Mountain fire monitoring and fire prevention method and system based on cloud platform

Abstract

The invention relates to the technical field of disaster prevention and control, in particular to a mountain fire monitoring and fire prevention method and system based on a cloud platform, and the method comprises the following steps: submitting a fire prevention instruction group based on cloud label mapping, node screening, equipment linkage and heat source identification. According to the method, dynamic division and task node numbering are carried out on a monitoring area through spatial indexes, thermosensitive difference analysis of area image frames is completed in combination with linkage of air and ground equipment, abnormal high-temperature coordinates are automatically extracted based on gray difference sudden change, time sequence verification is carried out on trigger nodes and heat source coordinates, and a prevention and control response task is established. Automatic sensing, efficient response and continuous prevention and control of early-stage hidden dangers of a fire in a complex mountain environment are achieved, automation and intelligentization of the whole process from fire early warning, intelligent response to postmortem analysis are achieved through deep integration of multiple technologies, the fire prevention and control efficiency and safety in the complex mountain terrain are remarkably improved, and the fire prevention and control cost is lowered. And data support is provided for continuous optimization of the emergency plan.

Description

Technical Field

[0001] This invention relates to the field of disaster prevention and control technology, and in particular to a cloud-based method and system for monitoring and preventing mountain fires. Background Technology

[0002] The field of disaster prevention and control technology mainly involves comprehensive methods for real-time monitoring, early warning, and emergency response to natural or man-made disasters. This includes disaster identification mechanisms, monitoring technology deployment, early warning information transmission, disaster data analysis, and response measures deployment. Typically, this is achieved through deploying multi-source sensor networks, constructing disaster information collection systems, utilizing data transmission links to realize real-time disaster monitoring, and coordinating with command and dispatch systems for disaster response. Among these methods, mountain fire monitoring and prevention refers to addressing the characteristics of mountainous areas, which are prone to fires due to dense vegetation and complex terrain. This is achieved through methods such as manual patrols, setting up fixed-point cameras to capture fire images, constructing fixed lookout towers for visual fire source monitoring, and laying fire-retardant firebreaks in historically fire-prone areas during hot, dry seasons to monitor fires and intervene in potential hazards.

[0003] In current mountain fire monitoring and prevention processes, monitoring of mountain fires mainly relies on setting up fixed-point equipment and manual patrols. Due to the complex terrain, the deployment area of ​​equipment is limited, and image acquisition is prone to blind spots. Under the condition of high-frequency patrols in large areas, it is necessary to rely on manual judgment of changes in fire source images, lacking a highly timely early warning and response mechanism. It is difficult to efficiently identify monitoring areas in response to variable weather conditions and dynamic fire situations. Especially during high-risk periods such as nighttime or sudden changes in wind, information delays can easily lead to the spread of fires, resulting in monitoring omissions and slow responses. At the same time, the response to fires in complex mountain terrain is slow, the coordination efficiency is low, and there is a lack of effective data review, which cannot meet the needs of dynamic early warning and task-oriented scheduling in areas with significant terrain differences. Summary of the Invention

[0004] To address the technical problems existing in the prior art, this invention provides a cloud-based mountain fire monitoring and prevention method, comprising the following steps: To achieve the above objectives, the present invention adopts the following technical solution: a mountain fire monitoring and prevention method based on a cloud platform, comprising the following steps: S1: Obtain GIS 3D topographic raster layer, surface slope information, meteorological wind speed sequence and temperature and humidity monitoring data of mountainous areas, generate labels for each topographic grid, and nest and map the temperature and humidity data, and upload it to the cloud to generate a set of topographic meteorological factor labels; S2: Based on the terrain and meteorological factor tag set, filter the tag areas and extract the center coordinate points, establish a node set for each area, and upload it to the cloud platform task node interface after establishing a unique number identifier to generate a monitoring trigger node sequence; S3: Based on the monitoring trigger node sequence, obtain the UAV number and ground sensor device number in the area, determine the relationship between the current operating status and the number position, combine and assign them and submit them to the execution queue, and obtain the air-ground device linkage number table. S4: Based on the air-ground equipment linkage number table, start the equipment corresponding to the number and collect regional image frames, calculate the difference between the single pixel thermal radiation gray value in the image frame and the average value of adjacent pixels in the same frame and determine whether it is greater than the thermal sensitivity difference threshold, mark the abnormal temperature change feature area, extract the corresponding layer coordinates and upload them to the cloud platform to obtain the high temperature heat source coordinate sequence. S5: Based on the high-temperature heat source coordinate sequence, verify whether the coordinate points in the sequence already exist in the monitoring trigger node sequence, submit all tasks to the cloud platform fire prevention task execution queue and mark the timestamp, and output the mountain fire monitoring and fire prevention instruction group.

[0005] As a further embodiment of the present invention, the terrain and meteorological factor label set includes terrain label categories, time-series nested meteorological factors, geographic grid index numbers, and cloud storage paths; the monitoring trigger node sequence includes node spatial coordinates, unique identifiers, node-assigned areas, and trigger priorities; the air-ground equipment linkage number table includes a set of statuses corresponding to equipment numbers, a set of coordinates corresponding to equipment numbers, combined assignment records, and equipment type identifiers; the high-temperature heat source coordinate sequence includes a set of abnormal heat source coordinate points, thermally sensitive grayscale differences, layer mapping indexes, and upload time stamps; and the mountain fire monitoring and fire prevention instruction group includes task node numbers, prevention and control task identifiers, fire prevention task timestamps, and task execution priorities.

[0006] As a further aspect of the present invention, the step of obtaining the topographic meteorological factor label set is as follows: S111: Obtain the 3D topographic raster layer and surface slope information of the mountainous area, extract the slope value of each raster point in the topographic raster layer, combine it with the wind speed value of the corresponding geographical location in the meteorological wind speed sequence, and perform index matching of the slope value and wind speed value of each raster point. By performing two-parameter grouping and statistical analysis on each slope value and its corresponding wind speed value, a topographic wind speed label matrix is ​​generated. S112: Based on the terrain and wind speed label matrix, extract the time series from the temperature and humidity monitoring data, perform coordinate mapping on the geographical coordinate points recorded in each time series data frame, and generate a nested meteorological factor matrix. S113: Based on the nested meteorological factor matrix, and combined with the label index of each grid point in the terrain wind speed label matrix, all meteorological factor data frames under the same label are sequentially aggregated. The aggregated data of each label is standardized and encoded, and the encoding results are uploaded to the cloud in one-to-one correspondence with the label index to generate a terrain meteorological factor label set.

[0007] As a further aspect of the present invention, the coordinate mapping specifically involves matching geographic coordinates with the location index of grid points in the terrain raster layer, while simultaneously nesting and mapping the temperature and humidity values ​​at the corresponding time to the corresponding grid point locations.

