A land natural disaster early warning management and control system based on unmanned aerial vehicle surveying and mapping

Abstract

This invention discloses a land natural disaster early warning and control system based on UAV mapping, belonging to the field of land resource early warning management technology. It includes a UAV swarm and a local control terminal for UAV swarm control and data analysis and processing. The local control terminal includes: a mapping planning module for planning the distribution and flight paths of UAVs within the mapping area; a data processing module for identifying, marking, and recording the locations of disaster hazard points in various terrains based on image data acquired by the UAVs and radar monitoring data; and an analysis and early warning module for conducting fixed-point inspections of the hazard-marked areas at different time intervals, calculating the migration speed and migration value of two sets of monitoring data, and conducting risk assessment and early warning based on the comparison of migration speed and migration value. This system efficiently and comprehensively acquires and processes data, avoiding losses caused by untimely and inadequate responses due to the inability to assess the safety of personnel and property in the risk warning area.

Description

Technical Field

[0001] This invention relates to the field of land resource early warning management technology, specifically a land natural disaster early warning and control system based on UAV mapping. Background Technology

[0002] Natural disasters are often intricately linked to environmental damage, requiring further in-depth research into the patterns of natural disaster occurrence and evolution, as well as the mechanisms of disaster formation. This necessitates strengthening natural disaster monitoring, prediction, and early warning capabilities to reduce the losses caused by natural disasters.

[0003] Existing land natural disaster early warning and control systems rely heavily on manual inspections and data collection, resulting in low timeliness and accuracy of warnings, failing to meet practical needs. Using drones can expand the mapping range while ensuring flexibility and efficiency, enabling the identification of hidden hazards. However, drone mapping is severely affected by weather conditions, and existing early warning and control systems can only analyze and issue warnings based on individual data, making it difficult to establish correlations and learning between data points. In urgent situations, it is difficult to acquire, process, and analyze data in a timely manner, thus hindering the timely provision of early warning information. Therefore, we propose a land natural disaster early warning and control system based on drone mapping. Summary of the Invention

[0004] The purpose of this invention is to provide a land natural disaster early warning and control system based on unmanned aerial vehicle (UAV) mapping.

[0005] The technical problem solved by this invention is:

[0006] (1) How to divide the survey area and plan the flight path of the UAV according to different terrains, so as to solve the problem that the flight path of the UAV cannot match different terrains when collecting data, resulting in a decrease in the accuracy and globality of the data;

[0007] (2) How to perform data calculation and preprocess the acquired image data through the data processing module, and determine the anomalies from multiple dimensions to solve the problem of complex and inefficient image comparison;

[0008] (3) How to compare and analyze two or more data through the analysis and early warning module, calculate the migration amount and migration speed, and thus identify areas with excessively fast migration speed, thereby solving the problem that existing technologies cannot detect risk areas in a timely manner;

[0009] (4) How to assess the risk factor of the risk warning area through the analysis and early warning module, and take different risk response measures according to different assessment results, so as to solve the loss caused by the inability of the existing technology to assess the safety of personnel and property in the risk warning area, resulting in untimely and inadequate response;

[0010] (5) How to use the fast matching early warning module to directly perform correlation analysis and matching on the basis of a large amount of historical data, so as to quickly carry out early warning and control, and solve the problem of untimely early warning and control in emergency situations.

[0011] This invention can be achieved through the following technical solution: a land natural disaster early warning and control system based on UAV mapping, including a UAV swarm and a local control terminal for UAV swarm control and data analysis and processing, wherein the UAV swarm is equipped with an image acquisition device, acoustic radar and sensors;

[0012] The local control terminal includes:

[0013] The surveying and planning module plans the distribution and flight routes of drones within the surveying area;

[0014] The data processing module identifies, marks, and records the locations of potential disaster sites in various terrains based on image data acquired by the UAV and radar monitoring data.

[0015] The analysis and early warning module conducts fixed-point inspections of the hazard-marked areas at different time intervals, calculates the migration speed and migration value of the two monitoring data, and conducts risk assessment and early warning based on the comparison of migration speed and migration value.

[0016] A further technical improvement of the present invention is that the route planning steps of the surveying and planning module include:

[0017] Step 1: Delineate the survey area on the map and divide it into grids according to the terrain;

[0018] Step 2: Allocate the number of drones within the grid area based on the area occupied by each terrain feature;

[0019] Step 3: Develop different drone navigation strategies based on different terrain structures.

