incorporates physical flame characteristics derived from optical flow and contours. Subsequently, scene classification and low-score filtering are employed to develop a hierarchical association strategy, reducing the impact of false detections and interfering objects. Additionally, a virtual trajectory generation module is proposed, employing a kinematic model to maintain trajectory continuity during flame occlusion. Locally evaluated on the 1080P-resolution FireMOT UAV wildfire dataset, AO-OCSORT achieves a 5.4% improvement in MOTA over advanced baselines at 28.1 FPS, meeting real-time requirements. This improvement enhances the reliability of fire front localization, which is crucial for forest fire management. Furthermore, AO-OCSORT demonstrates strong generalization, achieving 41.4% MOTA on VisDrone, 80.9% on MOT17, and 92.2% MOTA on DanceTrack. Full article

Forest fire detection is vital for ecological conservation and disaster management. Existing visual detection methods exhibit instability in smoke-obscured or illumination-variable environments. Although multimodal fusion has demonstrated potential, effectively resolving inconsistencies in smoke features across diverse modalities remains a significant challenge. This issue stems from the inherent ambiguity between regions characterized by high temperatures in infrared imagery and those with elevated brightness levels in visible-light imaging systems. In this paper, we propose MCDet, an RGB-T forest fire detection framework incorporating target-aware fusion. To alleviate feature cross-modal ambiguity, we design a Multidimensional Representation Collaborative Fusion module (MRCF), which constructs global feature interactions via a state-space model and enhances local detail perception through deformable convolution. Then, a content-guided attention network (CGAN) is introduced to aggregate multidimensional features by dynamic gating mechanism. Building upon this foundation, the integration of WIoU further suppresses vegetation occlusion and illumination interference on a holistic level, thereby reducing the false detection rate. Evaluated on three forest fire datasets and one pedestrian dataset, MCDet achieves a mean detection accuracy of 77.5%, surpassing advanced methods. This performance makes MCDet a practical solution to enhance early warning system reliability. Full article

Forest fires in the Hindu Kush Himalaya (HKH) region are increasing in frequency and severity, driven by climate variability, prolonged dry periods, and human activity. Nepal, a critical part of the HKH, recorded over 22,700 forest fire events in the past decade, with fire incidence nearly doubling in 2023. Despite this growing threat, operational early warning systems remain limited. This study presents Nepal’s first high-resolution early fire risk outlook system, developed by adopting the Canadian Fire Weather Index (FWI) using meteorological forecasts from the High-Impact Weather Assessment Toolkit (HIWAT). The system generates daily and two-day forecasts using a fully automated Python-based workflow and publishes results as Web Map Services (WMS). Model validation against MODIS, VIIRS, and ground fire records for 2023 showed that over 80% of fires occurred in zones classified as Moderate to Very High risk. Spatiotemporal analysis confirmed fire seasonality, with peaks in mid-April and over 65% of fires occurring in forested areas. The system’s integration of satellite data and high-resolution forecasts improves the spatial and temporal accuracy of fire danger predictions. This research presents a novel, scalable, and operational framework tailored for data-scarce and topographically complex regions. Its transferability holds substantial potential for strengthening anticipatory fire management and climate adaptation strategies across the HKH and beyond. Full article

Wildfires represent an increasing global concern, threatening ecosystems, human settlements, and economies. Chile, characterized by diverse climatic zones and extensive forested areas, has been particularly vulnerable to wildfire events over recent decades. In this context, real, long-term data are essential to understand wildfire dynamics and to design effective early warning and prevention systems. This paper introduces a unique dataset containing detailed wildfire occurrence and damage information across Chilean municipalities from 1985 to 2024. Derived from official records by the National Forestry Corporation of Chile CONAF, this dataset encompasses key variables such as the number of fires, total burned area, estimated material damages, and the number of affected individuals. It provides an invaluable resource for researchers and policymakers aiming to improve fire risk assessments, model fire behavior, and develop AI-driven early detection systems. The temporal span of nearly four decades offers opportunities for longitudinal analyses, the study of climate change impacts on fire regimes, and the evaluation of historical prevention strategies. Furthermore, by presenting a complete spatial coverage at the municipal level, it allows fine-grained assessments of regional vulnerabilities and resilience. Full article

