A forest fire danger grade dynamic evaluation and early warning method based on multi-source data fusion

Abstract

The application provides a forest fire risk grade dynamic evaluation and early warning method based on multi-source data fusion, which obtains forest image monitoring data, judges the fire risk based on the forest image monitoring data, obtains multi-source monitoring data in a target forest area, processes the multi-source monitoring data in time and space, determines multi-dimensional fire risk characteristic parameters, analyzes the risk according to the multi-dimensional fire risk characteristic parameters, determines the fire risk grade corresponding to the target forest area, obtains an infrared detection image of fire smoke, constructs a smoke plume vertical temperature distribution model, identifies the fire risk spread intensity according to the smoke plume vertical temperature distribution model, identifies the diffusion risk according to the fire risk spread intensity and the multi-source monitoring, and obtains the fire risk diffusion grade; and generates forest fire risk early warning information according to the corresponding fire risk grade and the fire risk diffusion grade, can dynamically adjust the fire risk grade according to the smoke plume vertical temperature distribution, and improves the accuracy of the fire risk grade evaluation.

Description

Technical Field

[0001] This application relates to the field of fire risk monitoring technology, and more specifically, to a method for dynamic assessment and early warning of forest fire risk levels based on multi-source data fusion. Background Technology

[0002] Forest fires, as a highly destructive natural disaster, pose a serious threat to the ecological environment and forest resources. Furthermore, influenced by global climate change and frequent extreme weather events, the frequency and extent of forest fires are showing a continuous increasing trend. Forest fire risk identification and early warning are crucial for reducing disaster losses. Constructing a scientific and efficient forest fire risk early warning system can not only enable the early identification of potential fire risks but also allow for timely adjustments to warning levels based on the fire's development, providing effective data for fire prevention resource allocation and fire suppression decisions.

[0003] Existing forest fire risk early warning methods typically construct risk assessment models based on factors such as meteorological conditions, vegetation status, and surface environment. They analyze changes in environmental factors to determine the likelihood of forest fires. However, these methods mainly focus on risk prediction before a fire occurs and lack effective perception of the heat release status, fire development trend, and spatial diffusion capacity after a fire has occurred. This makes it difficult to dynamically adjust the fire risk level. Therefore, how to combine multi-source monitoring information before and after a fire to dynamically assess the forest fire risk level has become an urgent technical problem to be solved. Summary of the Invention

[0004] This application provides a method for dynamic assessment and early warning of forest fire risk levels based on multi-source data fusion, which can dynamically adjust the fire risk level according to the vertical temperature distribution of the smoke plume, thereby improving the accuracy of fire risk level assessment.

[0005] In the first aspect, this application provides a method for dynamic assessment and early warning of forest fire risk level based on multi-source data fusion. This method can be executed by a network device, or it can be executed by a chip configured in the network device. This application does not limit the method in this regard.

[0006] Specifically, the method includes:

[0007] Acquire forest image monitoring data and make fire risk assessment based on the forest image monitoring data;

[0008] When it is determined that there is no fire risk, multi-source monitoring data within the target forest area is acquired, and the multi-source monitoring data is subjected to spatiotemporal correlation processing to determine multi-dimensional fire risk characteristic parameters.

[0009] Risk analysis is conducted based on the aforementioned multidimensional fire risk characteristic parameters to determine the fire risk level corresponding to the target forest area;

[0010] When a fire hazard exists, infrared detection images of fire smoke are acquired, and a vertical temperature distribution model of the smoke plume is constructed. The intensity of fire spread is identified based on the vertical temperature distribution model of the smoke plume. Based on the intensity of fire spread and the multi-source monitoring data, the spread risk is identified, and the fire spread level is obtained.

[0011] Forest fire risk warning information is generated based on the corresponding fire risk level and fire spread level.

[0012] In conjunction with the first aspect, in certain implementations of the first aspect, determining fire risk based on the forest image monitoring data specifically includes:

[0013] The forest image monitoring data is divided into target areas to obtain multiple spatial monitoring units;

[0014] For any spatial monitoring unit, abnormal regions are identified by pixel thresholds, and abnormal feature parameters of the abnormal regions are extracted.

[0015] Fire hazard identification is performed based on the aforementioned abnormal characteristic parameters to determine whether there is a fire hazard in the space monitoring unit.

[0016] In conjunction with the first aspect, in some implementations of the first aspect, the spatiotemporal correlation processing of the multi-source monitoring data specifically includes: dividing the target forest area into multiple spatial monitoring units, obtaining the spatial coordinate information corresponding to each monitoring data, mapping the multi-source monitoring data to the corresponding spatial monitoring units; establishing a unified time window, and performing time synchronization processing on the multi-source monitoring data collected within a preset time range.

[0017] In conjunction with the first aspect, in certain implementations of the first aspect, risk analysis based on the aforementioned multidimensional fire risk characteristic parameters to determine the fire risk level corresponding to the target forest area specifically includes:

[0018] Based on the multidimensional fire hazard feature parameters, feature association judgment is performed to generate dynamic feature weights between different fire hazard feature parameters.

[0019] The fire risk level of the target forest area is obtained by fusing and calculating the multidimensional fire risk feature parameters based on the dynamic feature weights.

[0020] In conjunction with the first aspect, in certain implementations of the first aspect, acquiring infrared detection images of fire smoke and constructing a vertical temperature distribution model of the smoke plume specifically includes:

[0021] Acquire a sequence of infrared detection images of fire smoke. For any infrared detection image, perform spatial correction on the infrared detection image according to the device imaging parameters and the preset spatial position relationship to determine the spatial position of each pixel in the infrared detection image.

[0022] The infrared detection image is used to identify thermal anomaly regions to determine the undetermined plume region;

[0023] Identify the undetermined smoke plume region in continuous infrared images, identify the temperature diffusion characteristics of the undetermined smoke plume region through continuous infrared images, extract the temperature diffusion characteristics of the undetermined smoke plume region for rule determination, and determine the fire smoke plume region.

[0024] A height-stratified model is established based on the spatial distribution characteristics of the fire plume region, and a vertical temperature distribution model of the plume is constructed based on the plume temperature characteristics corresponding to different height layers.

[0025] In conjunction with the first aspect, in certain implementations of the first aspect, identifying the intensity of fire spread based on the vertical temperature distribution model of the smoke plume specifically includes:

[0026] Based on the plume vertical temperature distribution model, the plume temperature distribution state corresponding to different time periods is analyzed, and the plume thermal diffusion feature vector is extracted.