[0020] A further technical improvement of the present invention is that: when the data processing module performs data processing, it constructs a three-dimensional coordinate system with the drone's takeoff position as the origin, and uses this three-dimensional coordinate system to record the coordinates of the drone at the time of data acquisition, so that data can be acquired at the same coordinate position when monitoring again in the future, thereby reducing the difficulty of data adjustment and processing.

[0021] A further technical improvement of this invention lies in: identifying and marking potential disaster sites in different terrains, specifically including:

[0022] To address the potential landslide hazard in mountainous areas, data processing is performed from three directions: gully areas, bare vegetation areas, and steep slope areas. The processed data is then compared with a set threshold to determine the presence of a landslide hazard. Landslide hazard points are marked and their coordinates are recorded.

[0023] For plain areas, sonic radar is used to simulate the fault structure under the ground surface and calculate the risk index of landslides in the corresponding areas. Based on the risk index, landslide hazard points are marked and their coordinates are recorded.

[0024] For water flow and lakes, image data and water depth data at the water-land interface in the two areas are acquired respectively. The image data is matched with the abnormal sample image. When the matching degree exceeds the set value, the hazard is marked and the coordinates are recorded. At the same time, the distribution of silt in the riverbed or lakebed is simulated, and the hazard is marked and the coordinates are recorded.

[0025] A further technical improvement of the present invention is that the analysis and early warning module compares the data acquired from two or more adjacent locations at the hazard marker and calculates the migration value and migration speed. When the migration speed exceeds a certain range, the hazard marker is upgraded to a risk marker and the frequency of on-site monitoring by drones is increased.

[0026] A further technical improvement of the present invention is that the analysis and early warning module also predicts the migration amount of the corresponding data within a certain period of time by using the weather conditions in the future and the distribution of weather in the same period of previous years for correction. The migration amount is then compared with the migration amount limit threshold of the relevant data and a response strategy is formulated based on the comparison result.

[0027] A further technical improvement of the present invention is that when the predicted migration amount exceeds the migration amount limit threshold of the relevant data, the hidden danger mark is upgraded to a risk mark and the frequency of on-site monitoring by drones is increased; when a risk mark appears, the risk warning area is released to the outside world through the network and other media to remind people in the relevant area to take precautions.

[0028] A further technical improvement of this invention lies in that: the analysis and early warning module performs a hazard assessment in the risk warning area and its surrounding production activity areas, and the hazard assessment formula is:

[0029]

[0030] Where L represents the distance between the facility area and the residential area in the production activity area and the coordinates of the risk marker point; ρ represents the population density in the production activity area; V represents the value of the facilities in the production activity area; R represents the risk type impact index; i and j represent the numbers of the production facility area and the personnel living area, respectively; W represents the risk assessment coefficient, with a value range of [0,1].

[0031] A further technical improvement of this invention lies in: implementing different control strategies in real time based on the hazard assessment coefficient.

[0032] When the calculated risk assessment coefficient W is less than or equal to 0.3, the risk warning area will continue to be monitored.

[0033] When the calculated hazard assessment coefficient W is between 0.3 and 0.6, the area should be evacuated and isolated.

[0034] When the calculated risk assessment coefficient W is above 0.6, proactive human intervention is carried out. Proactive intervention measures include targeted blasting, landslide prevention steps, river dredging, and embankment reinforcement.

[0035] A further technical improvement of the present invention is that the system also includes a rapid matching and early warning module. Based on historical data, the rapid matching and early warning module decomposes the acquired image data and radar monitoring data, performs correlation analysis with the historical data, obtains the set of historical data with the highest correlation, and directly performs prediction and control measures based on the processing records and results of the historical data at that time.

[0036] Compared with the prior art, the present invention has the following beneficial effects:

[0037] 1. The surveying and mapping planning module divides the surveying area and plans the flight path of the UAV according to different terrains, so that the flight path of the UAV during data collection matches different terrains, thereby improving the accuracy and globality of the data acquisition process.

[0038] 2. The data processing module performs data calculations and preprocesses the acquired image data, and identifies anomalies from multiple dimensions. It quickly identifies potential hazard areas or anomalies, marks them accordingly, and records their coordinates. This avoids the computational burden caused by extensive image comparisons and data calculations, and greatly improves data processing efficiency.

