Research on Forest Fire Detection and Segmentation Based on MST++ Hyperspectral Reconstruction Technology

Forest fires pose a potential threat to the ecological and environmental systems and natural resources, impacting human lives. However, automated surveillance system for early forest fire detection can mitigate such calamities and protect the environment. Therefore, we propose a UAV-based forest fire fighting system with integrated artificial intelligence (AI) capabilities for continuous forest surveillance and fire detection. The major contributions of the proposed research are fourfold. Firstly, we explain the detailed working mechanism along with the key steps involved in executing the UAV-based forest fire fighting system. Besides, a robust forest fire detection system requires precise and efficient classification of forest fire imagery against no-fire. Moreover, we have curated a novel dataset (DeepFire) containing diversified real-world forest imagery with and without fire to assist future research in this domain. The DeepFire dataset consists of 1900 colored images with 950 each for fire and no-fire classes. Next, we investigate the performance of various supervised machine learning classifiers for the binary classification problem of detecting forest fire. Furthermore, we propose a VGG19-based transfer learning solution to achieve improved prediction accuracy. We assess and compare the performance of several machine learning approaches such as k -nearest neighbors, random forest, naive Bayes, support vector machine, logistic regression, and the proposed approach for accurately identifying fire and no-fire images in the DeepFire dataset. The simulation results demonstrate the efficacy of the proposed approach in terms of accurate classification, where it achieves the mean accuracy of 95% with 95.7% precision and 94.2% recall.

With the influence of climate change and human activities, the frequency and scale of forest fires have been increasing continuously, posing a significant threat to the environment and human safety. Therefore, rapid and accurate forest fire detection has become essential for effective control of forest fires. This study proposes a Forest Fire Detection and Segmentation Model (FFDSM) based on unmanned aerial vehicle (UAV) infrared images to address the problems of forest fire occlusion and the poor adaptability of traditional forest fire detection methods. The FFDSM integrates the YOLO (You Only Look Once) v5s-seg, Efficient Channel Attention (ECA), and Spatial Pyramid Pooling Fast Cross-Stage Partial Channel (SPPFCSPC) to improve the detection accuracy of forest fires of different sizes. The FFDSM enhances the detection and extraction capabilities of forest fire features, enabling the accurate segmentation of forest fires of different sizes and shapes. Furthermore, we conducted ablation and controlled experiments on different attention mechanisms, spatial pyramid pooling (SPP) modules, and fire sizes to verify the effectiveness of the added modules and the adaptability of the FFDSM model. The results of the ablation experiment show that, compared to the original YOLOv5s-seg model, the models fused with the ECA and SPPFCSPC achieve an improved accuracy, with FFDSM showing the greatest improvement. FFDSM achieves a 2.1% increase in precision, a 2.7% increase in recall, a 2.3% increase in mAP@0.5, and a 4.2% increase in mAP@0.5:0.95. The results of the controlled experiments on different attention mechanisms and SPP modules demonstrate that the ECA+SPPFCSPC model (FFDSM) performs the best, with a precision, recall, mAP@0.5, and mAP@0.5:0.95 reaching 0.959, 0.870, 0.907, and 0.711, respectively. The results of the controlled experiment on different fire sizes show that FFDSM outperforms YOLOv5s-seg for all three fire sizes, and it performs the best for small fires, with a precision, recall, mAP@0.5, and mAP@0.5:0.95 reaching 0.989, 0.938, 0.964, and 0.769, respectively, indicating its good adaptability for early forest fire detection. The results indicate that the forest fire detection model based on UAV infrared images (FFDSM) proposed in this study exhibits a high detection accuracy. It is proficient in identifying obscured fires in optical images and demonstrates good adaptability in various fire scenarios. The model effectively enables real-time detection and provides early warning of forest fires, providing valuable support for forest fire prevention and scientific decision making.