[0027] A smoke plume identification model is constructed, and the smoke plume thermal diffusion feature vector is clustered based on the smoke plume identification model. The corresponding fire hazard spread intensity is determined based on the clustering results.

[0028] In conjunction with the first aspect, in certain implementations of the first aspect, generating forest fire risk warning information based on the corresponding fire risk level and fire spread level specifically includes:

[0029] Obtain the fire risk level and fire spread level corresponding to the target forest area;

[0030] When the fire risk level is higher than the preset risk level threshold, risk prevention and early warning information is generated as forest fire risk early warning information.

[0031] When the fire risk spread level is higher than the preset spread level threshold, a high-level fire emergency warning message is generated as a forest fire risk warning message.

[0032] In conjunction with the first aspect, in some implementations of the first aspect, after generating forest fire risk warning information based on the corresponding fire risk level and fire spread level, the method further includes: sending the forest fire risk warning information to a fire risk remote monitoring platform.

[0033] Secondly, this application provides a computer terminal device, which includes a memory and a processor. The memory stores code, and the processor is configured to acquire the code and execute the above-described method for dynamic assessment and early warning of forest fire risk levels based on multi-source data fusion.

[0034] Thirdly, this application provides a computer-readable storage medium storing at least one computer program, which is loaded and executed by a processor to perform the operations described above in the method for dynamic assessment and early warning of forest fire risk levels based on multi-source data fusion.

[0035] The technical solutions provided by the embodiments disclosed in this application have the following beneficial effects:

[0036] This application provides a method for dynamic assessment and early warning of forest fire risk levels based on multi-source data fusion. The method involves acquiring forest image monitoring data and determining the fire risk based on this data. When no fire risk is detected, multi-source monitoring data within the target forest area is acquired, and spatiotemporal correlation processing is performed on the multi-source monitoring data to determine multi-dimensional fire risk characteristic parameters. Risk analysis is conducted based on these multi-dimensional fire risk characteristic parameters to determine the corresponding fire risk level for the target forest area. When a fire risk exists, infrared detection images of fire smoke are acquired, and a vertical temperature distribution model of the smoke plume is constructed. The fire spread intensity is identified based on the smoke plume vertical temperature distribution model, and diffusion risk is identified based on the fire spread intensity and multi-source monitoring data to obtain the fire diffusion level. Forest fire risk early warning information is generated based on the corresponding fire risk level and fire diffusion level.

[0037] Therefore, this application considers that under fire hazard conditions, high-temperature smoke forms a plume with spatial diffusion characteristics. The vertical temperature distribution of the plume is directly related to the fire's heat release capacity, combustion intensity, and energy diffusion range. By integrating pre-fire environmental risk information with the thermodynamic characteristics of the plume after the fire, a phased dynamic assessment of forest fire risk status is achieved. Before a fire occurs, the potential fire risk level is determined through multi-source monitoring data. After a fire occurs, the fire risk diffusion level is adjusted in real time by the changes in the vertical temperature distribution of the plume, so that the early warning results can dynamically change with the development of the fire. Compared with the traditional assessment method that relies solely on static factors such as meteorology and vegetation, this improves the accuracy and real-time performance of forest fire risk level judgment.

[0038] In summary, this application enables dynamic adjustment of fire hazard levels based on the vertical temperature distribution of the smoke plume, thereby improving the accuracy of fire hazard assessment. Attached Figure Description

[0039] Figure 1 This is an exemplary flowchart illustrating a method for dynamic assessment and early warning of forest fire risk levels based on multi-source data fusion, according to some embodiments of this application.

[0040] Figure 2 This is a schematic diagram of the structure of a monitoring and evaluation system based on multi-source data fusion provided in an embodiment of this application;

[0041] Figure 3 This is a structural schematic diagram of a fire warning device according to some embodiments of this application;

[0042] Figure 4 This is a schematic diagram of the structure of a computer terminal device that implements a method for dynamic assessment and early warning of forest fire risk levels based on multi-source data fusion, according to some embodiments of this application. Detailed Implementation

[0043] This application acquires forest image monitoring data and uses this data to determine fire risk. When no fire risk is detected, it acquires multi-source monitoring data within the target forest area and performs spatiotemporal correlation processing on the multi-source monitoring data to determine multi-dimensional fire risk characteristic parameters. Based on these multi-dimensional fire risk characteristic parameters, it conducts risk analysis to determine the corresponding fire risk level of the target forest area. When a fire risk exists, it acquires infrared detection images of fire smoke and constructs a vertical temperature distribution model of the smoke plume. Based on this model, it identifies the intensity of fire spread and identifies the diffusion risk based on the intensity of fire spread and multi-source monitoring to obtain the fire diffusion level. Based on the corresponding fire risk level and fire diffusion level, it generates forest fire risk warning information. This system can dynamically adjust the fire risk level according to the vertical temperature distribution of the smoke plume, improving the accuracy of fire risk level assessment.

[0044] To better understand the above technical solutions, a detailed explanation will be provided below with reference to the accompanying drawings and specific implementation methods. (Reference) Figure 1 The figure is an exemplary flowchart of a method for dynamic assessment and early warning of forest fire risk levels based on multi-source data fusion, according to some embodiments of this application. The method 100 for dynamic assessment and early warning of forest fire risk levels based on multi-source data fusion mainly includes the following steps:

[0045] In step S101, forest image monitoring data is acquired, and fire risk is determined based on the forest image monitoring data.

[0046] Optionally, in some embodiments, the forest image monitoring data can be obtained by visible light camera equipment, drone inspection equipment or other image acquisition equipment within the target forest area, and the forest image monitoring data includes forest area image frames in a continuous time series.

[0047] In a specific implementation, before determining the fire risk based on the forest image monitoring data, the method further includes: performing image preprocessing on the forest image monitoring data, wherein the image preprocessing includes denoising, brightness correction, and image size normalization on multiple image frames in the forest image monitoring data.

[0048] Preferably, in some embodiments, determining the fire risk based on the forest image monitoring data specifically includes:

[0049] The forest image monitoring data is divided into target areas to obtain multiple spatial monitoring units;

[0050] For any spatial monitoring unit, abnormal regions are identified by pixel thresholds, and abnormal feature parameters of the abnormal regions are extracted.

[0051] Fire hazard identification is performed based on the aforementioned abnormal characteristic parameters to determine whether there is a fire hazard in the space monitoring unit.