[0039] 3. By analyzing and comparing data from two or more sources using the early warning module, the migration volume and speed are calculated. When the migration speed exceeds a certain range, the hazard marker is upgraded to a risk marker, and the frequency of on-site monitoring by drones is increased. This allows for the timely detection of risk warning areas and timely risk control to prevent major safety accidents. Simultaneously, a hazard coefficient assessment is conducted on the risk warning areas, and different hazard response measures are taken based on different assessment results to avoid losses caused by untimely or inadequate responses due to the inability to assess the safety of personnel and property in the risk warning areas.

[0040] 4. The rapid matching and early warning module can directly perform correlation analysis and matching based on a large amount of historical data, thereby enabling rapid early warning and control. This is beneficial for quickly issuing early warnings and generating control plans in emergency situations. Attached Figure Description

[0041] To facilitate understanding by those skilled in the art, the present invention will be further described below with reference to the accompanying drawings.

[0042] Figure 1 This is a system block diagram of the present invention. Detailed Implementation

[0043] To further illustrate the technical means and effects of the present invention in achieving its intended purpose, the following detailed description of the specific implementation methods, structures, features, and effects of the present invention, in conjunction with the accompanying drawings and preferred embodiments, is provided.

[0044] Please see Figure 1 As shown, a land natural disaster early warning and control system based on UAV mapping includes a UAV swarm, image acquisition devices, sensors, a local control terminal, and a wireless communication module. Each UAV in the UAV swarm is equipped with an image acquisition device and sensors to acquire images or monitor data within the required mapping area. The local control terminal includes a data processing module, an analysis and early warning module, a rapid matching and early warning module, and a storage module. The local control terminal and the UAV swarm are wirelessly connected via the wireless communication module. The local control terminal controls the UAV swarm and transmits data in real time through the wireless communication module.

[0045] The local control terminal also includes a surveying and planning module, which plans the distribution and flight paths of UAVs within the surveying area. Specific steps include:

[0046] Step 1: Obtain a satellite map from the internet and delineate the area to be surveyed on the satellite map. Within this delineated area, further divide it into grids according to the terrain, which includes mountains, plains, rivers, and lakes, etc.

[0047] Step 2: Calculate the area within the corresponding terrain grid and label it as S. i Where i represents the number of different terrain types, the number of grid areas for the above-mentioned different terrain types is allocated according to the distribution density of UAVs per unit area. For example, N UAVs are set up for mapping within an area of ​​10 square kilometers, so the base number of UAVs in each grid area is S. i N;

[0048] Due to the varying complexity of geological conditions in different terrains, the distribution density of drones needs to be adjusted. The adjustment factor ranges from 1 to 1.36, and the adjustment factor increases with the complexity of the geological conditions. The order of geological complexity is as follows: mountains, rivers, plains, and lakes.

[0049] Step 3: Within each grid area of ​​different terrains, plan the corresponding flight routes. For example, for mountainous terrain, data is acquired by gradually decreasing altitude along the ridge of the mountain; for plains terrain, data is acquired by radiating outwards from the center; for river terrain, data is acquired according to the direction of the river; and for lake terrain, data sampling points are randomly and evenly set in the water area according to the water boundary of the lake.

[0050] When the drone flies and collects data according to the above-mentioned route plan, it acquires images through an image acquisition device, which is a high-definition camera. At the same time, the drone is equipped with a sonic radar to detect distance data. The sonic radar can penetrate the covering above the ground and detect the fault structure below the ground.

[0051] These image and distance data are transmitted to the storage module of the local control terminal. The data processing module preprocesses the image data of various terrains, including data enhancement and noise reduction. The image data also includes the UAV coordinates at the time of image acquisition. The UAV coordinates are the current position coordinates of a three-dimensional coordinate system established with the UAV's takeoff position as the origin. Thus, all UAV coordinates can be established in the same three-dimensional coordinate system. The significance of recording coordinates in the image data is that the UAV can fly to the corresponding coordinates to take pictures in two image data acquisitions at different times. In this way, the image data acquired in two acquisitions can be directly compared without cropping and position correction, thereby reducing the difficulty of subsequent image data processing, improving timeliness, and having a positive significance for disaster early warning and control.

[0052] The data processing module identifies and marks potential disaster sites in various terrains based on image data and radar monitoring data, specifically including:

[0053] For mountainous terrain, data measurements are taken in the image data for gully areas, steep slope areas, and bare vegetation areas, such as gully depth and width, slope and height difference in steep slope areas, and bare area in bare vegetation areas.

[0054] When bare vegetation areas overlap with gully areas or steep slope areas, it is directly determined that there is a potential landslide hazard in that area.