Forests are a vital natural resource that directly influences the ecosystem. Recently, forest fire has been a serious issue due to natural and man-made climate effects. For early forest fire detection, an artificial intelligence-based forest fire detection method in smart city application is presented to avoid major disasters. This research presents a review of the vision-based forest fire localization and classification methods. Furthermore, this work makes use of the forest fire detection dataset, which solves the classification problem of discriminating fire and no-fire images. This work proposes a deep learning method named FFireNet, by leveraging the pre-trained convolutional base of the MobileNetV2 model and adding fully connected layers to solve the new task, that is, the forest fire recognition problem, which helps in classifying images as forest fires based on extracted features which are symmetrical. The performance of the proposed solution for classifying fire and no-fire was evaluated using different performance metrics and compared with other CNN models. The results show that the proposed approach achieves 98.42% accuracy, 1.58% error rate, 99.47% recall, and 97.42% precision in classifying the fire and no-fire images. The outcomes of the proposed approach are promising for the forest fire classification problem considering the unique forest fire detection dataset.

Forest fires are occurring throughout the year with an increasing intensity in the summer and autumn periods. These events are mainly caused by the actions of humans, but different nature and environmental phenomena, like lightning strikes or spontaneous combustion of dried leafs or sawdust, can also be credited for their occurrence. Regardless of the reasons for the ignition of the forest fires, they usually cause devastating damage to both nature and humans. Forest fires are also considered as a main contributor to the air pollution, due to the fact that during every fire huge amounts of gases and particle mater are released in the atmosphere. To fight forest fires, different solutions were employed throughout the years. They ware primary aimed at the early detection of the fires. The simplest of these solutions is the establishment of a network of observation posts - both cheap and easy to accomplish, but also time-consuming for the involved people. The constant evolution of the information and communication technologies has led to the introduction of a new generation of solutions for early detection and even prevention of forest fires. ICT-based networks of cameras and sensors and even satellite-based solutions were developed and used in the last decades. These solutions have greatly decreased the direct involvement of humans in the forest fire detection process, but have also proven to be expensive and hard to maintain. In this paper we will discuss and present two different emerging solutions for early detection of forest fires. The first of these solutions involves the use of unmanned aerial vehicles (UAVs) with specialized cameras. Several different scenarios for the possible use of the drones for forest fire detection will be presented and analysed, including a solution with the use of a combination between a fixed-wind and a rotary-wing UAVs. In the next chapter of the paper, we will present and discuss the possibilities for development of systems for early forest fire detection using LoRaWAN sensor networks and we will analyse and present some of the hardware and software components for the realisation of such sensor networks. The paper will also provide another point-of-view, which will present the involvement of students in the development and in the use of both systems and we will analyse the advantages and the benefits, which the students will gain from their work on and with these solutions.

In this study, the design of a microcontroller based early forest fire detection and warning system is described. The system can be installed in a forest and it consists of a network of microcontroller based fire detection sensors and long-range RF transmitter–receiver modules. As soon as a fire is detected, the system sends warning messages to the relevant emergency command and control centres so that appropriate action can be taken as soon as possible to put off the fire and minimize the damage. Objectives: The objective of this study is to develop and design a low power microcontroller based early forest fire detection system. Methods/statistical analysis: The system described in this study consists of a network of microcontroller based fire detection sensors and long-range raw RF transmitter/receiver modules, scattered around a forest. The system uses low power microcontrollers which are kept in deep sleep mode most of the time, thus conserving power which is one of the novelties of the designed system. Findings: One of the aims of this study has been to design a low power early forest fire detection and warning system. The designed system makes use of specially selected microcontrollers which can operate in deep sleep modes and as a result they can be configured to consume low currents. The results showed that the current consumption in deep sleep mode was only 10 mA, which is in general much lower than the current consumption of most ordinary microcontrollers. With such low power consumption, it should be possible to operate the overall system for longer hours before the batteries need charging via solar energy cells. Most forest fire detection systems rely on Bluetooth, ZigBee, WiFi, or similar technology for communication, which is not practical to implement in a forest environment. The designed system makes use of raw RF for communication which is normally available and very reliable. Application/improvements: The early forest fire detection and warning system described in this article can be improved by using arrays of multiple scattered sensors. Keywords: Forest Fire, Early Fire Detection and Warning, Early Warning System, ESP32, DevkitC.