[0052] In specific implementation, for any spatial monitoring unit, the image pixel set corresponding to that spatial monitoring unit is obtained, and pixel feature analysis is performed on the image pixel set. Optionally, in some embodiments, based on the relatively stable color distribution and brightness variation law of the forest background area under normal conditions, statistical analysis is performed on the gray value, color component, and brightness information corresponding to each pixel to construct the pixel feature range corresponding to the normal forest background. For example, a background pixel model is established based on historically collected normal forest environment images. For any pixel in the spatial monitoring unit, its corresponding gray value, color value, and brightness value are obtained, and the variation range of the corresponding pixel feature value within multiple consecutive time periods is statistically analyzed. The variation range is taken as the normal state pixel feature range corresponding to that pixel. Then, the pixel difference between the spatial monitoring unit image corresponding to the current collection time and the background image under normal conditions is calculated. For any pixel, the degree of difference between its current pixel feature value and the background pixel feature value is calculated. When the degree of difference exceeds a preset abnormal threshold, the corresponding pixel is marked as an abnormal pixel.

[0053] Connectivity analysis is performed on multiple consecutive abnormal pixels, and adjacent abnormal pixels are aggregated. Regions are divided based on the grayscale distribution range of the abnormal pixels to obtain the flame and smoke regions to be detected. For the smoke region, the area ratio, texture features, grayscale uniformity, and continuous diffusion rate are extracted as feature parameters. For the flame region, the proportion of bright pixels, color change amplitude, area growth rate, and duration are extracted as feature parameters. Based on a preset rule-based judgment model, the feature parameters of the region to be detected are judged to identify whether there is a fire hazard in the spatial monitoring unit. For example, when the region to be detected simultaneously meets the conditions of continuous smoke presence, continuous expansion of smoke area, and increased brightness of flame region, the corresponding spatial monitoring unit is determined to have a fire hazard anomaly. When only short-term smoke changes or local brightness changes are detected, and the continuous diffusion condition is not met, the corresponding spatial monitoring unit is determined not to have a fire hazard anomaly.

[0054] Optionally, in some other embodiments, when there are many misjudgments caused by environmental factors in a single moment image, time-series correlation analysis can be performed on forest image monitoring data over multiple consecutive time periods. By comparing the image changes between adjacent time frames, the dynamic change characteristics of abnormal areas can be determined. When it is detected that an area continuously exhibits smoke characteristics and the smoke area expands over time, or when there is a bright flame area with obvious dynamic change characteristics, it is considered that there is a fire hazard in the area. This application will not elaborate further on this.

[0055] In step S102, when it is determined that there is no fire hazard, multi-source monitoring data within the target forest area is acquired, and the multi-source monitoring data is subjected to spatiotemporal correlation processing to determine multi-dimensional fire hazard characteristic parameters.

[0056] Optionally, in some embodiments, multi-source monitoring data corresponding to the target forest area is acquired. This multi-source monitoring data includes meteorological monitoring data, vegetation status monitoring data, soil environment monitoring data, and remote sensing monitoring data. Specifically, the meteorological monitoring data includes parameters such as ambient temperature, air humidity, wind speed, wind direction, and rainfall, used to characterize the aridity of the target forest area; the vegetation status monitoring data includes parameters such as vegetation coverage, vegetation index, and vegetation water content, used to characterize the distribution of combustibles within the forest area; the soil environment monitoring data includes parameters such as soil moisture content, soil temperature, and soil moisture variation trends, used to characterize the aridity of surface combustibles; and the remote sensing monitoring data includes parameters such as surface temperature, vegetation remote sensing index, and historical anomalous heat source information, used to supplement environmental status information over a large area.

[0057] In practice, due to differences in sampling time, spatial resolution, and data representation formats among monitoring devices from different sources, spatiotemporal correlation processing is performed on the multi-source monitoring data. Specifically, firstly, a unified spatial grid is established based on the geographic spatial range of the target forest area, dividing the target forest area into multiple spatial monitoring units. The spatial coordinate information corresponding to each monitoring data point is then acquired, and monitoring data from different sources are mapped to their corresponding spatial monitoring units. Simultaneously, a unified time window is established to address the differences in the acquisition time intervals of different monitoring data, and time synchronization processing is performed on different types of monitoring data collected within a preset time range. For example, real-time environmental parameters collected by meteorological sensors, surface state parameters periodically collected by remote sensing equipment, and moisture parameters continuously collected by soil sensors are mapped to the same time window, ensuring temporal consistency of monitoring data from multiple sources corresponding to the same spatial monitoring unit.

[0058] In some embodiments, reference Figure 2 As shown, Figure 2This is a schematic diagram of a multi-source data fusion monitoring and evaluation system provided in this application embodiment. This application employs a multi-source data fusion monitoring and evaluation system for the acquisition and preprocessing of multi-source monitoring data. The multi-source data fusion monitoring and evaluation system includes: a visible light camera, a remote sensing satellite, a multi-source environmental monitoring station, an image preprocessing and segmentation module, a pixel feature analysis and anomaly identification module, a multi-source data spatiotemporal correlation module, and a multi-dimensional feature extraction module. Specifically, the visible light camera is used to acquire continuous image monitoring data corresponding to the target forest area and send the acquired forest image data to the data processing module; the remote sensing satellite is used to acquire large-scale remote sensing monitoring data corresponding to the target forest area, including surface temperature information, vegetation remote sensing index information, and historical abnormal heat source information; the multi-source environmental monitoring station is used to collect environmental state data corresponding to the target forest area, including ambient temperature, air humidity, wind speed, wind direction, rainfall, soil moisture content, and vegetation state parameters.

[0059] The pixel feature analysis and anomaly recognition module is connected to the image preprocessing and segmentation module. It is used to perform feature analysis on the image pixel information corresponding to each spatial monitoring unit, extract grayscale features, color features, and brightness change features, and identify abnormal areas in the forest area based on pixel differences. The multi-source data spatiotemporal correlation module is connected to the remote sensing satellite and the multi-source environmental monitoring station, respectively. It is used to perform time synchronization and spatial matching of data from different sources, with different sampling periods, and different spatial scales, and to establish multi-source correlation data corresponding to the target forest area. The multi-dimensional feature extraction module is connected to the pixel feature analysis and anomaly recognition module and the multi-source data spatiotemporal correlation module. It is used to fuse image anomaly features, remote sensing monitoring features, and environmental monitoring features to extract multi-dimensional fire risk feature parameters for forest fire risk assessment. The image preprocessing and segmentation module is connected to the visible light camera device. It is used to perform image denoising, brightness correction, and size normalization on the acquired forest image monitoring data, and to divide the data into multiple spatial monitoring units based on spatial regions.