[0055] The ratio of gully depth to gully width is calculated to obtain the gully depth-to-width ratio. When the depth-to-width ratio does not exceed a set threshold, it is considered that there is a potential landslide hazard at this location.

[0056] Based on the slope and height difference of the steep slope area, a curve of slope change in the height direction is established in a virtual plane rectangular coordinate system. The rate of change of the slope is obtained by differentiating the curve. The greater the rate of change, the more drastic the change of slope. This type of terrain is prone to landslides. When the rate of change is greater than a set threshold, it is considered that there is a landslide hazard.

[0057] When the area of ​​bare vegetation exceeds a certain value and the slope of the bare vegetation area exceeds a certain value, the corresponding area is considered to have a potential landslide risk.

[0058] The landslide hazards mentioned above are marked accordingly, and their corresponding coordinate positions are recorded. The coordinates can be directly obtained from the same three-dimensional coordinate system as the UAV coordinates. When the UAV acquires image data, since the location of the hazard marker is relatively fixed and the location of the UAV can be measured, the coordinate data of the hazard marker location is extracted from the same three-dimensional coordinate system as the UAV coordinates and recorded.

[0059] For plain areas, the propagation of the sonic radar bands beneath the ground surface is mainly used to detect the fault structure beneath the ground surface. Based on the wave band reflection received by the sonic radar, the surface layer structure under the UAV flight path is simulated in the data processing module. This allows it to identify whether there are faults under the ground surface under the corresponding flight path. If they exist, the fault structure can be simulated, including the fault area, fault depth, and the thickness of the entire fault.

[0060] The risk index of fault collapse is positively correlated with fault area and fault thickness, and inversely correlated with fault depth. Weighting coefficients are assigned to fault area, fault thickness, and fault depth as x, y, and z, respectively, with the coefficients of x, y, and z increasing sequentially.

[0061] Once the risk index is obtained, if the risk index is higher than the set value, the corresponding hidden danger markers and coordinate records are made for the area where the fault structure is located.

[0062] For rivers and lakes, high-definition imaging technology is combined with acoustic detection technology to acquire image data at the water-land interface in the two areas, while simultaneously detecting the landforms (riverbed structure or lakebed structure) under the water cover.

[0063] Specifically, the boundary between land and water in rivers mainly consists of riverbanks. For image data of riverbanks and dams, feature extraction, feature selection, and dimensionality reduction are performed. Extracted features include shape, color, and texture. Based on these features, a clustering algorithm is used to group the extracted image data, dividing the feature data points into multiple clusters according to the aforementioned division range. Outliers are identified in these clusters using the following methods:

[0064] A1: Calculate the size (i.e., the number of data points contained in each cluster) and density (i.e., the distribution of data points in the cluster) of each cluster;

[0065] A2: For each cluster, calculate the distance between data points within it, using Euclidean distance or other distance metrics.

[0066] A3: For each cluster, calculate the average distance and standard deviation of the data points within it, which can be used to measure the density or dispersion within the cluster.

[0067] A4: For each cluster, by comparing its size, density, average distance, standard deviation, and dispersion, it can be determined whether there are outliers in the cluster;

[0068] If a cluster has a small number of data points, a low density, a large average distance, and a large standard deviation and dispersion, then it is determined that there is abnormal data in that cluster. The above-mentioned size concepts are all determined by comparing with conventional thresholds.

[0069] Data within clusters containing anomalous data is extracted and automatically cross-matched with anomalous sample images in the storage module, and the matching degree is calculated. The matching degree is related to shape similarity, texture overlap, and color value difference. If the matching degree is higher than 70%, the dam area corresponding to the image data is identified as having hidden dangers, and the hidden dangers are marked and the coordinates are recorded.

[0070] Simultaneously, based on the reflection of sound waves by the riverbed or lakebed, a virtual model of the corresponding riverbed or lakebed can be constructed, thereby determining the distribution of silt on the riverbed or lakebed and obtaining the distance between the silt-corresponding area and the water surface. When the distance between the silt and the water surface is less than or equal to a set value, it is considered that there is a hidden danger in the corresponding water area, and the hidden danger is marked and the coordinates are recorded.