Self-powered forest fire alarm system based on impedance matching effect between triboelectric nanogenerator and thermosensitive sensor

Forest fires are destructive natural disasters that not only destroy vast forested areas but also threaten biodiversity, degrade air quality, damage agricultural land, and accelerate climate change. Due to rising global temperatures, prolonged droughts, and human-induced factors, the frequency and intensity of forest fires are increasing year by year. Consequently, the early detection and rapid classification of forest fires are critical for preventing loss of life and property and ensuring the effective management of disaster response processes. This study aims to present a deep learning-based approach for the early detection and classification of forest fires. In this context, four advanced convolutional neural network (CNN) architectures (Xception, InceptionV3, DenseNet121, and EfficientNetV2), which have shown outstanding success in image classification tasks in recent years, were comparatively evaluated for the classification of forest fire images. Training and testing procedures were performed using the Forest Fire Images dataset, consisting of fire and non-fire classes. The experimental results revealed that all models performed well in forest fire classification; however, the Xception model demonstrated superior performance, exhibiting higher accuracy than the others. These results emphasize that deep learning architectures are effective tools for the rapid and accurate classification of forest fires, thereby making significant contributions to forest fire monitoring and management strategies.

As we all know, the forest is considered as one of the most important and indispensable resources, the prevention and detection of the forest fire, have been researched hotly in worldwide Forest Fire Prevention Departments. Based on the deficiencies of conventional forest fire detection on real time and monitoring accuracy, the wireless sensor network technique for forest fire detection was introduced, together with satellite monitoring, aerial patrolling and manual watching, an omni-bearing and stereoscopic air and ground forest-fire detection pattern was found so that the decision for fire-extinguishing or fire prevention can be made rightly and real-timely by related government departments. A cluster-based wireless sensor network paradigm for forest fire real-time detection was put forward in this paper. Some key questions were discussed emphatically, such as the ad hoc network related technology, the node hardware designing, the forest-fire forecasting model and the propagation characteristic of UHF wireless signal and so on.

Forest fires are caused naturally by lightning, high atmospheric temperatures, and dryness. Forest fires have ramifications for both climatic conditions and anthropogenic ecosystems. According to various research studies, there has been a noticeable increase in the frequency of forest fires in India. Between 1 January and 31 March 2022, the country had 136,604 fire points. They activated an alerting system that indicates the location of a forest fire detected using MODIS sensor data from NASA Aqua and Terra satellite images. However, the satellite passes the country only twice and sends the information to the state forest departments. The early detection of forest fires is crucial, as once they reach a certain level, it is hard to control them. Compared with the satellite monitoring and detection of fire incidents, video-based fire detection on the ground identifies the fire at a faster rate. Hence, an unmanned aerial vehicle equipped with a GPS and a high-resolution camera can acquire quality images referencing the fire location. Further, deep learning frameworks can be applied to efficiently classify forest fires. In this paper, a cheaper UAV with extended MobileNet deep learning capability is proposed to classify forest fires (97.26%) and share the detection of forest fires and the GPS location with the state forest departments for timely action.