Forest fires are a global environmental threat to human life and ecosystems. This study compiles smoke alarm images from five high-definition surveillance cameras in Foshan City, Guangdong, China, collected over one year, to create a smoke-based early warning dataset. The dataset presents two key challenges: (1) high false positive rates caused by pseudo-smoke interference, including non-fire conditions like cooking smoke and industrial emissions, and (2) significant regional data imbalances, influenced by varying human activity intensities and terrain features, which impair the generalizability of traditional pre-train–fine-tune strategies. To address these challenges, we explore the use of visual language models to differentiate between true alarms and false alarms. Additionally, our method incorporates a prompt tuning strategy which helps to improve performance by at least 12.45% in zero-shot learning tasks and also enhances performance in few-shot learning tasks, demonstrating enhanced regional generalization compared to baselines. Full article

Forest fires pose significant threats to ecosystems, economies, and human lives. However, existing forest fire risk assessments are over-reliant on field data and expert-derived indices. Here, we assessed the nationwide forest fire risk in South Korea using a dataset of 2289 and 4578 fire and non-fire events between 2020 and 2023. Twelve remote sensing-based environmental variables were exclusively derived from Google Earth Engine, including climate, vegetation, topographic, and socio-environmental factors. After removing the snow equivalent variable owing to high collinearity, we trained three machine learning models: random forest, XGBoost, and artificial neural network, and evaluated their ability to predict forest fire risks. XGBoost showed the best performance (F1 = 0.511; AUC = 0.76), followed by random forest (F1 = 0.496) and artificial neural network (F1 = 0.468). DEM, NDVI, and population density consistently ranked as the most influential predictors. Spatial prediction maps from each model revealed consistent high-risk areas with some local prediction differences. These findings demonstrate the potential of integrating cloud-based remote sensing with machine learning for large-scale, high-resolution forest fire risk modeling and have implications for early warning systems and effective fire management in vulnerable regions. Future predictions can be improved by incorporating seasonal, real-time meteorological, and human activity data. Full article

In recent years, the frequency and severity of large-scale forest fires have increased globally, threatening forest ecosystems, human lives, and property while potentially triggering cascading ecological and social crises. Despite significant advancements in remote sensing-based forest fire monitoring, early warning systems, and fire risk zoning, post-fire thermal damage assessment remains insufficiently addressed. This study introduces an innovative approach combining multispectral imaging with a dual annealing constrained optimization algorithm to enable dynamic monitoring of fire temperature distribution. Based on this method, we develop a dynamic thermal damage assessment model to quantify thermal impacts during forest fires. The proposed model provides valuable insights for defining thermal damage zones, optimizing evacuation strategies, and supporting firefighting operations, ultimately enhancing emergency response and forest fire management efficiency. Full article

This study investigates the latency of lightning-caused fires in the boreal coniferous forests of the Greater Khingan Mountains, employing advanced machine learning techniques to analyze the relationship between meteorological factors, lightning characteristics, and fire ignition and smoldering processes. Using the Random Forest Model (RFM) combined with Recursive Feature Elimination with Cross-Validation (RFECV) and SHapley Additive exPlanations (SHAP), the study identifies key factors influencing fire latency. Two methods, Min distance and Min latency, were used to determine ignition lightning, with the Min distance method proving more reliable. The results show that lightning-caused fires cluster spatially and peak temporally between May and July, aligning with lightning activity. The Fine Fuel Moisture Code (FFMC) and precipitation were identified as the most influential factors. This study underscores the importance of fuel moisture and weather conditions in determining latency of lightning-caused fire, offering valuable insights for enhancing early warning systems. Despite limitations in data resolution and the exclusion of topographic factors, this study advances our understanding of lightning-fire latency mechanisms and provides a foundation for more effective wildfire management strategies under climate change. Full article