[0008] As a further aspect of the present invention, the step of obtaining the monitoring trigger node sequence is as follows: S211: Based on the topographic and meteorological factor tag set, establish a tag area retrieval according to the tag index field, perform grouping and clustering processing on all tag categories in the tag set, calculate the maximum and minimum values ​​of the horizontal and vertical coordinates of the grid points in two-dimensional space according to the spatial index data of the grid points in the tag group, and generate a tag area center point coordinate set by extracting the coordinates of the geometric boundary center point of each group of grid areas as the representative location point of the corresponding area. S212: Based on the coordinate set of the center points of the tag area, perform clustering according to the tag number to which each group of center points belongs, merge each group of center points into the node set indicated by the corresponding number, perform spatial index serialization processing on each node set, and set an independent number as a unique identifier for each node set to obtain the spatial node number set. S213: Based on the set of spatial node numbers, the node numbers and node coordinates in the set are combined and encapsulated into a node index structure, and written into the cloud platform task node interface in the form of a structure. Remote calls to node metadata are executed and mapped and mounted to establish a monitoring trigger node sequence.

[0009] As a further aspect of the present invention, the steps for obtaining the air-to-ground equipment linkage number table are as follows: S311: Based on the monitoring trigger node sequence, read the spatial coordinates in the index structure of each node, retrieve the UAV number and ground sensor device number in the matching area, collect the running status information and coordinate positioning information at the current timestamp, associate and bind each device number with its status data, and obtain the device running status mapping table. S312: Based on the device operation status mapping table, aggregate the UAV numbers and ground sensor numbers in the area according to the spatial distribution of nodes, combine the device numbers in each node into a task unit, filter the deployable devices according to the current operation status, record the status flags corresponding to each number, and submit the combined device numbers to establish a device deployment combination matrix. S313: Based on the equipment allocation and combination matrix, the equipment numbers of each group are serially encoded to construct an executable linkage task index number. The numbering standard verification and uniqueness detection are performed on each group of linkage numbers. After summarizing, the results are written into the task execution queue of the cloud platform, and the task status is set to generate an air-ground equipment linkage number table.

[0010] As a further aspect of the present invention, the step of obtaining the high-temperature heat source coordinate sequence is as follows: S411: Based on the air-ground equipment linkage number table, extract the set of UAV number and ground equipment number, and send a start command to trigger the image acquisition function. Perform synchronous image frame capture operation in the task area, perform pixel-by-pixel decoding on each image frame, convert the thermal radiation gray value of each pixel into a two-dimensional matrix structure, and generate a regional image frame matrix set. S412: Based on the thermal radiation grayscale matrix of each frame image in the regional image frame matrix set, perform neighborhood window analysis on all pixels, extract the average grayscale value of adjacent pixels, construct a pixel difference matrix, perform item-by-item thermal threshold judgment on the difference matrix and mark abnormal regions, calculate and obtain the binary mask image of the abnormal region, identify and extract the set of boundary coordinates of connected regions, and establish the coordinate set of the abnormal temperature change layer. S413: Based on each coordinate pair in the coordinate set of the abnormal temperature change layer, match the device number and timestamp index embedded in the original image frame, map each set of coordinates back to the corresponding device and acquisition time, convert it into a heat source location vector with spatiotemporal attributes, and submit it to the cloud platform spatial database to establish a high-temperature heat source coordinate sequence.

[0011] As a further aspect of the present invention, during the image frame capture operation, the acquired image frame is a thermal infrared image type with a grayscale range between 0 and 255.

[0012] As a further aspect of the present invention, the steps for obtaining the mountain fire monitoring and fire prevention command group are as follows: S511: Based on the coordinate points in the high-temperature heat source coordinate sequence, read the latitude and longitude values ​​point by point, and perform spatial matching with the node center coordinate set in the monitoring trigger node sequence to mark the registered nodes and unregistered area points, and generate a registered node comparison matrix. S512: Based on the registered node comparison matrix, filter the registered nodes, establish task prevention and control identifiers for the node data, and attach a task number prefix to form a unique prevention and control task number. Store all task identifier records in the task buffer to be executed to form a node prevention and control task list and generate a prevention and control task identifier set. S513: Based on the aforementioned prevention and control task identifier set, arrange all record structures in ascending order by timestamp and write them in batches into the cloud platform fire prevention task execution queue. Perform integrity verification on each task structure, and generate a task receipt identifier in the cloud and return a confirmation status code to obtain the mountain fire monitoring and fire prevention instruction group.

[0013] A cloud-based mountain fire monitoring and prevention system includes: The factor label generation module is used to execute S1: acquire GIS 3D topographic raster layers, surface slope information, meteorological wind speed sequence and temperature and humidity monitoring data of mountainous areas, generate labels for each topographic grid, nest and map the temperature and humidity data, and upload them to the cloud to generate a set of topographic meteorological factor labels; The node sequence extraction module is used to execute S2: based on the terrain and meteorological factor label set, filter the label areas and extract the center coordinate points, establish the node set of each area, and upload it to the cloud platform task node interface after establishing a unique number identifier to generate the monitoring trigger node sequence; The equipment task allocation module is used to execute S3: based on the monitoring trigger node sequence, obtain the UAV number and ground sensor device number in the area, determine the relationship between the current operating status and the number position, combine and assign them and submit them to the queue to be executed, and obtain the air-ground device linkage number table; The heat source area identification module is used to execute S4: based on the air-ground equipment linkage number table, start the equipment corresponding to the number and collect regional image frames, calculate the difference between the single pixel thermal radiation gray value in the image frame and the average value of adjacent pixels in the same frame and determine whether it is greater than the thermal sensitivity difference threshold, mark the abnormal temperature change feature area, extract the corresponding layer coordinates and upload them to the cloud platform to obtain the high temperature heat source coordinate sequence. The fire prevention instruction issuance module is used to execute S5: based on the high-temperature heat source coordinate sequence, verify whether the coordinate points in the sequence already exist in the monitoring trigger node sequence, submit all tasks to the cloud platform fire prevention task execution queue and mark the timestamp, and output the mountain fire monitoring fire prevention instruction group.

[0014] Compared with the prior art, the advantages and positive effects of the present invention are as follows: In this invention, the monitoring area is dynamically divided and task node numbers are generated through spatial indexing, enabling proactive identification of high-risk areas and precise assignment of task targets. Combined with the linkage of air and ground equipment, thermal difference analysis of regional image frames is completed. Based on the abrupt change in grayscale difference, abnormal high-temperature coordinates are automatically extracted, forming a heat source perception mechanism oriented towards changing characteristics. The trigger node and heat source coordinates are verified in sequence to establish prevention and control response tasks, avoiding blind spots in human inspections and static monitoring dead zones. This improves the timeliness of response and the accuracy of fire identification in high-risk areas, and realizes automatic perception, efficient response, and continuous prevention and control of early fire hazards in complex mountainous environments. Through the deep integration of multiple technologies, the entire process from fire early warning and intelligent response to post-event analysis is automated and intelligent, significantly improving the efficiency and safety of fire prevention and control in complex mountainous terrain, and providing data support for the continuous optimization of emergency plans. Attached Figure Description

[0015] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0016] Figure 1 This is a schematic diagram of the steps of the present invention; Figure 2 This is a flowchart of the process for obtaining the topographic and meteorological factor tag set in this invention; Figure 3 This is a flowchart of the process for obtaining the monitoring trigger node sequence in this invention; Figure 4 This is a flowchart for obtaining the air-to-ground equipment linkage number table of the present invention; Figure 5 This is a flowchart of the process for obtaining the high-temperature heat source coordinate sequence in this invention; Figure 6 This is a flowchart of the process for obtaining the mountain fire monitoring and fire prevention command group in this invention; Figure 7 This is a system module diagram of the present invention. Detailed Implementation

[0017] The technical solution of the present invention will now be described with reference to the accompanying drawings.