[0060] In some embodiments, this application performs data standardization processing on multi-source monitoring data after spatiotemporal matching, transforming data of different dimensions into a unified feature space. Specifically, normalization processing is performed based on the variation range of each monitoring parameter to eliminate dimensional differences between different parameters such as temperature, humidity, vegetation index, and water content, resulting in standardized multi-source associated data. Simultaneously, feature extraction is performed based on the standardized multi-source associated data to construct multi-dimensional fire risk feature parameters corresponding to the target forest area. Specifically, based on the correlation between various monitoring parameters and the occurrence of forest fires, environmental dryness characteristics, vegetation combustibility characteristics, thermal anomaly characteristics, and diffusion condition characteristics are extracted to characterize the fire risk state, forming a set of multi-dimensional fire risk characteristic parameters. Among them, the environmental dryness characteristics can be calculated based on ambient temperature, air humidity, and rainfall, and are used to reflect the degree of water loss in forest areas; the vegetation combustibility characteristics can be determined based on vegetation index, vegetation coverage, and vegetation water content, and are used to reflect the accumulation of combustibles; the thermal anomaly characteristics can be determined based on changes in surface temperature and historical heat source information, and are used to reflect areas of abnormal temperature rise; the diffusion condition characteristics can be determined based on wind speed, wind direction, and topographical parameters, and are used to reflect the spread capability after a fire occurs.

[0061] In step S103, risk analysis is performed based on the multidimensional fire risk characteristic parameters to determine the fire risk level corresponding to the target forest area.

[0062] Optionally, in some embodiments, determining the fire risk level of the target forest area based on the multidimensional fire risk characteristic parameters specifically includes:

[0063] Based on the multidimensional fire hazard feature parameters, feature association judgment is performed to generate dynamic feature weights between different fire hazard feature parameters.

[0064] The fire risk level of the target forest area is obtained by fusing and calculating the multidimensional fire risk feature parameters based on the dynamic feature weights.

[0065] In practical implementation, the multidimensional fire hazard characteristic parameters are organized according to spatial monitoring units to construct a fire hazard characteristic matrix corresponding to the target forest area. Each row in the fire hazard characteristic matrix corresponds to a spatial monitoring unit, and each column corresponds to a fire hazard influencing factor, including environmental dryness characteristics, vegetation flammability characteristics, thermal anomaly characteristics, and fire spread condition characteristics. Based on the fire hazard characteristic matrix, feature correlation analysis is performed to determine the degree of correlation between different fire hazard characteristic parameters. Specifically, the correlation between any two fire hazard characteristic parameters is calculated, and the correlation strength between each fire hazard characteristic parameter is determined based on the synchronous changes of different fire hazard characteristic parameters during historical fire events.

[0066] For example, when the ambient temperature rises, the air humidity decreases, and the vegetation moisture content decreases simultaneously, it indicates that the dryness of the area is increasing and there is a strong correlation between multiple characteristic parameters. When there is a significant difference in the changing trend of a characteristic parameter from other parameters, the influence of that characteristic parameter in the risk assessment process is reduced. Furthermore, a fire risk characteristic correlation matrix is ​​constructed based on the degree of correlation of the characteristics, and the influence weight of each fire risk characteristic parameter is dynamically adjusted in combination with the changing trend of each fire risk characteristic parameter in the current monitoring period.

[0067] In some embodiments, the changes in fire hazard characteristic parameters and the corresponding fire risk results within multiple historical time periods are obtained. The contribution of different fire hazard characteristic parameters under different risk states is analyzed, and a characteristic parameter contribution relationship model is established. Based on the matching degree between the real-time fire hazard characteristic parameters corresponding to the current target forest area and the historical risk state, the weight values ​​of the corresponding fire hazard characteristic parameters are adjusted to generate dynamic characteristic weights. For example, when the target forest area has not experienced rainfall for a long time and the soil moisture content continues to decrease, the weight of the fire hazard characteristic parameter representing the degree of dryness is increased. When sustained high temperatures and increased wind speed are detected, the weights of the environmental temperature characteristic and diffusion condition characteristic are increased, so that the risk assessment results under different environmental conditions are more consistent with the actual fire hazard change patterns.

[0068] In specific implementation, the multi-dimensional fire risk feature parameters are fused and calculated based on the dynamic feature weights to obtain the fire risk level corresponding to the target forest area. Specifically, the multi-dimensional fire risk feature parameters corresponding to each spatial monitoring unit are weighted and fused with the corresponding dynamic feature weights to calculate the fire risk score. Further, according to preset risk level classification rules, the fire risk score is converted into the corresponding fire risk level. For example, when the fire risk score is in the low-risk range, the target forest area is determined to have a low fire risk level; when the fire risk score reaches the medium or high-risk range, the corresponding medium and high fire risk levels are determined respectively.

[0069] In step S104, when a fire hazard exists, an infrared detection image of the fire smoke is acquired, and a vertical temperature distribution model of the smoke plume is constructed. The intensity of fire spread is identified based on the vertical temperature distribution model of the smoke plume, and the diffusion risk is identified based on the intensity of fire spread and the multi-source monitoring data to obtain the fire spread level.

[0070] Optionally, in some embodiments, during the process of acquiring infrared detection images of fire smoke and constructing a vertical temperature distribution model of the smoke plume, the infrared detection images can be acquired by infrared cameras installed on forest lookout towers, drone platforms, or mobile inspection equipment. The infrared cameras generate corresponding infrared detection images by acquiring infrared radiation information corresponding to the fire area and the surrounding smoke plume area. In some embodiments, other devices or equipment capable of acquiring fire smoke temperature can also be used; this application does not limit this approach.

[0071] Preferably, in some embodiments, acquiring infrared detection images of fire smoke and constructing a vertical temperature distribution model of the smoke plume specifically includes:

[0072] Acquire a sequence of infrared detection images of fire smoke. For any infrared detection image, perform spatial correction on the infrared detection image according to the device imaging parameters and the preset spatial position relationship to determine the spatial position of each pixel in the infrared detection image.

[0073] The infrared detection image is used to identify thermal anomaly regions to determine the undetermined plume region;

[0074] Identify the undetermined smoke plume region in continuous infrared images, identify the temperature diffusion characteristics of the undetermined smoke plume region through continuous infrared images, extract the temperature diffusion characteristics of the undetermined smoke plume region for rule determination, and determine the fire smoke plume region.