[0071] The coordinate system model marked with hazard markers and coordinate records is mapped and fitted onto the network satellite map. The analysis module performs UAV fixed-point patrols of the hazard-marked areas at different time intervals, compares the image data and radar monitoring data of the hazard-marked areas in two separate instances, and calculates the migration value between the two sets of data. The migration value is then adjusted to eliminate the influence of weather factors to calculate the migration speed. The calculation formula is as follows:

[0072]

[0073] in:

[0074] v represents the time T. a At time T b The data migration speed of the corresponding hazard-marked area within the time interval;

[0075] and They represent time T, respectively. a and time T b Values ​​of relevant data from drones for the same hazard-marked area;

[0076] α, β, and γ represent the influence coefficients on migration speed under three weather conditions (dry, wet, and rain), respectively, and α < β < γ. The values ​​of α, β, and γ are all between 0 and 1, excluding the endpoints.

[0077] t1, t2, and t3 represent the time intervals T and T, respectively. a and time T b Within the time interval, the number of days each of the three weather conditions occurred, and T b -T a = t1 + t2 + t3.

[0078] When the migration speed is faster than the normal migration speed, the hazard marker is upgraded to a risk marker and the frequency of on-site monitoring by drones is increased. The latest monitoring data is used for early warning and control. It should be noted that the normal migration speed is fast, which refers to the average migration speed of the corresponding terrain based on big data.

[0079] By obtaining local weather conditions over a certain period (one week, half a month, or one month) via the internet and adjusting the forecast based on the distribution of days with different weather types during the same period in previous years, the number of days with three different weather conditions is determined within the next certain period. Based on the migration speed formula and the number of days with the corresponding weather conditions, the migration amount of the corresponding data during this period can be predicted. A migration amount threshold is set for the local land environment and composition to match the actual local conditions. When the predicted migration amount exceeds the migration amount threshold, the hazard marker is upgraded to a risk marker, and the frequency of on-site monitoring by drones is increased. When a risk marker appears, a risk warning area is issued to the outside world through the internet and other media to remind people in the relevant areas to take appropriate precautions.

[0080] Within the risk warning area and its surrounding areas, a hazard assessment is conducted on areas where people engage in production activities, such as farmland, irrigation areas, and residential areas. The hazard includes the risk to production facilities and personnel. Relevant indices for the hazard assessment include the distance L between the facility area and residential area in the production activity area and the coordinates of the risk marker, the population density ρ in the production activity area, the value V of the facilities in the production activity area, and the risk type impact index R. These factors are then substituted into the hazard assessment formula:

[0081]

[0082] Among them, the risk type impact index R is related to the risk type, which includes landslides, land subsidence, river siltation, and dam breaches. In the above order, the risk type impact index gradually increases.

[0083] i and j represent the production facility area and the personnel living area, respectively. i and j are consecutive positive integers. k represents the distance influence coefficient, which ranges from 0.3 to 0.8.

[0084] The population density ρ represents the ratio of the number of people to the area of ​​a designated region;

[0085] W represents the risk assessment coefficient, with a value range of [0,1].

[0086] When the calculated risk assessment coefficient W is less than or equal to 0.3, follow-up monitoring can continue.

[0087] When the calculated hazard assessment coefficient W is between 0.3 and 0.6, the area should be evacuated and isolated.

[0088] When the calculated risk assessment coefficient W is above 0.6, active human intervention will be carried out. Active intervention measures include targeted blasting, landslide prevention steps, river dredging, and embankment reinforcement.

[0089] The rapid matching and early warning module, with a large amount of historical data accumulated in the storage module, breaks down the acquired image data and radar monitoring data and performs correlation analysis with the historical data to obtain the closest set of historical data. Based on the processing records and results of the historical data at that time, prediction and control measures can be directly implemented, which can save a lot of resources and time. This has an important reference role in handling emergencies and can promptly form a handling plan.

[0090] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the present invention. Any simple modifications, equivalent changes and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the scope of the present invention.