[0018] In embodiments of the present invention, words such as "exemplarily," "for example," etc., are used to indicate that something is an example, illustration, or description. Any embodiment or design described as "exemplary" in the present invention should not be construed as being more preferred or advantageous than other embodiments or designs. Specifically, the use of the word "exemplary" is intended to present the concept in a concrete manner. Furthermore, in embodiments of the present invention, the meaning expressed by "and / or" can be both, or either one.

[0019] In the embodiments of this invention, the terms "image" and "picture" may sometimes be used interchangeably. It should be noted that, without emphasizing the distinction between them, they convey the same meaning. Similarly, the terms "of," "corresponding (relevant)," and "corresponding" may sometimes be used interchangeably. It should be noted that, without emphasizing the distinction between them, they convey the same meaning.

[0020] In this embodiment of the invention, sometimes a subscript such as W1 may be written in a non-subscript form such as W1. When the difference is not emphasized, the meaning they express is the same.

[0021] To make the technical problems, technical solutions and advantages of the present invention clearer, a detailed description will be given below in conjunction with the accompanying drawings and specific embodiments.

[0022] Please see Figure 1 This invention provides a cloud-based method for monitoring and preventing mountain fires, comprising the following steps: S1: Obtain the 3D topographic raster layer, surface slope information, meteorological wind speed sequence and temperature and humidity monitoring data of the mountainous area; generate labels for each topographic grid according to the correspondence between slope and wind speed; and map the temperature and humidity data to each grid point according to the nested time index; and upload it to the cloud to generate a set of topographic meteorological factor labels. S2: Based on the terrain and meteorological factor tag set, filter the tag areas by spatial index and extract the center coordinate points, establish a node set for each area, and upload the set to the cloud platform task node interface after establishing a unique number identifier for the set to generate a monitoring trigger node sequence. S3: Based on the monitoring trigger node sequence, obtain the UAV number and ground sensor device number in each area, determine the relationship between the current operating status and the number location, combine and assign the numbers synchronously, build a task allocation record on the cloud platform and submit it to the queue to be executed, and obtain the air-ground device linkage number table. S4: Based on the device numbers in the air-ground device linkage number table, start the device corresponding to the number and collect regional image frames. Calculate the difference between the single pixel thermal radiation gray value in the image frame and the average value of adjacent pixels in the same frame and determine whether it is greater than the thermal sensitivity difference threshold. Identify the area that meets the condition and mark it as an abnormal temperature change feature area. Extract the layer coordinates corresponding to the abnormal temperature change feature area and upload them to the cloud platform to obtain the high temperature heat source coordinate sequence. S5: Based on the high-temperature heat source coordinate sequence, verify whether each coordinate point in the sequence already exists in the monitoring trigger node sequence, establish prevention and control task identifiers for all registered nodes, submit all tasks to the cloud platform fire prevention task execution queue and mark the timestamp, and output the mountain fire monitoring and fire prevention instruction group.

[0023] The terrain and meteorological factor label set includes terrain label categories, time-series nested meteorological factors, geographic grid index numbers, and cloud storage paths. The monitoring trigger node sequence includes node spatial coordinates, unique number identifiers, node-belonging areas, and trigger priorities. The air-ground equipment linkage number table includes equipment number-corresponding status sets, equipment number-corresponding coordinate sets, combined assignment records, and equipment type identifiers. The high-temperature heat source coordinate sequence includes abnormal heat source coordinate point sets, thermal sensitivity grayscale differences, layer mapping indexes, and upload time stamps. The mountain fire monitoring and fire prevention instruction group includes task node numbers, prevention and control task identifiers, fire prevention task timestamps, and task execution priorities.

[0024] Please see Figure 2 The specific steps of S1 are as follows: S111: Obtain the 3D topographic raster layer and surface slope information of the mountainous area, extract the slope value of each raster point in the topographic raster layer, combine it with the wind speed value of the corresponding geographical location in the meteorological wind speed sequence, and perform index matching of the slope value and wind speed value of each raster point. By performing two-parameter grouping and statistical analysis on each slope value and its corresponding wind speed value, a topographic wind speed label matrix is ​​generated. In the process of acquiring GIS 3D topographic raster layers and surface slope information for mountainous areas, the first step involves extracting the 3D spatial index values ​​and corresponding geographic coordinates of each raster cell based on the topographic layer data provided in the main step. After retrieving the surface slope information, the data is matched to the corresponding position of each raster cell, and the slope of each raster point is extracted. This operation is based on the rate of change of the z-value gradient in the topographic layer. The ratio of the elevation difference to the horizontal distance between adjacent points is converted into an angle value by inverse switching and multiplying by 180 / π. For example, taking cell G1 as an example, its adjacent cell has an elevation difference of 3.5m and a horizontal distance of 16m, so the slope value is arctan(3.5 / 16)×180 / π=12.38 degrees, recorded as 12.5 degrees. Next, the meteorological wind speed sequence is obtained, and the wind speed value consistent with the spatial location of each raster point is read. Assuming that the wind speed corresponding to G1 is 3.8m / s, The wind speed and slope values ​​are paired. After pairing, a bivariate grouping operation is performed on the slope and wind speed data of all grid points. The slope values ​​are divided into 5 levels: 0–10 degrees, 10–20 degrees, 20–30 degrees, 30–40 degrees, and above 40 degrees. At the same time, the wind speed values ​​are divided into 5 levels: 0–2 m / s, 2–4 m / s, 4–6 m / s, 6–8 m / s, and above 8 m / s. G1, with a slope of 12.5 degrees and a wind speed of 3.8 m / s, is assigned to the 10–20 degree slope and 2–4 m / s wind speed range and labeled as category A. Then, all grid numbers in the same range are grouped into one category to generate cluster labels. Next, the same group assignment operation is performed to add label categories to all cluster units. The labels are written as the matrix codes into a two-dimensional grid structure consistent with the GIS layer structure to form a label matrix. Finally, the terrain wind speed label matrix is ​​generated. As shown in Table 1, different grid cells are assigned to different label categories based on the combination of their slope and wind speed values, for subsequent meteorological factor mapping and classification processing.