[0075] A height-stratified model is established based on the spatial distribution characteristics of the fire plume region, and a vertical temperature distribution model of the plume is constructed based on the plume temperature characteristics corresponding to different height layers.

[0076] In its specific implementation, this application considers that during the actual deployment of infrared thermal imaging equipment, the installation height, shooting angle, focal length, and monitoring distance will affect the spatial correspondence of the target area in the infrared image. Therefore, spatial correction is required for the infrared detection image. For any given infrared detection image, based on the imaging parameters of the infrared thermal imaging equipment and the preset spatial relationship between the equipment and the fire area, coordinate mapping processing is performed on the infrared detection image to establish a transformation relationship between the infrared image pixel coordinate system and the actual forest spatial coordinate system. The imaging parameters include the equipment installation height, pitch angle, horizontal rotation angle, focal length, and image resolution; the spatial relationship includes the distance information and spatial orientation information between the infrared equipment and the target forest area.

[0077] In some embodiments, based on the coordinate transformation relationship, the spatial position corresponding to each pixel in the infrared detection image is determined, and the infrared radiation intensity corresponding to each pixel is converted into temperature information to obtain an infrared temperature distribution map containing spatial coordinates and temperature values. Then, based on a preset fire smoke feature recognition threshold, thermal anomaly regions are identified on the temperature distribution map corresponding to the infrared detection image to determine undetermined plume regions. For example, continuously distributed thermal anomaly pixels are aggregated, and spatially adjacent thermal anomaly pixels with similar temperature changes are grouped into the same anomaly region, resulting in multiple undetermined thermal anomaly regions. For each undetermined thermal anomaly region, corresponding temperature feature parameters are extracted, including the region's average temperature, maximum temperature, temperature gradient, and the proportion of high-temperature area. Regions that meet preset temperature conditions are determined as undetermined plume regions.

[0078] It should be noted that, to avoid misidentifying high-temperature areas on the ground, equipment heat sources, and other non-fire heat sources as smoke plume areas, this application improves the robustness of fire smoke plume area identification by performing temporal correlation analysis on the undetermined smoke plume areas in continuous infrared images. Specifically, temperature diffusion characteristics of the undetermined smoke plume areas are identified through continuous infrared images, and these characteristics are extracted for rule-based determination. In the process of determining the fire smoke plume area, the temperature diffusion characteristics include the change in the center position of the smoke plume area, the vertical diffusion velocity, the temperature attenuation gradient, the expansion ratio of the high-temperature area, and... The continuous existence time, wherein the temperature diffusion characteristics are input into a preset rule judgment model for judgment, when the undetermined smoke plume region simultaneously meets the preset spatial movement characteristics, temperature change characteristics and duration conditions, the undetermined smoke plume region is determined to be a fire smoke plume region; otherwise, the region is determined to be a non-fire thermal anomaly region. For example, fire smoke plume regions usually have the characteristics of spreading upward from the fire source region, gradually decreasing temperature and changing region range over time. Based on the historical fire smoke plume diffusion characteristics, a rule judgment model is constructed to identify misjudged regions in the undetermined smoke plume region, thereby improving the accuracy of fire smoke plume region identification.

[0079] In some embodiments, the location of the fire is used as the reference benchmark for the vertical height of the smoke plume. Based on the upward diffusion direction of the fire smoke plume, the fire plume region is divided into multiple height layers, each corresponding to a certain spatial height range. Temperature distribution information within each height layer is obtained. For example, the fire plume region is divided into a near-fire source layer, an intermediate diffusion layer, and a far-end diffusion layer based on the smoke plume diffusion height. Temperature characteristic parameters corresponding to each height layer are statistically analyzed, including average temperature, maximum temperature, temperature change amplitude, proportion of high-temperature area, and temperature change rate. Furthermore, a vertical temperature distribution model of the smoke plume is constructed based on the smoke plume temperature characteristics corresponding to different height layers. Specifically, the smoke plume height is used as a spatial variable, and the temperature characteristics of the corresponding height layer are used as model parameters. A fitting function between the smoke plume height and temperature change is established as the vertical temperature distribution model of the smoke plume.

[0080] It should be noted that after a forest fire occurs, the burning area continuously releases heat, causing the surrounding air to heat up and form an updraft. This updraft carries smoke particles produced by the fire, forming a plume. The movement of the plume is influenced by the energy released by the fire source, the intensity of combustion, and environmental diffusion conditions. Its spatial temperature distribution is directly related to the fire's development status. Existing forest fire risk assessment methods typically rely on environmental factors such as meteorological conditions and vegetation status to determine the probability of a fire. They have limited ability to observe the heat release intensity and spread trend after a fire occurs, and are particularly difficult to identify the current fire energy based on the plume temperature distribution, thereby analyzing the fire's spread intensity. This application constructs a vertical temperature distribution model for the plume, using the temperature change patterns of the plume at different heights to characterize the fire's heat release capacity and spatial diffusion status. By analyzing the high-temperature plume's expansion height, temperature gradient changes, and temperature change trends, it achieves dynamic identification of fire risk diffusion energy and fire risk spread intensity, improving the accuracy of risk assessment at different fire development stages. In some embodiments, identifying the fire risk spread intensity based on the plume's vertical temperature distribution model specifically includes:

[0081] Based on the vertical temperature distribution model of the plume, the temperature distribution state of the plume in different time periods is analyzed, and the thermal diffusion feature vector of the plume is extracted.

[0082] A smoke plume identification model is constructed, and the smoke plume thermal diffusion feature vector is clustered based on the smoke plume identification model. The corresponding fire hazard spread intensity is determined based on the clustering results.

[0083] In specific implementation, the vertical temperature distribution model of the plume is obtained for multiple consecutive time periods, and the plume temperature characteristics corresponding to different height layers in each time period are extracted. These temperature characteristics include the average temperature, maximum temperature, temperature gradient, proportion of high-temperature regions, and temperature change rate for each height layer. Difference analysis is performed on the plume temperature distribution states in adjacent time periods to determine the dynamic characteristics of plume thermal diffusion. A plume thermal diffusion feature vector is constructed based on the plume temperature characteristic parameters and the dynamic characteristics of plume thermal diffusion. This feature vector characterizes the spatial distribution of energy released by the current fire and its trend over time. Specifically, the plume thermal diffusion feature vector includes the plume high-temperature holding height feature, vertical temperature gradient feature, high-temperature region expansion feature, and temperature change rate feature. The high-temperature holding height reflects the current fire heat influence range, the vertical temperature gradient reflects the degree of heat attenuation along the plume height, the high-temperature region expansion feature reflects the diffusion trend of the fire's thermal anomaly region, and the temperature change rate reflects the changing state of the fire's released energy.