Claims

1. A land natural disaster early warning and control system based on unmanned aerial vehicle (UAV) mapping, comprising a UAV swarm and a local control terminal for UAV swarm control and data analysis and processing, characterized in that, The drone swarm is equipped with image acquisition devices, acoustic radar, and sensors; The local control terminal includes: The surveying and planning module plans the distribution and flight routes of drones within the surveying area; The data processing module identifies, marks, and records the locations of potential disaster sites in various terrains based on image data acquired by the UAV and radar monitoring data. The analysis and early warning module conducts fixed-point inspections of the hazard-marked areas at different time intervals, calculates the migration speed and migration value of the two monitoring data sets, and performs risk assessment and early warning based on the comparison of the migration speed and migration value. Specifically: The analysis and early warning module compares two or more adjacent data acquisitions at the hazard marker and calculates the migration value and migration speed. When the migration speed exceeds a certain range, the hazard marker is upgraded to a risk marker, and the frequency of on-site monitoring by drones is increased. The formula for calculating the migration speed is: ； in: Indicates at time At the time The data migration speed of the corresponding hazard-marked area within the time interval; and They represent the times respectively. and time Values ​​of relevant data from drones for the same hazard-marked area; , , These represent the influence coefficients on migration speed under three weather conditions: dry, humid, and rainy. < < , , , The values ​​of are all between 0 and 1, excluding endpoints; , and They represent the times respectively. and time Within the time interval, the number of days each of the three weather conditions occurred, and ; The analysis and early warning module also predicts the migration amount of corresponding data during the period by using the weather conditions in the future and correcting them with the distribution of weather in the same period of previous years. The migration amount is then compared with the migration amount limit threshold of relevant data, and a response strategy is formulated based on the comparison results. When the predicted migration volume exceeds the migration volume limit threshold of the relevant data, the hazard marker is upgraded to a risk marker and the frequency of on-site monitoring by drones is increased; when a risk marker appears, the risk warning area is released to the outside world through the Internet and other media to remind people in the relevant area to take precautions. The analysis and early warning module conducts a hazard assessment within the risk warning area and its surrounding production activity areas. The hazard assessment formula is as follows: ; Where L represents the distance between the facility area and the residential area in the production activity area and the coordinates of the risk marker point; ρ represents the population density in the production activity area; V represents the value of the facilities in the production activity area; R represents the risk type impact index; i and j represent the numbers of the production facility area and the personnel living area, respectively; W represents the risk assessment coefficient, with a value range of [0,1].

2. The land natural disaster early warning and control system based on UAV mapping according to claim 1, characterized in that, The surveying and planning module includes the following steps for route planning: Step 1: Delineate the survey area on the map and divide it into grids according to the terrain; Step 2: Allocate the number of drones within the grid area based on the area occupied by each terrain feature; Step 3: Develop different drone navigation strategies based on different terrain structures.

3. A land natural disaster early warning and control system based on UAV mapping according to claim 1, characterized in that, When processing data, the data processing module constructs a three-dimensional coordinate system with the drone's takeoff position as the origin, and uses this three-dimensional coordinate system to record the coordinates of the drone at the moment of data acquisition, so that data can be acquired at the same coordinate position during subsequent monitoring, reducing the difficulty of data adjustment and processing.

4. A land natural disaster early warning and control system based on UAV mapping according to claim 1, characterized in that, Identify and mark potential disaster sites in different terrains, specifically including: To address the potential landslide hazard in mountainous areas, data processing is performed from three directions: gully areas, bare vegetation areas, and steep slope areas. The processed data is then compared with a set threshold to determine the presence of a landslide hazard. Landslide hazard points are marked and their coordinates are recorded. For plain areas, sonic radar is used to simulate the fault structure under the ground surface and calculate the risk index of landslides in the corresponding areas. Based on the risk index, landslide hazard points are marked and their coordinates are recorded. For water flow and lakes, image data and water depth data at the water-land interface in the two areas are acquired respectively. The image data is matched with abnormal sample images. When the matching degree exceeds the set value, hazard marking and coordinate recording are performed. At the same time, the distribution of silt in the riverbed or lakebed is simulated, and hazard marking and coordinate recording are performed.

5. A land natural disaster early warning and control system based on UAV mapping according to claim 1, characterized in that, Different control strategies are implemented in real time based on the risk assessment coefficient: When the calculated risk assessment coefficient W is less than or equal to 0.3, the risk warning area will continue to be monitored. When the calculated hazard assessment coefficient W is between 0.3 and 0.6, the area should be evacuated and isolated. When the calculated risk assessment coefficient W is above 0.6, proactive human intervention is carried out. Proactive intervention measures include targeted blasting, landslide prevention steps, river dredging, and embankment reinforcement.

6. A land natural disaster early warning and control system based on UAV mapping according to claim 1, characterized in that, It also includes a rapid matching and early warning module. Based on historical data, the rapid matching and early warning module breaks down the acquired image data and radar monitoring data and performs correlation analysis with the historical data to obtain the set of historical data with the highest correlation. Based on the processing records and results of the historical data at that time, it directly makes predictions and implements control measures.

Images