[0025] S112: Based on the terrain and wind speed label matrix, extract the time series from the temperature and humidity monitoring data, perform coordinate mapping on the geographic coordinate points recorded in each time series data frame (by matching the geographic coordinates with the position index of the grid points in the terrain raster layer, and simultaneously nesting the temperature and humidity values ​​at the corresponding time to the corresponding grid point positions), and generate a nested matrix of meteorological factors. Based on the terrain wind speed label matrix, during the meteorological factor mapping process, the time series information from the temperature and humidity monitoring data is first retrieved, and the geographic coordinate index corresponding to each raster number in the terrain layer is obtained. For example, rasters G1 to G5 represent five different spatial locations, and the observed values ​​of temperature and humidity corresponding to time indices T1 and T2 are extracted. Among them, G1 has a temperature of 18.5℃ and a humidity of 68% at time T1, and a temperature of 19.1℃ and a humidity of 70% at time T2. Subsequently, according to the spatial index of each raster point in the GIS layer, these time series data are nested and mapped to the corresponding rasters. Within the unit, it is stored in the form of a two-dimensional array, that is, each raster unit corresponds to a time series vector, which contains temperature and humidity records at multiple times. The mapping method is to perform Euclidean distance matching on geographic coordinate points in the temperature and humidity data, and combine the time index to establish the mapping relationship between the raster index and meteorological data. Then, the temperature and humidity vector is filled into the original GIS raster structure to complete the two-dimensional nesting of space and time, and construct a complete time series meteorological matrix. For example, the temperature of G2 in T1 and T2 is 20.2℃ and 21.3℃, and the humidity is 72% and 75%, respectively, which finally form a nested meteorological factor matrix. As shown in Table 2, the temperature and humidity data of each grid point are mapped to the corresponding geographic grid according to the time series, forming a nested structure of meteorological factors for classification and aggregation analysis.

[0026] S113: Based on the nested meteorological factor matrix and combined with the label index of each grid point in the topographic wind speed label matrix, all meteorological factor data frames under the same label are sequentially aggregated. The aggregated data of each label is standardized and encoded, and the encoding results are uploaded to the cloud in one-to-one correspondence with the label index to generate a topographic meteorological factor label set. Based on the constructed meteorological factor nesting matrix, the time series data carried by each grid point are classified and aggregated according to label categories. First, the label category information in the terrain and wind speed label matrix is ​​retrieved, and all grid numbers are assigned to their corresponding groups according to their labels. For example, G1 and G3 both belong to label category A, G4 to B, G2 to C, and G5 to D. Then, the meteorological data within each group are averaged to extract their representative statistical indicators. For example, the temperatures of G1 and G3 corresponding to label A at time T1 are 18.5℃ and 17.9℃, and the humidity is 68% and 70%, respectively. At time T2, the temperatures were 19.1℃ and 18.2℃, and the humidity was 70% and 71%, respectively. Therefore, the mean of the temperature sequence is (18.5 + 17.9 + 19.1 + 18.2) / 4 = 18.175℃, and the mean of the humidity sequence is (68 + 70 + 70 + 71) / 4 = 69.75%. This process is repeated to calculate the statistical average for each tag category. After obtaining the complete tag aggregation result, the mean vectors of temperature and humidity are normalized to 0–1. The normalization uses a linear mapping method, setting the minimum value to 0 and the maximum value to 1. The normalization formula is as follows: , Represents the normalized value. This represents the minimum value. Indicates the maximum value. The current value is represented, and the results are then structured and organized. The label category, temperature series mean, and humidity series mean are uploaded as a whole to the cloud database to generate a topographic meteorological factor label set. Table 3 lists the mean values ​​of temperature and humidity sequences for each label category, which are used to further construct meteorological factor classification features or subsequent training data.

[0027] Please see Figure 3 The specific steps of S2 are as follows: S211: Based on the topographic and meteorological factor tag set, establish tag area retrieval according to the tag index field, perform grouping and clustering processing on all tag categories in the tag set, and calculate the maximum and minimum values ​​of the horizontal and vertical coordinates of the grid points in two-dimensional space according to the spatial index data of the grid points within the tag group. By extracting the coordinates of the geometric boundary center point of each group of grid areas as the representative location point of the corresponding area, a tag area center point coordinate set is generated. Based on the topographic and meteorological factor label set, the defined label category fields are first extracted from the label set. A label area index table is then established by retrieving the raster number set corresponding to each label. Next, spatial boundary calculations are performed on the raster point coordinates corresponding to each label category. The minimum and maximum values ​​of the x and y coordinates on the two-dimensional plane are extracted. The center value of the area in the x-axis direction is obtained by adding the maximum and minimum x-coordinates and dividing by 2. Similarly, the center value in the y-axis direction is obtained by adding the maximum and minimum y-coordinates and dividing by 2. This constructs a set of center point coordinates for the area. For example, if the raster numbers corresponding to label category A are G1 and G3, where the coordinates of G1 are (108.2...). The coordinates of G3 are (108.4, 32.3), so its x-coordinate center point is (108.2+108.4) / 2=108.3, and its y-coordinate center point is (32.1+32.3) / 2=32.2. Therefore, the center point coordinates of the A-type area are (108.3, 32.2). The same method is used to calculate the center point coordinates of the grid points of the other label categories to obtain the complete center point coordinate set. For example, if the B-type label only contains G4, its center coordinates are the coordinates of G4 (108.6, 32.3). After the center points of all areas are calculated, each type of label and its corresponding center point are uniformly organized into a center point coordinate set. As shown in Table 4, the center point coordinates of each label area have been extracted and organized based on the boundary data, generating a set of center point coordinates for each label area.

[0028] S212: Based on the coordinate set of the center points of the tag area, perform clustering according to the tag number to which each group of center points belongs, merge each group of center points into the node set indicated by the corresponding number, perform spatial index serialization processing on each node set, and assign an independent number as a unique identifier to each node set to obtain the spatial node number set. Based on the coordinate set of the center points of the tag area, the system first groups the center points according to the tag field to which each center point belongs. A node allocation table is then established for the center points of each tag group. For example, tag A corresponds to two center point coordinates, namely G1 and G3, with the combined center coordinates being (108.3, 32.2). Tags B, C, and D each correspond to only one center point, namely (108.6, 32.3), (109.0, 32.4), and (109.5, 32.6), respectively. Subsequently, a spatial node set is constructed for the center points within each tag group. All coordinate points are mapped to a two-dimensional space and an index sequence is set. Then, an equal-length number sequence is generated based on the number of node groups. The node identifier is generated using a fixed numbering format "NXYY", where N is the node prefix, X is a fixed character, and YY is an auto-incrementing numerical number, starting from N001 and numbered sequentially. Finally, four node sets are established, identified as N001, N002, N003, and N004, respectively. The node numbers are then associated with the corresponding center coordinates one by one, and the resulting numbered node sets are generated. As shown in Table 5, the construction and numbering mapping of the node set have been completed, resulting in a spatial node number set.

[0029] S213: Based on the set of spatial node numbers, combine the node numbers and node coordinates in the set into a node index structure, and write it to the cloud platform task node interface in the form of a structure to perform remote calls to node metadata and perform mapping and mounting, and establish a monitoring trigger node sequence. Based on the set of spatial node numbers, extract the coordinate information of the main center point corresponding to each number and construct an index structure object. Each structure contains a node number field and a center point coordinate field. The structure data format is {number: Nxxx, coordinates: (x, y)}, where the number field comes from the node number field in the table, and the coordinate field comes from the combination of the x and y coordinates of the main center point. For example, the N001 structure is represented as {number: N001, coordinates: (108.3, 32.2)}. All structure objects are encapsulated in the form of data frames and connected to the task node interface module of the cloud platform to perform remote node registration operations. The structure push task is completed through a batch upload mechanism, and the upload status is marked as "uploaded" after successful upload, indicating that the node triggering mechanism has completed the mapping operation, thereby establishing a monitoring trigger node sequence. As shown in Table 6, all node index structures have been created and uploaded, indicating that the monitoring task has been successfully mounted on the cloud platform.

