Forest Monitoring System Research Articles

A Forest Fire Smoke Monitoring System Based on a Lightweight Neural Network for Edge Devices

Forest resources are one of the indispensable resources of the earth, which are the basis for the survival and development of human society. With the swift advancements in computer vision and artificial intelligence technology, the utilization of deep learning for smoke detection has achieved remarkable results. However, the existing deep learning models have poor performance in forest scenes and are difficult to deploy because of numerous parameters. Hence, we introduce an optimized forest fire smoke monitoring system for embedded edge devices based on a lightweight deep learning model. The model makes full use of the multi-scale variable attention mechanism of Transformer architecture to strengthen the ability of image feature extraction. Considering the needs of application scenarios, we propose an improved lightweight network model LCNet for feature extraction, which can reduce the parameters and enhance detecting ability. In order to improve running speed, a simple semi-supervised label knowledge distillation scheme is used to enhance the overall detection capability. Finally, we design and implement a forest fire smoke detection system on an embedded device, including the Jetson NX hardware platform, high-definition camera, and detection software system. The lightweight model is transplanted to the embedded edge device to achieve rapid forest fire smoke detection. Also, an asynchronous processing framework is designed to make the system highly available and robust. The improved model reduces three-fourths of the parameters and increases speed by 3.4 times with similar accuracy to the original model. This demonstrates that our system meets the precision demand and detects smoke in time.

Monitoring loss and degradation of forests and shrubs in the North of Chile using Landsat time series data sets from 1998 to 2018

Deforestation and degradation are among the main sources of climate-change-inducing greenhouse gas emissions. Implementing sustainable forest management and conservation programs, such as Reducing Emissions from Deforestation and Forest Degradation in Developing Countries (REDD+), requires a forest monitoring system and carbon inventory to define reference emissions. Most REDD+ projects focus on dense forests, with limited studies conducted in arid and semi-arid regions where shrubs dominate. Despite diverse vegetation formations, these areas are often overlooked for their carbon sequestration potential. This study estimated gross and net deforestation and degradation of forest formation and from devegetation and degradation of shrub formation in Norte Grande (Far North), Chile, from 1998 to 2018 using remote sensing satellite imagery and projected out to 2118. The carbon equivalent (Ce) content of forest and shrub formations and their associated processes’ CO2 equivalent (CO2-e) emissions were also estimated. The results showed that the total Ce content in shrubs was higher than in forests because of the larger area. Although deforestation/devegetation and degradation tended to decrease across the study period, they will continue to affect forests and shrubs in the future, reaching a deforestation rate of 21.82 and 393.28 ha y−1 by 2118, respectively, and an annual degradation rate of 9535.31 and 255,850.65 kg m−2, respectively. Therefore, arid and semi-arid zones and shrub formations should be integrated into REDD+ strategies to promote their protection, given their potential to increase carbon reserves.

Enhancing deadwood reporting for forest ecosystems: Bridge equations to convert deadwood measured at any diameter threshold to reference diameters

National as well as international requirements have led to an increased need to quantify deadwood stocks in forest ecosystems given their important role not only in terms of carbon storage and regulation of the carbon cycle but also as biodiversity refugia. However, differences in definitions and field monitoring as well as gaps in existing data on deadwood mean that comparisons among countries and retrospective analyses are difficult. In this research, we propose two potential approaches to solve the most common gaps in forest deadwood monitoring. First, we develop bridging functions capable of converting deadwood measurements with a specific reference diameter to 7.5 cm (minimum diameter value in Spain) and 10.0 cm (the most common minimum value for international statistics) diameters for the main forest types while also addressing the effect of raising the minimum measurable size on the quantification of deadwood. Furthermore, we aim to calculate the ratios between the amount of standing deadwood, the most common indicator monitored in National Forest Inventories, and the entire deadwood pool as a proxy for estimating complete deadwood stocks when data are not available. For this objective, we use information obtained from the Spanish National Forest Inventory, linear models and 10-fold cross-validation. We estimate the percentage of deadwood omitted when the minimum deadwood size is increased for the main eight forest types in Spain as well as for the entire country, using two different approaches. The ratio between the amount of standing deadwood and the entire deadwood pool ranged between 0.14 and 0.45 depending on the forest type. The lowest values of this ratio were found in Open woodlands and the largest in Mediterranean conifers. The validation statistics (R2 ranging from 0.82 in Evergreen broadleaves to 0.97 in Macaronesian broadleaves) indicate that the bridging functions we propose are robust and accurate. However, the ratios between the amount of standing deadwood and the entire deadwood pool performed poorer (R2 ranging from 0.26 in Macaronesian conifers to 0.65 in Macaronesian broadleaves) and led to an overestimation of the total stocks. Our results are of value not only for the purposes of comparison and harmonization but also for the implementation of new forest monitoring systems.