Forest fires pose a significant ecological threat, particularly in the Diamer District, Gilgit-Baltistan, Pakistan, where climatic factors combined with human activities have resulted in severe fire incidents. The present study sought to investigate the correlation between the incidence of forest fires and critical meteorological elements, including temperature, humidity, precipitation, and wind speed, over a period of 25 years, from 1998 to 2023. We analyzed 169 recorded fire events, collectively burning approximately 109,400 hectares of forest land. Employing sophisticated machine learning algorithms, Random Forest (RF), and Gradient Boosting Machine (GBM) revealed that temperature and relative humidity during the critical fire season, which spans May through July, are key factors influencing fire activity. Conversely, wind speed was found to have a negligible impact. The RF model demonstrated superior predictive accuracy compared to the GBM model, achieving an RMSE of 5803.69 and accounting for 49.47% of the variance in the burned area. This study presents a novel methodology for predictive fire risk modeling under climate change scenarios in the region, offering significant insights into fire management strategies. Our results underscore the necessity for real-time early warning systems and adaptive management strategies to mitigate the frequency and intensity of escalating forest fires driven by climate change. Full article

Forest fires are among the most devastating natural disasters, causing significant environmental and economic damage. Effective early prediction mechanisms are critical for minimizing these impacts. In our previous work, we developed a smart and secure wireless sensor network (WSN) utilizing physical sensors to emulate forest fire dynamics and predict fire scenarios using machine learning. Building on this foundation, this study explores the integration of virtual sensors to enhance the prediction capabilities of the WSN. Virtual sensors were generated using polynomial regression models and incorporated into a supervector framework, effectively augmenting the data from physical sensors. The enhanced dataset was used to train a multi-layer perceptron neural network (MLP NN) to classify multiple fire scenarios, covering both early warning and advanced fire states. Our experimental results demonstrate that the addition of virtual sensors significantly improves the accuracy of fire scenario predictions, especially in complex situations like “Fire with Thundering” and “Fire with Thundering and Lightning”. The extended model’s ability to predict early warning scenarios such as lightning and smoke is particularly promising for proactive fire management strategies. This paper highlights the potential of combining physical and virtual sensors in WSNs to achieve superior prediction accuracy and scalability of the field without any extra cost. Such findings pave the way for deploying scalable (cost-effective), intelligent monitoring systems capable of addressing the growing challenges of forest fire prevention and management. We obtained significant results in specific scenarios based on the number of virtual sensors added, while in some scenarios, the results were less promising compared to using only physical sensors. However, the integration of virtual sensors enables coverage of much larger areas, making it a highly promising approach despite these variations. Future work includes further optimization of the virtual sensor generation process and expanding the system’s capability to handle large-scale forest environments. Moreover, utilizing virtual sensors will alleviate many challenges associated with the huge number of deployed physical sensors. Full article

The understory fine dead fuel moisture content (DFMC) is an important reference indicator for regional forest fire warnings and risk assessments, and determining it on a large scale is a critical goal. It is difficult to estimate understory fine DFMC directly from satellite images due to canopy shading. To address this issue, we used canopy meteorology estimated by Landsat images in combination with explanatory variables to construct random forest models of in-forest meteorology, and then construct random forest models by combining the meteorological factors and explanatory variables with understory fine DFMC obtained from the monitoring device to (1) investigate the feasibility of Landsat images for estimating in-forest meteorology; (2) explore the feasibility of canopy or in-forest meteorology and explanatory variables for estimating understory fine DFMC; and (3) compare the effects of each factor on model accuracy and its effect on understory fine DFMC. The results showed that random forest models improved in-forest meteorology estimation, enhancing in-forest relative humidity, vapor pressure deficit, and temperature by 50%, 34%, and 2.2%, respectively, after adding a topography factor. For estimating understory fine DFMC, models using vapor pressure deficit improved fit by 10.2% over those using relative humidity. Using in-forest meteorology improved fits by 36.2% compared to canopy meteorology. Including topographic factors improved the average fit of understory fine DFMC models by 123.1%. The most accurate model utilized in-forest vapor pressure deficit, temperature, topographic factors, vegetation index, precipitation data, and seasonal factors. Correlations indicated that slope, in-forest vapor pressure deficit, and slope direction were most closely related to understory fine DFMC. The regional understory fine-grained DFMC distribution mapped according to our method can provide important decision support for forest fire risk early warning and fire management. Full article