[0030] Please see Figure 4 The specific steps of S3 are as follows: S311: Based on the monitoring trigger node sequence, read the spatial coordinates in the index structure of each node, retrieve the UAV number and ground sensor device number in the matching area, collect the running status information and coordinate positioning information at the current timestamp, associate and bind each device number with its status data, and obtain the device running status mapping table. Based on the monitoring trigger node sequence, the spatial coordinate information in the node structure is retrieved. Within each spatial node, the device registration database is called to extract all UAV IDs and ground sensor IDs within the node's area. The extraction process is performed through a spatial coordinate comparison mechanism with a tolerance range of ±0.01 degrees. The matching rule is that if both the device's x and y coordinates are within ±0.01 degrees of the node's center coordinates, it is considered to belong to the node's coverage area. For example, if the center coordinates of node N001 are (108.3, 32.2), and the device ID is UAV-01 with coordinates (108.30, 32.20), then... If the device is identified as a matching device, the corresponding operating status field is extracted. For example, UAV-01 is in "standby" status and GS-12 is in "normal" status. Then, the device number, device type, current status, and device spatial coordinate fields are recorded, and a status mapping data table structure is constructed. The operating status field is extracted based on the operating mode marker field in the device status data frame. The field value includes categories such as "standby", "normal", "task in progress", "abnormal", and "charging". Finally, all status data and device numbers are grouped and summarized by node to form a complete status record table, generating a device operating status mapping table. As shown in Table 7, the number, status and spatial location of different types of equipment in each node area have been recorded, which facilitates the execution of subsequent allocation tasks.

[0031] S312: Based on the equipment operation status mapping table, aggregate the UAV numbers and ground sensor numbers in the area according to the spatial distribution of nodes, combine the equipment numbers in each node into a task unit, filter the deployable equipment according to the current operation status, record the status flags corresponding to each number, and submit the combined equipment numbers to establish an equipment deployment combination matrix. Based on the equipment operation status mapping table, clustering is first performed according to node number. Equipment type splitting is performed on the equipment number under each node, and UAV number set and ground sensor number set are aggregated respectively. In the aggregation results, equipment numbers marked as "standby" or "normal" in the operation status field are selected as the current deployable equipment set. Boolean filtering is performed on the equipment status field. If the operation status field value is equal to "standby" or "normal", the equipment is included in the deployment pool. For example, UAV-01 in node N001 is in the "standby" status and GS-12 is in the "normal" status, both of which meet the deployment conditions. UAV-02 in node N002 is in the "task" status, which does not meet the conditions but can be preset as a backup. An empty GS number indicates that there is no ground equipment. Only airspace equipment is retained to perform tasks. After filtering, UAV and ground equipment numbers are grouped and merged by node to form deployment task combination relationships. The merging rule is that the equipment numbers are connected by a plus sign to form a combination identifier, and the combination status is recorded as "deployable". Finally, an equipment deployment combination matrix is ​​generated. As shown in Table 8, the equipment combinations that meet the allocation conditions have been generated according to the node number, which will be used for subsequent linkage number generation operations.

[0032] S313: Based on the equipment allocation and combination matrix, each group of equipment numbers is serially encoded to construct an executable linkage task index number. The numbering standard verification and uniqueness detection are performed on each group of linkage numbers. After being summarized, the data is written into the task execution queue of the cloud platform, and the task status is set to generate an air-ground equipment linkage number table. Based on the equipment allocation and combination matrix, each group of equipment numbers in the matrix is ​​treated as a task combination unit to perform a linkage number generation operation. The numbers are concatenated according to the preset air-to-ground linkage format, which is "UAV number + ground equipment number". The linkage pair combination string is formed by connecting them with a plus sign. For example, UAV-01 and GS-12 are combined to form "UAV-01+GS-12", and UAV-02 without ground equipment combination is generated as "UAV-02+-". Then, all generated linkage numbers are checked for number standardization. The check rule is whether the number contains a valid equipment prefix and connection symbol structure. If it meets the rule, it is judged as "passed" and the task status is recorded as "pending execution". Then, all qualified linkage number contents are written to the task execution scheduling queue of the cloud platform to complete the number submission process and finally generate the air-to-ground equipment linkage number table. As shown in Table 9, all generated linkage numbers have passed the format check and completed the task status marking.

[0033] Please see Figure 5 The specific steps of S4 are as follows: S411: Based on the air-to-ground equipment linkage number table, extract the UAV number and ground equipment number set, and send a start command to trigger the image acquisition function. Perform synchronous image frame capture operation within the task area (the acquired image frames are thermal infrared images with a grayscale range of 0 to 255). Perform pixel-by-pixel decoding on each image frame, convert the thermal radiation grayscale value of each pixel into a two-dimensional matrix structure, and generate a set of regional image frame matrices. Based on the combined number field in the air-to-ground equipment linkage number table, the table's field content is first parsed, splitting each group of numbers into individual UAV numbers and ground equipment numbers. The startup command set in the task scheduling module is then invoked to issue a startup command to each equipment number. Upon receiving the startup command, the equipment immediately enters the image acquisition preparation state, with the startup task identified as "image frame capture." The startup timestamp is based on the current server time, for example, 12:00:00. After startup, each equipment performs image acquisition initialization parameter settings, setting the resolution to 640×480 pixels, the frame rate to 25fps, and the acquisition mode to thermal infrared mode, with the corresponding thermal radiation grayscale range set to 0~255. After initialization, it enters the acquisition state. Each acquired thermal infrared image frame is stored in a two-dimensional grayscale matrix, with all pixels within the frame numbered by row and column index and their grayscale values ​​recorded. For example, in the image with frame number IMG_0001, the pixel at position (100, 120) has a grayscale value of 190. To ensure temporal correlation, each image frame is accompanied by a metadata tag after acquisition, including the image frame number, acquisition device number, acquisition timestamp, and region location information. This metadata is stored in the image frame header and along with the image frame. Subsequently, all image frame data is grouped by node number. This data structure is grouped according to the acquisition task, with each group containing at least 3 image frames. The acquisition process lasts for 5 seconds, during which each device acquires an average of 125 frames. All image frames are uniformly named and written to the device's cache folder for subsequent interpolation analysis to generate a regional image frame matrix set.