[0084] Optionally, in some embodiments, a pre-trained single-hidden-layer neural network is used as a plume recognition model to cluster and identify the thermal diffusion feature vectors of the plume, and the corresponding fire spread intensity is determined based on the clustering results. The following is a specific embodiment of the method adopted in this application to cluster and identify the thermal diffusion feature vectors of the plume to determine the corresponding fire spread intensity:

[0085] The heat diffusion feature vector of the plume corresponding to the current fire area is input in the form of a data vector. The clustering recognition model is used to classify and learn the heat diffusion features of the plume corresponding to different fire states, and determines the current plume state category through the clustering output results. According to the mapping relationship between the plume state category and the fire risk spread intensity level, the fire risk spread intensity corresponding to the current fire area is obtained. The different dimensions of the heat diffusion feature vector of the plume may include feature parameters such as the plume high temperature holding height, the plume vertical temperature gradient, the area ratio of the high temperature region, the high temperature region expansion rate, the average temperature of the plume, the maximum temperature of the plume, and the temperature change rate.

[0086] The method employs pre-acquired infrared monitoring data of plumes corresponding to multiple historical forest fire events, along with corresponding fire hazard spread intensity annotations, as training samples. Multiple historical plume thermal diffusion feature vectors are input into the clustering recognition model for training. Specifically, the clustering recognition model classifies plume states with similar temperature distribution characteristics and diffusion change patterns into the same category based on the feature similarity between the input plume thermal diffusion feature vectors, and outputs the corresponding plume state clustering results. Based on the corresponding fire development in historical fire events, the fire hazard spread intensity is labeled for different plume state clustering results. For example, if a cluster category's plume characteristics are characterized by low plume height, concentrated high-temperature areas, and a small rate of temperature change, then that cluster category is labeled as a low spread intensity category; if a cluster category's plume characteristics are characterized by a continuous expansion of the high-temperature area to higher heights, a rapid increase in the area of ​​the high-temperature area, and a continuous rise in temperature, then that cluster category is labeled as a high spread intensity category.

[0087] Specifically, the clustering identification results are correlated with the corresponding fire spread intensity labeling results to establish a mapping relationship between smoke plume state categories and fire spread intensity levels. When the number of historical fire samples corresponding to a certain smoke plume state category reaches a preset number, and the correlation between the smoke plume state category and the corresponding fire spread intensity level meets preset conditions, the fire spread intensity level corresponding to the smoke plume state category is determined.

[0088] Furthermore, in practical applications, the smoke plume thermal diffusion feature vector corresponding to the current fire area is input into the clustering identification model. By calculating the feature distance between the current smoke plume thermal diffusion feature vector and each smoke plume state category, the smoke plume state category to which the current smoke plume thermal diffusion feature vector belongs is determined. Based on the pre-established mapping relationship between smoke plume state categories and fire risk spread intensity levels, the fire risk spread intensity corresponding to the current fire area is determined. When the correlation between the clustering result and the fire risk spread intensity labeling result corresponding to historical fire states is lower than a preset threshold, the cluster center parameter or feature weight parameter in the clustering identification model is adjusted, and the clustering analysis of the smoke plume thermal diffusion feature vector is re-executed until the correlation between the clustering result and the fire risk spread intensity labeling result reaches a preset standard, and the mapping relationship between smoke plume state categories and fire risk spread intensity is obtained.

[0089] Preferably, in some embodiments, the fire spread risk identification based on the fire spread intensity and the multi-source monitoring data to obtain the fire spread level specifically includes:

[0090] Obtain multi-source monitoring data corresponding to the current fire area, and extract environmental characteristic parameters based on the multi-source monitoring data, which includes meteorological monitoring data, vegetation status data, topographic environment data, and historical fire risk data;

[0091] A fire spread feature vector is constructed based on the fire spread intensity and the multi-source monitoring data.

[0092] Based on the fire risk diffusion feature vector, a diffusion risk correlation analysis is performed to generate dynamic diffusion weights; the fire risk diffusion features in the fire risk diffusion feature vector are fused and calculated according to the dynamic diffusion weights to obtain the fire risk diffusion risk value and perform fire risk diffusion level mapping.

[0093] Specifically, for meteorological monitoring data, ambient temperature, air humidity, wind speed, and wind direction characteristics are extracted to characterize the thermal environment of the fire area and the diffusion direction of flames and plumes affected by external airflow; for vegetation status data, vegetation coverage, vegetation moisture content, and combustible material distribution characteristics are extracted to characterize the continuity of combustible materials in the target area; for topographic environment data, topographic slope, altitude change, and topographic orientation characteristics are extracted to characterize the resistance and trend of fire propagation along the topographic direction.

[0094] Furthermore, a fire spread feature vector is constructed based on the fire spread intensity and the multi-source monitoring data.

[0095] The fire risk spread feature vector includes the current fire spread intensity feature, wind-driven feature, fuel continuity feature, and terrain propagation feature.

[0096] Specifically, the current fire spread intensity characteristics are determined by the aforementioned vertical temperature distribution model of the smoke plume, reflecting the energy released and heat diffusion capacity of the current fire; the wind-driven characteristics are determined based on wind speed, wind direction, and wind speed change trends, reflecting the propagation capacity of flames and high-temperature smoke plumes after being affected by external airflow; the fuel continuity characteristics are determined based on vegetation coverage and vegetation moisture content, reflecting whether the surrounding area has conditions for continuous combustion; and the terrain propagation characteristics are determined based on terrain slope and spatial direction relationship, reflecting the possibility of fire spreading along the terrain direction.

[0097] Furthermore, based on the fire risk diffusion feature vector, diffusion risk correlation analysis is performed to determine the degree of correlation between different influencing factors, and dynamic diffusion weights are generated according to the degree of influence of each influencing factor on the fire diffusion process.

[0098] For example, when the current fire spread intensity is high, and strong wind speed, low vegetation moisture content, and continuous combustible vegetation distribution are detected in the target area, the weights corresponding to wind-driven characteristics and fuel continuity characteristics are increased; when the current fire spread intensity is high but the surrounding environment has high humidity and high vegetation moisture content, the risk contribution corresponding to environmental diffusion factors is reduced.