Forest fires can destroy forest and inflict great damage to the ecosystem. Fortunately, forest fire detection with video has achieved remarkable results in enabling timely and accurate fire warnings. However, the traditional forest fire detection method relies heavily on artificially designed features; CNN-based methods require a large number of parameters. In addition, forest fire detection is easily disturbed by fog. To solve these issues, a lightweight YOLOX-L and defogging algorithm-based forest fire detection method, GXLD, is proposed. GXLD uses the dark channel prior to defog the image to obtain a fog-free image. After the lightweight improvement of YOLOX-L by GhostNet, depth separable convolution, and SENet, we obtain the YOLOX-L-Light and use it to detect the forest fire in the fog-free image. To evaluate the performance of YOLOX-L-Light and GXLD, mean average precision (mAP) was used to evaluate the detection accuracy, and network parameters were used to evaluate the lightweight effect. Experiments on our forest fire dataset show that the number of the parameters of YOLOX-L-Light decreased by 92.6%, and the mAP increased by 1.96%. The mAP of GXLD is 87.47%, which is 2.46% higher than that of YOLOX-L; and the average fps of GXLD is 26.33 when the input image size is 1280 × 720. Even in a foggy environment, the GXLD can detect a forest fire in real time with a high accuracy, target confidence, and target integrity. This research proposes a lightweight forest fire detection method (GXLD) with fog removal. Therefore, GXLD can detect a forest fire with a high accuracy in real time. The proposed GXLD has the advantages of defogging, a high target confidence, and a high target integrity, which makes it more suitable for the development of a modern forest fire video detection system.

Natural hazards are a challenge for the society. Scientific community efforts have been severely increased assessing tasks about prevention and damage mitigation. The most important points to minimize natural hazard damages are monitoring and prevention. This work focuses particularly on forest fires. This phenomenon depends on small-scale factors and fire behavior is strongly related to the local weather. Forest fire spread forecast is a complex task because of the scale of the phenomena, the input data uncertainty and time constraints in forest fire monitoring. Forest fire simulators have been improved, including some calibration techniques avoiding data uncertainty and taking into account complex factors as the atmosphere. Such techniques increase dramatically the computational cost in a context where the available time to provide a forecast is a hard constraint. Furthermore, an early mapping of the fire becomes crucial to assess it. In this work, a non-supervised method for forest fire early detection and mapping is proposed. As main sources, the method uses daily thermal anomalies from MODIS and VIIRS combined with land cover map to identify and monitor forest fires with very few resources. This method relies on a clustering technique (DBSCAN algorithm) and on filtering thermal anomalies to detect the forest fires. In addition, a concave hull (alpha shape algorithm) is applied to obtain rapid mapping of the fire area (very coarse accuracy mapping). Therefore, the method leads to a potential use for high-resolution forest fire rapid mapping based on satellite imagery using the extent of each early fire detection. It shows the way to an automatic rapid mapping of the fire at high resolution processing as few data as possible.

In recent years, the frequency and severity of forest fire occurrence have increased, compelling the research communities to actively search for early forest fire detection and suppression methods. Remote sensing using computer vision techniques can provide early detection from a large field of view along with providing additional information such as location and severity of the fire. Over the last few years, the feasibility of forest fire detection by combining computer vision and aerial platforms such as manned and unmanned aerial vehicles, especially low cost and small-size unmanned aerial vehicles, have been experimented with and have shown promise by providing detection, geolocation, and fire characteristic information. This paper adds to the existing research by proposing a novel method of detecting forest fire using color and multi-color space local binary pattern of both flame and smoke signatures and a single artificial neural network. The training and evaluation images in this paper have been mostly obtained from aerial platforms with challenging circumstances such as minuscule flame pixels, varying illumination and range, complex backgrounds, occluded flame and smoke regions, and smoke blending into the background. The proposed method has achieved F1 scores of 0.84 for flame and 0.90 for smoke while maintaining a processing speed of 19 frames per second. It has outperformed support vector machine, random forest, Bayesian classifiers and YOLOv3, and has demonstrated the capability of detecting challenging flame and smoke regions of a wide range of sizes, colors, textures, and opacity.