Wildfires, which are distinguished by their destructive nature and challenging suppression, present a significant threat to ecological environments and socioeconomic systems. In order to address this issue, the development of efficient and accurate fire detection technologies for early warning and timely response is essential. This paper addresses the complexity of forest and mountain fire detection by proposing YOLO-CSQ, a drone-based fire detection method built upon an improved YOLOv8 algorithm. Firstly, we introduce the CBAM attention mechanism, which enhances the model’s multi-scale fire feature extraction capabilities by adaptively adjusting weights in both the channel and spatial dimensions of feature maps, thereby improving detection accuracy. Secondly, we propose an improved ShuffleNetV2 backbone network structure, which significantly reduces the model’s parameter count and computational complexity while maintaining feature extraction capabilities. This results in a more lightweight and efficient model. Thirdly, to address the challenges of varying fire scales and numerous weak emission targets in mountain fires, we propose a Quadrupled-ASFF detection head for weighted feature fusion. This enhances the model’s robustness in detecting targets of different scales. Finally, we introduce the WIoU loss function to replace the traditional CIoU object detection loss function, thereby enhancing the model’s localization accuracy. The experimental results demonstrate that the improved model achieves an mAP@50 of 96.87%, which is superior to the original YOLOV8, YOLOV9, and YOLOV10 by 10.9, 11.66, and 13.33 percentage points, respectively. Moreover, it exhibits significant advantages over other classic algorithms in key evaluation metrics such as precision, recall, and F1 score. These findings validate the effectiveness of the improved model in mountain fire detection scenarios, offering a novel solution for early warning and intelligent monitoring of mountain wildfires. Full article

Due to its unique geographical and climatic conditions, the Liangshan Prefecture region is highly prone to large fires. There is an urgent need to study the growth rate of fire-burned areas to fill the research gap in this region. To address this issue, this study uses the Grey Wolf Optimizer (GWO) algorithm to optimize the hyperparameters in the eXtreme Gradient Boosting (XGBoost) model, constructing a GWO-XGBoost model. Finally, the optimized ensemble model (GWO-XGBoost) is used to create a fire growth rate warning map for the Liangshan Prefecture in Sichuan Province, China, filling the research gap in forest fire studies in this area. This study comprehensively selects factors such as monthly climate, monthly vegetation, terrain, and socio–economic aspects and incorporates monthly reanalysis data from forest fire assessment systems in Canada, the United States, and Australia as features to construct the forest fire dataset. After collinearity tests to filter redundant features and Pearson correlation analysis to explore features related to the burned area growth rate, the Synthetic Minority Oversampling Technique (SMOTE) is used to oversample the positive class samples. The GWO algorithm is used to optimize the hyperparameters in the XGBoost model, constructing the GWO-XGBoost model, which is then compared with XGBoost, Random Forest (RF), and Logistic Regression (LR) models. Model evaluation results showed that the GWO-XGBoost model, with an AUC value of 0.8927, is the best-performing model. Using the SHapley Additive exPlanations (SHAP) value analysis method to quantify the contribution of each influencing factor indicates that the Ignition Component (IC) value from the United States National Fire Danger Rating System contributes the most, followed by the average monthly temperature and the population density. The growth rate warning map results indicate that the southern part of the study area is the key fire prevention area. Full article

In recent years, forest fires have been occurring frequently around the globe, affected by extreme weather and dry climate, causing serious economic losses and environmental pollution. In this context, timely detection of forest fire smoke is crucial for realizing real-time early warning of fires. However, fire and smoke from forest fires can spread to cover large areas and may affect distant areas. In this paper, a lightweight joint forest fire and smoke detection network, SmokeFireNet, is proposed, which employs ShuffleNetV2 as the backbone for efficient feature extraction, effectively addressing the computational efficiency challenges of traditional methods. To integrate multi-scale information and enhance the semantic feature extraction capability, a feature pyramid network (FPN) and path aggregation network (PAN) are introduced in this paper. In addition, the FPN network is optimized by a lightweight DySample upsampling operator. The model also incorporates efficient channel attention (ECA), which can pay more attention to the detection of forest fires and smoke regions while suppressing irrelevant features. Finally, by embedding the receptive field block (RFB), the model further improves its ability to understand contextual information and capture detailed features of fire and smoke, thus improving the overall detection accuracy. The experimental results show that SmokeFireNet is better than other mainstream target detection algorithms in terms of average AP_all of 86.2%, FPS of 114, and GFLOPs of 8.4, and provides effective technical support for forest fire prevention work in terms of average precision, frame rate, and computational complexity. In the future, the SmokeFireNet model is expected to play a greater role in the field of forest fire prevention and make a greater contribution to the protection of forest resources and the ecological environment. Full article