[0099] Furthermore, the fire risk diffusion feature vector is fused and calculated based on the dynamic diffusion weight to obtain the fire risk diffusion risk value. Specifically, the current fire spread intensity and various environmental diffusion feature parameters are normalized and weighted and fused according to the corresponding preset influence weights to obtain the comprehensive diffusion risk value. Further, the current fire risk diffusion level is determined based on the relationship between the fire risk diffusion risk value and the preset diffusion level threshold. For example, when the fire risk diffusion risk value is low, it indicates that the current fire energy is weak and is determined to be a low diffusion level; when the fire risk diffusion risk value is in the medium range, it is determined to be a medium diffusion level; when the fire risk diffusion risk value is in the high range, it indicates that the current fire releases strong energy and is determined to be a high diffusion level.

[0100] In step S105, forest fire risk warning information is generated based on the corresponding fire risk level and fire spread level.

[0101] Optionally, in some embodiments, generating forest fire risk warning information based on the corresponding fire risk level and fire spread level specifically includes:

[0102] Obtain the fire risk level and fire spread level corresponding to the target forest area;

[0103] When the fire risk level is higher than the preset risk level threshold, risk prevention and early warning information is generated as forest fire risk early warning information.

[0104] When the fire risk spread level is higher than the preset spread level threshold, a high-level fire emergency warning message is generated as a forest fire risk warning message.

[0105] Optionally, in some embodiments, after generating forest fire risk warning information based on the corresponding fire risk level and fire spread level, the method further includes: sending the forest fire risk warning information to a fire risk remote monitoring platform, whereby platform personnel take corresponding protective measures according to the warning level.

[0106] Optionally, in some embodiments, the forest fire risk warning information is dynamically updated based on the fire risk level and fire spread level obtained within a continuous monitoring period. For example, when the fire risk level continues to rise over multiple consecutive monitoring periods, the warning level for the corresponding area is raised; when a fire occurs, if the fire spread level continues to increase, the warning information is updated and the emergency response level is raised.

[0107] In another aspect, in some embodiments, this application provides a fire hazard warning device, with reference to... Figure 3The figure is a schematic diagram of the exemplary hardware and / or software structure of a fire hazard warning device according to some embodiments of this application. The fire hazard warning device 200 includes: a fire hazard determination module 201, a multi-source data acquisition module 202, a multi-source data analysis module 203, and a fire hazard warning module 204, which are described below:

[0108] The fire risk assessment module 201 is used to acquire forest image monitoring data and make a fire risk assessment based on the forest image monitoring data.

[0109] The multi-source data acquisition module 202 is used to acquire multi-source monitoring data within the target forest area and perform spatiotemporal correlation processing on the multi-source monitoring data to determine multi-dimensional fire risk characteristic parameters.

[0110] The multi-source data analysis module 203 is used to perform risk analysis based on the multi-dimensional fire risk characteristic parameters and determine the fire risk level corresponding to the target forest area.

[0111] The multi-source data analysis module 203 is also used to acquire infrared detection images of fire smoke, construct a vertical temperature distribution model of the smoke plume, identify the fire spread intensity based on the vertical temperature distribution model of the smoke plume, identify the spread risk based on the fire spread intensity and the multi-source monitoring data, and obtain the fire spread level.

[0112] The fire risk warning module 204 is used to generate forest fire risk warning information based on the corresponding fire risk level and fire spread level.

[0113] The foregoing detailed an example of a dynamic assessment and early warning method for forest fire risk levels based on multi-source data fusion provided in the embodiments of this application. It is understood that the corresponding device includes hardware structures and / or software modules for performing each function in order to achieve the above functions.

[0114] Those skilled in the art should readily recognize that, based on the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein, this application can be implemented in hardware or a combination of hardware and computer software. Whether a certain function in the application is executed in a manner that drives hardware or computer software depends on the specific application and design constraints of the technical solution. Therefore, those skilled in the art can use different methods to implement the described function for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0115] In addition, this application also provides a computer terminal device, which includes a memory and a processor. The memory stores code, and the processor is configured to acquire the code and execute the above-described method for dynamic assessment and early warning of forest fire risk levels based on multi-source data fusion.

[0116] In some embodiments, reference Figure 4 The figure is a schematic diagram of the structure of a computer terminal device implementing a dynamic assessment and early warning method for forest fire risk levels based on multi-source data fusion, according to some embodiments of this application. The dynamic assessment and early warning method for forest fire risk levels based on multi-source data fusion in the above embodiments can... Figure 4 The computer terminal device 300 shown is used to implement this, and the computer terminal device 300 includes at least one communication bus 301, communication interface 302, processor 303 and memory 304.

[0117] The processor 303 can be a general-purpose central processing unit (CPU), an application-specific integrated circuit (ASIC), or one or more devices used to control the execution of a dynamic assessment and early warning method for forest fire risk levels based on multi-source data fusion as described in this application.

[0118] The communication bus 301 may include a path for transmitting information between the aforementioned components.

[0119] Memory 304 may be a read-only memory (ROM) or other type of static storage device capable of storing static information and instructions, random access memory (RAM) or other type of dynamic storage device capable of storing information and instructions, or electrically erasable programmable read-only memory (EEPROM), compact disc read-only memory (CD-ROM) or other optical disc storage, optical disc storage (including compressed optical discs, laser discs, optical discs, digital universal optical discs, Blu-ray discs, etc.), magnetic disks or other magnetic storage devices, or any other medium capable of carrying or storing desired program code in the form of instructions or data structures and accessible by a computer, but not limited thereto. Memory 304 may exist independently and be connected to processor 303 via communication bus 301. Memory 304 may also be integrated with processor 303.

[0120] The memory 304 stores program code for executing the scheme of this application, and its execution is controlled by the processor 303. The processor 303 executes the program code stored in the memory 304. The program code may include one or more software modules. In the above embodiments, the determination of the fire hazard spread level can be achieved by the processor 303 and one or more software modules in the program code in the memory 304.

[0121] Communication interface 302 uses any transceiver-like device for communicating with other devices or communication networks, such as Ethernet, radio access network (RAN), wireless local area networks (WLAN), etc.

[0122] Optionally, the computer terminal device 300 may also include a power supply 305 for providing power to various devices or circuits in the real-time computer terminal device.

[0123] In a specific implementation, as one example, a computer terminal device may include multiple processors, each of which may be a single-core (single-CPU) processor or a multi-core (multi-CPU) processor. Here, a processor may refer to one or more devices, circuits, and / or processing cores used to process data (e.g., computer program instructions).