[0034] S412: Based on the thermal radiation grayscale matrix of each frame in the regional image frame matrix set, perform neighborhood window analysis on all pixels, extract the average grayscale value of adjacent pixels, construct a pixel difference matrix, perform item-by-item thermal threshold judgment on the difference matrix, and mark abnormal areas using the formula: ; The process involves obtaining a binary mask image of the abnormal region, identifying and extracting the coordinate set of the boundary of the connected region, and establishing a coordinate set for the abnormal temperature change layer; among which... Indicates position The abnormal marker value, Indicates the position in the original image frame grayscale value, for Neighborhood window set This represents the total number of pixels in the neighborhood. The thermal difference threshold; Based on the image number field and grayscale matrix field in the regional image frame matrix set, the image data is accessed frame by frame. Neighborhood extraction is performed on the effective pixels in each frame, with a selected neighborhood window of 3×3. That is, for any pixel location... its neighborhood Calculate the average gray value of the neighborhood, including its eight adjacent pixels in the top, bottom, left, right, and diagonal directions. Then compared with the grayscale value of the target pixel The difference is calculated as follows: For example, if a pixel is located at (100, 120) and has a grayscale value of 190, and the grayscale values ​​of its 8 neighboring pixels are 180, 185, 182, 190, 188, 184, 186, and 183 respectively, then the average value of the neighboring pixels is... Difference The calculated differences between all pixels are recorded in a difference matrix. Then, a thermal difference threshold is set. For each The process involves determining if a point's value is greater than 35. If so, the point is considered an anomaly, and the corresponding position in the mask is assigned a value of 1. Otherwise, it is assigned a value of 0. This process is repeated for all image frames to obtain a binary mask of the same size. Connected region identification is then performed on the set of all points with a value of 1 in the mask. The coordinates of the edge contours of the connected regions are aggregated, and their layer coordinates in the image are extracted. For example, in the image with the mask number IMG_0003, the distribution of its anomaly points forms three connected regions. The boundary coordinates of the first region are [(85, 93), (85, 94), (86, 94)]. This coordinate set is named Layer_A1. All image frames are then processed to establish the coordinate set of the abnormal temperature variation layer.

[0035] In this formula, the core computational logic is used to determine whether a pixel in a thermal infrared image belongs to an abnormal temperature change region, constructing a discrimination mechanism based on the intensity of the difference between the pixel and the average value of its neighborhood. The left side of the formula represents the anomaly marker value. Its value depends on the outcome of the conditional expression on the right. In the expression on the right, This represents the absolute difference between the grayscale value of a pixel and the average grayscale value of its neighboring pixels, where It is the grayscale value of the target pixel. It is its neighborhood set. This represents the total number of pixels in the neighborhood. It is calculated by summing the grayscale values ​​of the neighboring pixels and then dividing by the number of pixels. The mean value is obtained, reflecting the background thermal characteristics of the area. This mean value is then subtracted from the center pixel value to measure the local temperature fluctuation. Absolute value operations are used to eliminate the influence of positive and negative offsets on the judgment, ensuring that both heating and cooling are reflected as changes in intensity. This difference is compared with the thermal threshold. The comparison is used to establish the criteria for judging whether something is abnormal. If the value exceeds the threshold, it is determined that there is a significant temperature change at that point, and its value is set to 1; otherwise, it is 0. In summary, this formula constructs a spatial local temperature change feature recognition process through neighborhood averaging, absolute difference calculation, and threshold comparison, and logically realizes the effective determination of non-homogeneous temperature change points within a region.

[0036] An anomaly region binary mask is a two-dimensional image matrix generated based on each pixel in an image frame. It judges the intensity of the difference between the pixel's thermal radiation gray value and the average gray value of its neighborhood by whether the difference exceeds a set thermal sensitivity threshold. Essentially, it marks and abstracts the spatial location of potential temperature anomaly regions in the original thermal infrared image. Each pixel in the image corresponds to a Boolean value: a value of 1 indicates that the location is identified as an anomalous temperature change point, and a value of 0 indicates that the location is in a thermally stable state and does not belong to the anomalous response region. This mask removes the complex thermal intensity values ​​from the original image and only retains the binary anomaly judgment result. It has the characteristics of data simplification, clear boundaries, and clusterable structure, which facilitates subsequent operations such as connected component aggregation, contour extraction, and layer coordinate mapping. It is the basic data structure and logical carrier for extracting spatial temperature change feature regions from the original thermal data.

[0037] S413: Based on each coordinate pair in the coordinate set of the abnormal temperature change layer, match the device number and timestamp index embedded in the original image frame, map each set of coordinates back to the corresponding device and acquisition time, convert it into a heat source location vector with spatiotemporal attributes, and submit it to the cloud platform spatial database to establish a high temperature heat source coordinate sequence. Based on the coordinate index field of each layer marker in the abnormal temperature variation layer coordinate set, and by connecting to the acquisition device number field and timestamp field in the original image frame data source, the source device number and acquisition time are obtained through image number mapping. For example, the layer marker Layer_A1 corresponds to frame IMG_0003. The metadata of this frame records the device number as UAV-02 and the acquisition time as 2024-10-22 12:00:04. The coordinate set of this layer is bound to the metadata to form a triple structure {device number, acquisition time, coordinate set}. The structure is encoded in JSON format and is written in batches, with no less than 10 structures written each time. Each structure is bound to a unique linkage task number, such as LNK-002. All written structures are uniformly encoded and uploaded to the cloud platform's spatial database module via an HTTP interface. The target table is heat_source_coord_table. After the upload is completed, the interface return value is verified. If the status is 200, the structure is transferred to the task record table. The record format is: {task number, write time, number of data entries}, to complete the establishment of the time-series record and finally establish the high-temperature heat source coordinate sequence.

[0038] Please see Figure 6 The specific steps of S5 are as follows: S511: Based on the coordinate points in the high-temperature heat source coordinate sequence, read the latitude and longitude values ​​point by point, and perform spatial matching with the coordinate set of the node center in the monitoring trigger node sequence. The comparison method adopts the Euclidean distance judgment rule, and the matching threshold is set to 0.02 degrees for any heat source point. With the node center Calculate distance ,like If the heat source point is already registered in the node sequence, it is considered to be a registered area point; otherwise, it is marked as an unregistered area point, and a comparison matrix of registered nodes is generated. Based on the coordinate point structure in the high-temperature heat source coordinate sequence, each coordinate pair is extracted. and the coordinates of all center points in the monitoring trigger node sequence. Perform spatial comparison operations, specifically using Euclidean distance calculation, and perform distance judgment operations using the following comparison formula: By comparing each coordinate pair with the center points of all nodes sequentially, if a node's distance is less than or equal to a set matching threshold of 0.02 degrees, it is marked as a match; otherwise, it is marked as a miss. During execution, values ​​need to be extracted and distances calculated point by point. For example, if the heat source point is (108.31, 32.21) and a node is (108.30, 32.20), then the calculated distance is... Therefore, the point is successfully matched and assigned a value of 1. All matching results are encoded as 0 or 1 to form a matching marker matrix. At the same time, the original index of the participating point is recorded in each comparison pair to support subsequent mapping operations. All comparison results are stored in a two-dimensional matrix, where the rows correspond to the heat source coordinates and the columns correspond to the node numbers. Finally, a registered node comparison matrix is ​​generated.