Deep Learning-Based Forest Fire Risk Research on Monitoring and Early Warning Algorithms

With the development of image processing technology and video analysis technology, forest fire monitoring technology based on video recognition is more and more important in the field of forest fire prevention and control. The objects currently applied to forest fire video image monitoring system monitoring are mainly flames and smoke. This paper proposes a forest fire risk monitoring and early warning algorithm, which integrates a deep learning model, infrared monitoring and early warning, and forest fire weather index. The algorithm first obtains the current visible image and infrared image of the same forest area, utilizing a smoke detection model based on deep learning to detect smoke in the visible image, and obtains the confidence level of the occurrence of fire in said visible image. Then, it determines whether the local temperature value of said infrared image exceeds a preset warning value, and obtains a judgment result based on the infrared image. It calculates again a current FWI based on environmental data, and determines a current fire danger level based on the current FWI. Finally, it determines whether or not to carry out a fire warning based on said fire danger level, said confidence level of the occurrence of fire in said visible image, and said judgment result based on the infrared image. The experimental results show that the accuracy of the algorithm in this paper reaches 94.12%, precision is 96.1%, recall is 93.67, and F1-score is 94.87. The algorithm in this paper can improve the accuracy of smoke identification at the early stage of forest fire danger occurrence, especially by excluding the interference caused by clouds, fog, dust, and so on, thus improving the fire danger warning accuracy.

GIS and Artificial Intelligence Application in Smart Forest Ecosystem Sustainability Evaluation of Olokemeji Forest Reserve, Ogun State, Nigeria

The increase in human population over the years has accelerated growth in anthropogenic activities, which have led to the conversion of forest reserves to other land uses. In the sequel, it has become imperative for researchers to focus on the mapping of forest reserves through the use of GIS and Artificial Intelligence (AI) with time-efficient, automated, and low-cost methods to preserve the existing forest reserve and its sustainability evaluation implementation. This research aimed at utilizing GIS and artificial intelligence applications in smart forest ecosystem sustainability evaluation of Olokemeji forest reserve, Ogun State, with the following objectives: (i.) assessment of the current state of the forest ecosystem.(ii.) identify potential threats and risks to the study area and (iii.) develop sustainable management strategies for its conservation and preservation. In pursuance of this, GIS and AI were deployed in this study to assess the spatial characteristics of the forest ecosystem in Olokemeji forest reserve. Landsat imagery, ground coordinates, and a research questionnaire were the major data used. Object-based classification, Normalized Difference Vegetation Index (NDVI), and Land Use Land Cover in ArcGIS 10.2 software was deployed in data generation and analysis. The results showed that in 2013, about 1657.115 ha of the study area was occupied by dense forest cover while in 2023, it decreased to 1188.060 ha, with a difference of about 469.055 ha. By implementing smart forest monitoring and evaluation systems that use artificial intelligence, the government and commercial groups should set regulations focused on reducing the escalating risks to forest reserves.