[0124] The aforementioned computer terminal device can be a general-purpose computer terminal device or a dedicated computer terminal device. In specific implementations, the computer terminal device can be a desktop computer, a portable computer, a network server, a handheld computer (PDA), a mobile phone, a tablet computer, a wireless terminal device, a communication device, or an embedded device. This application does not limit the type of computer terminal device.

[0125] In addition, other aspects of this application also provide a computer-readable storage medium storing at least one computer program, which is loaded and executed by a processor to perform the operations described above in a method for dynamic assessment and early warning of forest fire risk levels based on multi-source data fusion.

[0126] In summary, the forest fire risk level dynamic assessment and early warning method based on multi-source data fusion disclosed in this application first acquires forest image monitoring data and makes a fire risk judgment based on the forest image monitoring data; when no fire risk is determined, multi-source monitoring data within the target forest area is acquired, and spatiotemporal correlation processing is performed on the multi-source monitoring data to determine multi-dimensional fire risk characteristic parameters; risk analysis is performed based on the multi-dimensional fire risk characteristic parameters to determine the fire risk level corresponding to the target forest area; when a fire risk exists, infrared detection images of fire smoke are acquired, and a vertical temperature distribution model of the smoke plume is constructed; the fire spread intensity is identified based on the vertical temperature distribution model of the smoke plume, and diffusion risk is identified based on the fire spread intensity and multi-source monitoring to obtain the fire spread level; forest fire risk early warning information is generated based on the corresponding fire risk level and fire spread level, which can dynamically adjust the fire risk level according to the vertical temperature distribution of the smoke plume, thus improving the accuracy of fire risk level assessment.

[0127] The above descriptions are merely embodiments of this application, and common knowledge such as specific technical solutions or characteristics in the solutions are not described in detail here. It should be noted that those skilled in the art can make several modifications and improvements without departing from the technical solutions of this application, and these should also be considered within the scope of protection of this application, without affecting the effectiveness of the implementation of this application or the practicality of the patent.

Claims

1. A method for dynamic assessment and early warning of forest fire risk levels based on multi-source data fusion, characterized in that, include: Acquire forest image monitoring data and make fire risk assessment based on the forest image monitoring data; When it is determined that there is no fire risk, multi-source monitoring data within the target forest area is acquired, and the multi-source monitoring data is subjected to spatiotemporal correlation processing to determine multi-dimensional fire risk characteristic parameters. Risk analysis is conducted based on the aforementioned multidimensional fire risk characteristic parameters to determine the fire risk level corresponding to the target forest area; When a fire hazard exists, infrared detection images of fire smoke are acquired, and a vertical temperature distribution model of the smoke plume is constructed. The intensity of fire spread is identified based on the vertical temperature distribution model of the smoke plume. Based on the intensity of fire spread and the multi-source monitoring data, the spread risk is identified, and the fire spread level is obtained. Forest fire risk warning information is generated based on the corresponding fire risk level and fire spread level. The acquisition of infrared detection images of fire smoke and the construction of a vertical temperature distribution model for the smoke plume specifically include: The process involves acquiring a sequence of infrared detection images of fire smoke. For any given infrared detection image, spatial correction is performed based on the device's imaging parameters and a preset spatial relationship to determine the spatial position of each pixel in the image. Thermal anomaly regions are identified within the infrared detection images to determine undetermined smoke plume regions. Undetermined smoke plume regions are identified in continuous infrared images, and temperature diffusion characteristics of these regions are identified using continuous infrared images. These temperature diffusion characteristics are then extracted and used for rule-based determination to identify the fire smoke plume regions. A height-layered model is established based on the spatial distribution characteristics of the fire smoke plume regions, and a vertical temperature distribution model of the smoke plume is constructed based on the smoke plume temperature characteristics corresponding to different height layers. Specifically, identifying the intensity of fire spread based on the vertical temperature distribution model of the smoke plume includes: Based on the vertical temperature distribution model of the plume, the temperature distribution state of the plume in different time periods is analyzed, and the heat diffusion feature vector of the plume is extracted. A plume identification model is constructed, and the heat diffusion feature vector of the plume is clustered based on the plume identification model. The corresponding fire spread intensity is determined based on the clustering results.

2. The method as described in claim 1, characterized in that, The fire risk assessment based on the aforementioned forest image monitoring data specifically includes: The forest image monitoring data is divided into target areas to obtain multiple spatial monitoring units; For any spatial monitoring unit, abnormal regions are identified by pixel thresholds, and abnormal feature parameters of the abnormal regions are extracted. Fire hazard identification is performed based on the aforementioned abnormal characteristic parameters to determine whether there is a fire hazard in the space monitoring unit.

3. The method as described in claim 1, characterized in that, The spatiotemporal correlation processing of the multi-source monitoring data specifically includes: dividing the target forest area into multiple spatial monitoring units, obtaining the spatial coordinate information corresponding to each monitoring data, mapping the multi-source monitoring data to the corresponding spatial monitoring units; establishing a unified time window, and performing time synchronization processing on the multi-source monitoring data collected within a preset time range.

4. The method as described in claim 1, characterized in that, Based on the aforementioned multidimensional fire risk characteristic parameters, risk analysis is conducted to determine the fire risk level corresponding to the target forest area, specifically including: Based on the multidimensional fire hazard feature parameters, feature association judgment is performed to generate dynamic feature weights between different fire hazard feature parameters. The fire risk level of the target forest area is obtained by fusing and calculating the multidimensional fire risk feature parameters based on the dynamic feature weights.

5. The method as described in claim 1, characterized in that, Forest fire risk warning information is generated based on the corresponding fire risk level and fire spread level, specifically including: Obtain the fire risk level and fire spread level corresponding to the target forest area; When the fire risk level is higher than the preset risk level threshold, risk prevention and early warning information is generated as forest fire risk early warning information. When the fire risk spread level is higher than the preset spread level threshold, a high-level fire emergency warning message is generated as a forest fire risk warning message.

6. The method as described in claim 1, characterized in that, After generating forest fire risk warning information based on the corresponding fire risk level and fire spread level, the process also includes sending the forest fire risk warning information to the fire risk remote monitoring platform.

7. A computer terminal device, characterized in that, The computer terminal device includes a memory and a processor, the memory storing code, and the processor being configured to retrieve the code, such that when the code is executed in the processor, the method as described in any one of claims 1-6 is implemented.

8. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions that, when executed on a computer, cause the method as described in any one of claims 1-6 to be implemented.

Images