[0039] S512: Based on the comparison matrix of registered nodes, filter the registered nodes, establish task prevention and control identifiers for the node data, and add a task number prefix to form a unique prevention and control task number. Store all task identifier records in the task buffer to be executed to form a node prevention and control task list and generate a prevention and control task identifier set. Based on the marker field in the registered node comparison matrix, a filtering operation is performed on the node numbers corresponding to heat source points with a value of 1 in the matrix. A prevention and control task identifier is generated for all filtered node numbers. The identifier structure is given the prefix code FIRE- and concatenated with the node number to form a task number. The task number is a string; for example, if the node number is N004, the task number is FIRE-N004. Each task number is bound to the current server time, with the timestamp format set to yyyy-MM-ddHH:mm:ss, for example, 2024-10-22. At 15:48:06, this information, along with the task number and node number, is used to construct a structure record. The structure format is {task number, node number, timestamp}. All structures are stored in a temporary task buffer stack for use by the task batch processing module. In addition, a uniqueness check is performed on all task structures to ensure that the task number is not repeated, the number field conforms to the registration format, and the timestamp format is compliant. After the check passes, the task is appended to the bottom of the task list. Each time a record is appended, the task counter is refreshed to construct an ordered task set with a counter number, and finally, a prevention and control task identifier set is generated.

[0040] S513: Based on the prevention and control task identifier set, all record structures are arranged in ascending order by timestamp and written in batches to the cloud platform fire prevention task execution queue. Integrity verification is performed on each task structure. At the same time, a task receipt identifier is generated in the cloud and a confirmation status code is returned to obtain the mountain fire monitoring and fire prevention instruction group. Based on all task structure entries in the prevention and control task identifier set, the fields of task number, node number, and timestamp are extracted. The task numbers are de-duplicated and sorted in ascending order by timestamp. The sorted task structures are then encapsulated into a JSON array format and connected to the cloud platform task submission interface. The interface performs a field integrity check to ensure that each structure contains three non-empty fields of type string, string, and timestamp. If all checks pass, a write operation is performed, writing the task data into the task execution queue list structure task_fire_queue. An interface receipt status code 200 is generated for all successfully written records, with the receipt content including the number of records written and the task start and end times. All successfully submitted data structures are redundantly backed up locally and merged into a unified scheduling object structure. The structure is assigned a scheduling identifier and submitted to the scheduling center for secondary scheduling operations. At the same time, the scheduling structure is written in JSON format to the output cache for subsequent instruction issuance interface loading, ultimately resulting in the mountain fire monitoring and fire prevention instruction group.

[0041] Please see Figure 7 A cloud-based mountain fire monitoring and prevention system includes: The factor label generation module is used to execute S1: acquire GIS 3D topographic raster layers, surface slope information, meteorological wind speed sequence and temperature and humidity monitoring data of mountainous areas, generate labels for each topographic grid, nest and map the temperature and humidity data, and upload them to the cloud to generate a set of topographic meteorological factor labels; The node sequence extraction module is used to execute S2: based on the terrain and meteorological factor label set, filter the label areas and extract the center coordinate points, establish the node set of each area, and upload it to the cloud platform task node interface after establishing a unique number identifier to generate the monitoring trigger node sequence; The equipment task allocation module is used to execute S3: based on the monitoring trigger node sequence, obtain the UAV number and ground sensor device number in the area, determine the relationship between the current operating status and the number position, combine and assign them and submit them to the queue to be executed, and obtain the air-ground device linkage number table; The heat source area identification module is used to execute S4: Based on the air-ground equipment linkage number table, start the equipment corresponding to the number and collect regional image frames, calculate the difference between the single pixel thermal radiation gray value in the image frame and the average value of adjacent pixels in the same frame and determine whether it is greater than the thermal sensitivity difference threshold, mark the abnormal temperature change feature area, extract the corresponding layer coordinates and upload them to the cloud platform to obtain the high temperature heat source coordinate sequence. The fire prevention instruction issuance module is used to execute S5: based on the high-temperature heat source coordinate sequence, verify whether the coordinate points in the sequence already exist in the monitoring trigger node sequence, submit all tasks to the cloud platform fire prevention task execution queue and mark the timestamp, and output the mountain fire monitoring fire prevention instruction group.

[0042] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A cloud-based method for monitoring and preventing mountain fires, characterized in that, Includes the following steps: S1: Obtain GIS 3D topographic raster layer, surface slope information, meteorological wind speed sequence and temperature and humidity monitoring data of mountainous areas, generate labels for each topographic grid, and nest and map the temperature and humidity data, and upload it to the cloud to generate a set of topographic meteorological factor labels; S2: Based on the terrain and meteorological factor tag set, filter the tag areas and extract the center coordinate points, establish a node set for each area, and upload it to the cloud platform task node interface after establishing a unique number identifier to generate a monitoring trigger node sequence; S3: Based on the monitoring trigger node sequence, obtain the UAV number and ground sensor device number in the area, determine the relationship between the current operating status and the number position, combine and assign them and submit them to the execution queue, and obtain the air-ground device linkage number table. S4: Based on the air-ground equipment linkage number table, start the equipment corresponding to the number and collect regional image frames, calculate the difference between the single pixel thermal radiation gray value in the image frame and the average value of adjacent pixels in the same frame and determine whether it is greater than the thermal sensitivity difference threshold, mark the abnormal temperature change feature area, extract the corresponding layer coordinates and upload them to the cloud platform to obtain the high temperature heat source coordinate sequence. S5: Based on the high-temperature heat source coordinate sequence, verify whether the coordinate points in the sequence already exist in the monitoring trigger node sequence, submit all tasks to the cloud platform fire prevention task execution queue and mark the timestamp, and output the mountain fire monitoring and fire prevention instruction group.

2. The mountain fire monitoring and prevention method based on a cloud platform according to claim 1, characterized in that, The terrain and meteorological factor label set includes terrain label categories, time-series nested meteorological factors, geographic grid index numbers, and cloud storage paths. The monitoring trigger node sequence includes node spatial coordinates, unique identifiers, node-assigned areas, and trigger priorities. The air-ground equipment linkage number table includes a set of statuses corresponding to equipment numbers, a set of coordinates corresponding to equipment numbers, combined assignment records, and equipment type identifiers. The high-temperature heat source coordinate sequence includes a set of abnormal heat source coordinate points, thermal grayscale differences, layer mapping indexes, and upload timestamps. The mountain fire monitoring and fire prevention instruction group includes task node numbers, prevention and control task identifiers, fire prevention task timestamps, and task execution priorities.

3. The mountain fire monitoring and prevention method based on a cloud platform according to claim 1, characterized in that, The steps for obtaining the topographic and meteorological factor tag set are as follows: S111: Obtain the 3D topographic raster layer and surface slope information of the mountainous area, extract the slope value of each raster point in the topographic raster layer, combine it with the wind speed value of the corresponding geographical location in the meteorological wind speed sequence, and perform index matching of the slope value and wind speed value of each raster point. By performing two-parameter grouping and statistical analysis on each slope value and its corresponding wind speed value, a topographic wind speed label matrix is ​​generated. S112: Based on the terrain and wind speed label matrix, extract the time series from the temperature and humidity monitoring data, perform coordinate mapping on the geographical coordinate points recorded in each time series data frame, and generate a nested meteorological factor matrix. S113: Based on the nested meteorological factor matrix, and combined with the label index of each grid point in the terrain wind speed label matrix, all meteorological factor data frames under the same label are sequentially aggregated. The aggregated data of each label is standardized and encoded, and the encoding results are uploaded to the cloud in one-to-one correspondence with the label index to generate a terrain meteorological factor label set.