IoT Based Early Forest Fire Detection and Monitoring System

IoT Based Early Forest Fire Detection and Monitoring System

Low-Cost Non-Contact Forest Inventory: A Case Study of Kieni Forest in Kenya

Forests are a vital source of food, fuel, and medicine and play a crucial role in climate change mitigation. Strategic and policy decisions on forest management and conservation require accurate and up-to-date information on available forest resources. Forest inventory data such as tree parameters, heights, and crown diameters must be collected and analysed to monitor forests effectively. Traditional manual techniques are slow and labour-intensive, requiring additional personnel, while existing non-contact methods are costly, computationally intensive, or less accurate. Kenya plans to increase its forest cover to 30% by 2032 and establish a national forest monitoring system. Building capacity in forest monitoring through innovative field data collection technologies is encouraged to match the pace of increase in forest cover. This study explored the applicability of low-cost, non-contact tree inventory based on stereoscopic photogrammetry in a recently reforested stand in Kieni Forest, Kenya. A custom-built stereo camera was used to capture images of 251 trees in the study area from which the tree heights and crown diameters were successfully extracted quickly and with high accuracy. The results imply that stereoscopic photogrammetry is an accurate and reliable method that can support the national forest monitoring system and REDD+ implementation.

Development of a deep learning-based surveillance system for forest fire detection and monitoring using UAV.

This study presents a surveillance system developed for early detection of forest fires. Deep learning is utilized for aerial detection of fires using images obtained from a camera mounted on a designed four-rotor Unmanned Aerial Vehicle (UAV). The object detection performance of YOLOv8 and YOLOv5 was examined for identifying forest fires, and a CNN-RCNN network was constructed to classify images as containing fire or not. Additionally, this classification approach was compared with the YOLOv8 classification. Onboard NVIDIA Jetson Nano, an embedded artificial intelligence computer, is used as hardware for real-time forest fire detection. Also, a ground station interface was developed to receive and display fire-related data. Thus, access to fire images and coordinate information was provided for targeted intervention in case of a fire. The UAV autonomously monitored the designated area and captured images continuously. Embedded deep learning algorithms on the Nano board enable the UAV to detect forest fires within its operational area. The detection methods produced the following results: 96% accuracy for YOLOv8 classification, 89% accuracy for YOLOv8n object detection, 96% accuracy for CNN-RCNN classification, and 89% accuracy for YOLOv5n object detection.

An Optimized Smoke Segmentation Method for Forest and Grassland Fire Based on the UNet Framework

Smoke, a byproduct of forest and grassland combustion, holds the key to precise and rapid identification—an essential breakthrough in early wildfire detection, critical for forest and grassland fire monitoring and early warning. To address the scarcity of middle–high-resolution satellite datasets for forest and grassland fire smoke, and the associated challenges in identifying smoke, the CAF_SmokeSEG dataset was constructed for smoke segmentation. The dataset was created based on GF-6 WFV smoke images of forest and grassland fire globally from 2019 to 2022. Then, an optimized segmentation algorithm, GFUNet, was proposed based on the UNet framework. Through comprehensive analysis, including method comparison, module ablation, band combination, and data transferability experiments, this study revealed that GF-6 WFV data effectively represent information related to forest and grassland fire smoke. The CAF_SmokeSEG dataset was found to be valuable for pixel-level smoke segmentation tasks. GFUNet exhibited robust smoke feature learning capability and segmentation stability. It demonstrated clear smoke area delineation, significantly outperforming UNet and other optimized methods, with an F1-Score and Jaccard coefficient of 85.50% and 75.76%, respectively. Additionally, augmenting the common spectral bands with additional bands improved the smoke segmentation accuracy, particularly shorter-wavelength bands like the coastal blue band, outperforming longer-wavelength bands such as the red-edge band. GFUNet was trained on the combination of red, green, blue, and NIR bands from common multispectral sensors. The method showed promising transferability and enabled the segmentation of smoke areas in GF-1 WFV and HJ-2A/B CCD images with comparable spatial resolution and similar bands. The integration of high spatiotemporal multispectral data like GF-6 WFV with the advanced information extraction capabilities of deep learning algorithms effectively meets the practical needs for pixel-level identification of smoke areas in forest and grassland fire scenarios. It shows promise in improving and optimizing existing forest and grassland fire monitoring systems, providing valuable decision-making support for fire monitoring and early warning systems.