Over the years, early detection of forest and land fires has been conducted using hotspot data provided by the National Institute of Aeronautics and Space (LAPAN), based on its interpretation of satellite images. The hotspot data have tremendously helped firefighting efforts and further enforcement. However, the system has several shortcomings, especially due to its inability to distinguish forest and land fires from other hot surfaces or fires caused by common human activities. Furthermore, this method also requires labor-intensive verification, and heavily relies on human factors for advanced analysis and validation. Recently, the DG of Law Enforcement of the Ministry of Environment and Forestry (DGLE MoEF) has been piloting a new approach through advancement in artificial intelligence, called Geospatial Artificial Intelligence (GeoAI). By utilizing recorded satellite image data from 2017 - 2019, the machine has been trained to recognize the pattern and tone of the image in burnt areas so that it can validate the presence of the burnt area based on the history of Sentinel-2 imagery for the past week at each cluster. DGLE MoEF found that the burnt area data processed by GeoAI has better accuracy than the hotspot count for forest and land fire identification. Moreover, GeoAI may ease forest and land fire analysis and verification by automatically overlaying forest area and company concessions at the burnt area. GeoAI’s innovation in forest and land fire monitoring can produce more accurate and complete early detection data of forest and land fires than currently available hotspot data. The results of hotspot clustering that detect fires may assist firefighters in rapidly extinguishing the fire, and support law enforcement officers in determining the appropriate target location. Therefore, GeoAI technology may increase the effectiveness and efficiency of resources allocated by law enforcement officers in providing better and more responsive public services. Keywords: GeoAI, geospatial technology, artificial intelligence, forest fire, hotspot

According to infrared imaging features of forest fire, we use image processing technology which is conducive to early detection and prevention of forest fires. We use image processing technology based on infrared imaging features of forest fire which is conducive to early detection and prevention of forest fires. In order to the timeliness and accuracy of fire detection, this paper proposes a forest fire detection method based on infrared image technology. We take gray histogram analysis to collected Cruising image. The image which will be detected is segmented by the adaptive dynamic threshold. Then the suspected ignitions are extracted in the image after segmentation. The ignition of forest fire which form image in the infrared image is almost circular. We use the circular degree of suspected ignition as the decision basis of the fire in the infrared image. Through the analysis of position correlation which is the same suspected ignition between adjacent frames, we judge whether there is a fire in the image. In order to verify the effectiveness of the method, we adopt image sequences of forest fire to do experiment. The experimental results show that the proposed algorithm under the conditions of different light conditions and complex backgrounds, which can effectively eliminate distractions and extract the fire target. The accuracy fire detection rate is above 95 percent. All fire can be detected. The method can quickly identify fire flame and high-risk points of early fire. The structure of method is clear and efficient which processing speed is less than 25 frames per second. So it meets the application requirement of real-time processing.

The frequent occurrence of forest fires causes irreparable damage to the environment and the economy. Therefore, the accurate detection of forest fires is particularly important. Due to the various shapes and textures of flames and the large variation in the target scales, traditional forest fire detection methods have high false alarm rates and poor adaptability, which results in severe limitations. To address the problem of the low detection accuracy caused by the multi-scale characteristics and changeable morphology of forest fires, this paper proposes YOLOv5s-CCAB, an improved multi-scale forest fire detection model based on YOLOv5s. Firstly, coordinate attention (CA) was added to YOLOv5s in order to adjust the network to focus more on the forest fire features. Secondly, Contextual Transformer (CoT) was introduced into the backbone network, and a CoT3 module was built to reduce the number of parameters while improving the detection of forest fires and the ability to capture global dependencies in forest fire images. Then, changes were made to Complete-Intersection-Over-Union (CIoU) Loss function to improve the network’s detection accuracy for forest fire targets. Finally, the Bi-directional Feature Pyramid Network (BiFPN) was constructed at the neck to provide the model with a more effective fusion capability for the extracted forest fire features. The experimental results based on the constructed multi-scale forest fire dataset show that YOLOv5s-CCAB increases AP@0.5 by 6.2% to 87.7%, and the FPS reaches 36.6. This indicates that YOLOv5s-CCAB has a high detection accuracy and speed. The method can provide a reference for the real-time, accurate detection of multi-scale forest fires.