4. The mountain fire monitoring and prevention method based on a cloud platform according to claim 3, characterized in that, The coordinate mapping specifically involves matching geographic coordinates with the location index of grid points in the terrain raster layer, while simultaneously nesting and mapping the corresponding temperature and humidity values ​​to the corresponding grid point locations.

5. The mountain fire monitoring and prevention method based on a cloud platform according to claim 1, characterized in that, The steps for obtaining the monitoring trigger node sequence are as follows: S211: Based on the topographic and meteorological factor tag set, establish a tag area retrieval according to the tag index field, perform grouping and clustering processing on all tag categories in the tag set, calculate the maximum and minimum values ​​of the horizontal and vertical coordinates of the grid points in two-dimensional space according to the spatial index data of the grid points in the tag group, and generate a tag area center point coordinate set by extracting the coordinates of the geometric boundary center point of each group of grid areas as the representative location point of the corresponding area. S212: Based on the coordinate set of the center points of the tag area, perform clustering according to the tag number to which each group of center points belongs, merge each group of center points into the node set indicated by the corresponding number, perform spatial index serialization processing on each node set, and set an independent number as a unique identifier for each node set to obtain the spatial node number set. S213: Based on the set of spatial node numbers, the node numbers and node coordinates in the set are combined and encapsulated into a node index structure, and written into the cloud platform task node interface in the form of a structure. Remote calls to node metadata are executed and mapped and mounted to establish a monitoring trigger node sequence.

6. The mountain fire monitoring and prevention method based on a cloud platform according to claim 1, characterized in that, The steps for obtaining the air-to-ground equipment linkage number table are as follows: S311: Based on the monitoring trigger node sequence, read the spatial coordinates in the index structure of each node, retrieve the UAV number and ground sensor device number in the matching area, collect the running status information and coordinate positioning information at the current timestamp, associate and bind each device number with its status data, and obtain the device running status mapping table. S312: Based on the device operation status mapping table, aggregate the UAV numbers and ground sensor numbers in the area according to the spatial distribution of nodes, combine the device numbers in each node into a task unit, filter the deployable devices according to the current operation status, record the status flags corresponding to each number, and submit the combined device numbers to establish a device deployment combination matrix. S313: Based on the equipment allocation and combination matrix, the equipment numbers of each group are serially encoded to construct an executable linkage task index number. The numbering standard verification and uniqueness detection are performed on each group of linkage numbers. After summarizing, the results are written into the task execution queue of the cloud platform, and the task status is set to generate an air-ground equipment linkage number table.

7. The mountain fire monitoring and prevention method based on a cloud platform according to claim 1, characterized in that, The steps for obtaining the high-temperature heat source coordinate sequence are as follows: S411: Based on the air-ground equipment linkage number table, extract the set of UAV number and ground equipment number, and send a start command to trigger the image acquisition function. Perform synchronous image frame capture operation in the task area, perform pixel-by-pixel decoding on each image frame, convert the thermal radiation gray value of each pixel into a two-dimensional matrix structure, and generate a regional image frame matrix set. S412: Based on the thermal radiation grayscale matrix of each frame image in the regional image frame matrix set, perform neighborhood window analysis on all pixels, extract the average grayscale value of adjacent pixels, construct a pixel difference matrix, perform item-by-item thermal threshold judgment on the difference matrix and mark abnormal regions, calculate and obtain the binary mask image of the abnormal region, identify and extract the set of boundary coordinates of connected regions, and establish the coordinate set of the abnormal temperature change layer. S413: Based on each coordinate pair in the coordinate set of the abnormal temperature change layer, match the device number and timestamp index embedded in the original image frame, map each set of coordinates back to the corresponding device and acquisition time, convert it into a heat source location vector with spatiotemporal attributes, and submit it to the cloud platform spatial database to establish a high-temperature heat source coordinate sequence.

8. The mountain fire monitoring and prevention method based on a cloud platform according to claim 7, characterized in that, During the image frame capture operation, the acquired image frame is a thermal infrared image with a grayscale range of 0 to 255.

9. The mountain fire monitoring and prevention method based on a cloud platform according to claim 1, characterized in that, The steps for obtaining the mountain fire monitoring and fire prevention command group are as follows: S511: Based on the coordinate points in the high-temperature heat source coordinate sequence, read the latitude and longitude values ​​point by point, and perform spatial matching with the node center coordinate set in the monitoring trigger node sequence to mark the registered nodes and unregistered area points, and generate a registered node comparison matrix. S512: Based on the registered node comparison matrix, filter the registered nodes, establish task prevention and control identifiers for the node data, and attach a task number prefix to form a unique prevention and control task number. Store all task identifier records in the task buffer to be executed to form a node prevention and control task list and generate a prevention and control task identifier set. S513: Based on the aforementioned prevention and control task identifier set, arrange all record structures in ascending order by timestamp and write them in batches into the cloud platform fire prevention task execution queue. Perform integrity verification on each task structure, and generate a task receipt identifier in the cloud and return a confirmation status code to obtain the mountain fire monitoring and fire prevention instruction group.

10. A cloud-based mountain fire monitoring and prevention system, characterized in that, The system is used to implement the cloud-based mountain fire monitoring and prevention method according to any one of claims 1-9, the system comprising: The factor label generation module is used to execute S1: acquire GIS 3D topographic raster layers, surface slope information, meteorological wind speed sequence and temperature and humidity monitoring data of mountainous areas, generate labels for each topographic grid, nest and map the temperature and humidity data, and upload them to the cloud to generate a set of topographic meteorological factor labels; The node sequence extraction module is used to execute S2: based on the terrain and meteorological factor label set, filter the label areas and extract the center coordinate points, establish the node set of each area, and upload it to the cloud platform task node interface after establishing a unique number identifier to generate the monitoring trigger node sequence; The equipment task allocation module is used to execute S3: based on the monitoring trigger node sequence, obtain the UAV number and ground sensor device number in the area, determine the relationship between the current operating status and the number position, combine and assign them and submit them to the queue to be executed, and obtain the air-ground device linkage number table; The heat source area identification module is used to execute S4: based on the air-ground equipment linkage number table, start the equipment corresponding to the number and collect regional image frames, calculate the difference between the single pixel thermal radiation gray value in the image frame and the average value of adjacent pixels in the same frame and determine whether it is greater than the thermal sensitivity difference threshold, mark the abnormal temperature change feature area, extract the corresponding layer coordinates and upload them to the cloud platform to obtain the high temperature heat source coordinate sequence. The fire prevention instruction issuance module is used to execute S5: based on the high-temperature heat source coordinate sequence, verify whether the coordinate points in the sequence already exist in the monitoring trigger node sequence, submit all tasks to the cloud platform fire prevention task execution queue and mark the timestamp, and output the mountain fire monitoring fire prevention instruction group.

Images