Digital Opportunity or a Threat? Adoption of Internet of Things (IoT) Monitoring Systems for Natural Resources in Germany

In Germany, modern digital technologies like Internet of Things (IoT) have been increasingly promoted for better environmental and natural resources management through “smart” monitoring and real-time data. However, adopting such advanced technologies is complex and brings a wide array of risks and challenges, and it remains unclear whether local actors are in fact willing and sufficiently equipped to adopt them. Using the Motivations and Abilities (MOTA) framework, and quantitative data collection and analysis methods, this study explores the motivations and abilities of German local (governmental) actors by focusing on the adoption of IoT-based forest and water monitoring systems. The findings reveal an early-stage adoption of IoT environmental monitoring, with limited awareness and no plans for adoption. The lack of willingness, however, is not attributed to a lack of motivation; it is, rather, influenced by perceived insufficient financial and technical capacities and resources. This study provides novel insights for understanding the complex relationship between actors’ behavior and the adoption of advanced digital technologies in the realm of environmental and natural resources management. The results provide a robust foundation for future research, and inform policy and practice aimed at facilitating digitalized natural resources management.

Improving the organization of communication in the forest fire monitoring system

The article discusses the issues of early detection of forest fires. Currently, traditional monitoring methods are gradually being replaced by modern technologies based on automation and the use of artificial intelligence. The system of remote video monitoring “Lesohranitel”, which has been widely used in many regions of Russia, is considered in detail. The main advantages of this system and the basic principle of operation are considered. The scheme of communication organization of the considered system is presented, the main ways for its improvement are defined.

An Internet of Things Platform for Forest Monitoring

Forests have a very important function in sustaining the natural life on our planet. However, wildfires, landslides, uncontrolled tree cutting, poaching, arson, and many other dangers threaten the forests and natural resources. Therefore, effective monitoring and observation of forests is crucial. This study explains the development and implementation of an Internet of Things (IoT)-enabled forest monitoring system as an innovative solution that will contribute to the protection of forests. The presented system provides real-time climate data in forestlands by using microcontrollers, low-cost sensors, data communication, and cloud platforms. It collects important information such as temperature, humidity, air quality, and ecological activities. Fire detection is achieved by associating the increase in CO gas concentration with the increase in temperature. Landslides are detected by measuring the acceleration of soil movement in 3 axes. Additionally, the system includes advanced machine learning-based acoustic tracking techniques to detect chainsaws, motor vehicles, screams, shouts, and gunshots. The IoT platform provides a web-based user interface and other tools to system users such as forest managers and researchers. These tools detect early signs of threats such as wildfires, landslides and illegal activities in forests. Our tests demonstrate the system's effectiveness in providing information for protecting and managing forests.

Prototype of Forest and Land Fire Monitoring and Detection System Using IoT-Based WSN Technology

Forest and land fires are a recurrent issue in several Indonesian regions, necessitating advanced technological solutions for early detection and monitoring due to their significant impact. This research project focuses on developing a prototype system utilizing IoT-based Wireless Sensor Network (WSN) technology to detect forest and land fires. The prototype offers real-time remote monitoring of temperature, humidity, smoke levels, and wind speed in the forest through an Android app or web interface. To validate its performance, the prototype was compared to existing BMKG tools, with results showing minor temperature errors (2.41%) and a humidity error of 17.68%. The anemometer sensor exhibited a slight 4-second data transmission delay. Importantly, the prototype excelled in fire detection, effectively identifying temperature and humidity changes within a 2-meter radius or more indicative of fire outbreaks. This pioneering prototype promises to significantly enhance early warning and response mechanisms for forest and land fires in Indonesia, contributing to more effective environmental conservation and disaster management.

Accuracy of the Copernicus High-Resolution Layer Forest Type (HRL FTY) assessed with domestic NFI sampling plots in Poland

Abstract Over the past years, several remote sensing maps of land cover have been produced, but they still exhibit certain differences compared to the real land use that reduce their value for climate and carbon cycle modelling as well as for national estimates of forest carbon stocks and their change. This paper outlines a straightforward framework for evaluating map accuracy and estimating uncertainty in land cover area, specifically for forest-related land cover maps in Poland for the year 2018. The study compares stratified field-based data from the National Forest Inventory (NFI) with remote sensing data on forest variables, at the pixel level, in order to identify suitable methods for accuracy and area uncertainty estimation. Additionally, the paper introduces and presents a variety of accuracy metrics applicable to assess overall uncertainties in GHG inventories. The results indicate that the High-Resolution Layer Forest Type (HRL FTY) product (part of the broader Copernicus Land Monitoring Service [CLMS] portfolio), assessed using NFI field-based information, achieved an overall accuracy (OA) of 69.2%. This metric varies among particular nature protection forms, with the highest observed ones in Natura 2000 sites of 70.45%. The primary source of map errors was associated with distinguishing between broad-leaved and coniferous forest areas. Improving future maps necessitates more precise differentiation between species to better support national forest monitoring systems for the purpose of greenhouse gas (GHG) inventories where information on the spatial distribution and variability of forests sources, biodiversity assessment, threat prevention, estimation of carbon content is becoming an important part of the associated reporting system.

Forest Monitoring and Fire Detection System

Forest fires/Wildfires are the most common disaster in forests which lead to severe destruction of forest wealth, natural habitat and bio-diversity. Our concept is to create a system which will help in early detection of forest fire by measuring natural factors which lead to forest fire. Our Forest Monitoring and Fire Detection System uses various sensors like heat, smoke, wind, temperature and humidity to predict and detect forest fire on an early basis. The sensors are connected to Esp8266 wi-fi module which after detecting sends data to cloud. The cloud then sends the data to the website. The website displays an alert so that immediate action can be taken. Early detection and preventive measures are necessary to protect forests from fires which will be done using this system.

Comparing ground below-canopy and satellite spectral data for an improved and integrated forest phenology monitoring system

Phenology monitoring allows a better understanding of forest functioning and climate impacts. Satellite indicators are used to upscale ground phenological observations, but often differential responses are observed, and data availability can be limited. In view of climate impacts, new tools capable to detect rapid phenological changes and to work at single species level are needed. This research compares indices derived by the TreeTalker© (TT + ) below canopy upward-looking spectral data and Sentinel 2 satellite data, used to assess the phenological behavior and changepoints in several European beech forests. Overall, a mismatch between the information derived by the two sensor types is evidenced, with main differences in: start/end and length of season and phenology changepoints; larger variability captured by TT + with respect to Sentinel 2 especially in the leaf on period; mixed signal response from multiple vegetation layers in Sentinel 2 data. The complementarity of satellite and TT + indices allow exploring the phenological responses from different vegetation layers. TT + higher temporal resolution demonstrates precision in capturing the phenological changepoints in beech forests, especially if satellite image availability is limited by cloud cover and leads to miss critical phenological dates. The best settings for TT + data collection and the advantages to have two spectral data sources for improved forest phenology monitoring are also commented. The TT+, collecting additional tree parameters, can be a valuable tool for an integrated monitoring system based on spectral signals from above and below the canopy, at high temporal frequency and high spatial resolution.

Evaluation of Tree-Growth Rate in the Laurentides Wildlife Reserve Using GEDI and Airborne-LiDAR Data

Loss of forest cover and derived effects on forest ecosystems services has led to the establishment of land management policies and forest monitoring systems, and consequently to the demand for accurate and multitemporal data on forest extent and structure. In recent years, spaceborne Light Detection and Ranging (LiDAR) missions, such as the Global Ecosystem Dynamics Investigation (GEDI) instrument, have facilitated the repeated acquisition of data on the vertical structure of vegetation. In this study, we designed an approach incorporating GEDI and airborne LiDAR data, in addition to detailed forestry inventory data, for estimating tree-growth dynamics for the Laurentides wildlife reserve in Canada. We estimated an average tree-growth rate of 0.32 ± 0.23 (SD) m/year for the study site and evaluated our results against field data and a time series of NDVI from Landsat images. The results are in agreement with expected patterns in tree-growth rates related to tree species and forest stand age, and the produced dataset is able to track disturbance events resulting in the loss of canopy height. Our study demonstrates the benefits of using spaceborne-LiDAR data for extending the temporal coverage of forestry inventories and highlights the ability of GEDI data for detecting changes in forests’ vertical structure.

Remote Sensing Based Deforestation Monitoring at the Change Hotspot Area in Hindu Kush Himalayan Region

Forests play an essential role in providing sustainable ecosystem services and livelihood options for a growing population. In the Hindu Kush Himalayan (HKH) region, forests have been continuously reduced due to increasing demand for timber, fuelwood, and agriculture. Identification of deforestation hotspots and monitoring changes in those hotspots will be highly useful for forest managers to prevent illegal deforestation. In this paper, we identified forest cover change hotspots and areas for annual monitoring in those hotspots. For that factors like land cover from 1990 and 2010, shuttle radar topography mission (SRTM) digitalelevation model (DEM), slope, curvature, and distance from the settlement and roads for all four countries which could influence loss in forest area and overplayed to determine forest hotspots, were considered. Land cover maps of 1990 and 2010 and other GIS layers were used for identification of hotspots using the model builder ArcGIS software. For monitoring of deforestation in the hotspot areas, Landsat 8 images (2013, 2014 and 2015) and geographic object-based image analysis (GEOBIA) technique was used. The method was validated in four study sites in Bangladesh, Nepal, Bhutan and Pakistan. The study revealed that the sites in Bangladesh have higher deforestation during 2013-2014 and 2014-2015 with forest loss of 1121.58 and 1773.18 ha respectively. The web-based forest monitoring system provides information on deforestation useful for forest managers to enforce annual management plans.

Classifying drivers of deforestation by using the deep learning based poly-highway forest convolution network

The worldwide deforestation rate worsens year after year, ultimately resulting in a variety of severe implications for both mankind and the environment. In order to track the success of forest preservation activities, it is crucial to establish a reliable forest monitoring system. Changes in forest status are extremely difficult to manually annotate due to the tiny size and subtlety of the borders involved, particularly in regions abutting residential areas. Previous forest monitoring systems failed because they relied on low-resolution satellite images and drone-based data, both of which have inherent limitations. Most government organizations still use manual annotation, which is a slow, laborious, and costly way to keep tabs on data. The purpose of this research is to find a solution to these problems by building a poly-highway forest convolution network using deep learning to automatically detect forest borders so that changes over time may be monitored. Here initially the data was curated using the dynamic decomposed kalman filter. Then the data can be augmented. Afterward the augmented image features can be fused using the multimodal discriminant centroid feature clustering. Then the selected area can be segmented using the iterative initial seeded algorithm (IISA). Finally, the level and the driver of deforestation can be classified using the poly-highway forest convolution network (PHFCN). The whole experimentation was carried out in a dataset of 6048 Landsat-8 satellite sub-images under MATLAB environment. From the result obtained the suggested methodology express satisfied performance than other existing mechanisms.

A Survey on Deep Learning Based Forest Environment Sound Classification at the Edge

Forest ecosystems are of paramount importance to the sustainable existence of life on earth. Unique natural and artificial phenomena pose severe threats to the perseverance of such ecosystems. With the advancement of artificial intelligence technologies, the effectiveness of implementing forest monitoring systems based on acoustic surveillance has been established due to the practicality of the approach. It can be identified that with the support of transfer learning, deep learning algorithms outperform conventional machine learning algorithms for forest acoustic classification. Further, a clear requirement to move the conventional cloud-based sound classification to the edge is raised among the research community to ensure real-time identification of acoustic incidents. This article presents a comprehensive survey on the state-of-the-art forest sound classification approaches, publicly available datasets for forest acoustics, and the associated infrastructure. Further, we discuss the open challenges and future research aspects that govern forest acoustic classification.