A watershed emergency drone management method and platform

Abstract

The present invention provides a watershed emergency drone management method and platform, relating to the technical field of drone management. The method comprises: determining a monitoring area and monitoring points based on the location of a hydropower station and historical geological disasters; developing a flight route based on the monitoring area and monitoring points with the goal of minimizing time cost, and generating a flight mission; acquiring flight data during the execution of the mission, the flight data comprising video data and corresponding pose data; preprocessing the video data to obtain a target image, extracting target image feature points, and updating a three-dimensional map based on the pose data; and displaying the updated three-dimensional map. The present invention can automatically plan drone flight routes, and using the data collected by the drone, it can quickly and accurately construct a new three-dimensional map and update the historical three-dimensional map, making the three-dimensional map more comprehensive, detailed, and accurate during use.

Description

Technical Field

[0001] The present invention relates to the technical field of drone management, and in particular to a watershed emergency drone management method and platform. Background Art

[0002] my country boasts vast land and numerous rivers, and its abundant water resources are crucial to our economic development. These resources can be used for hydropower generation, agricultural production, transportation, and other purposes. However, the complex terrain of river basins, with river channels and reservoirs often flanked by high mountains and canyons, makes mudslides and other geological disasters common. If a large-scale landslide occurs within a river channel or reservoir, the debris can block the river channel and easily form a barrier lake, significantly impacting the power generation of hydropower stations and the safety of surrounding residents.

[0003] In recent years, drone technology has developed rapidly. Due to its flexibility, speed, and freedom from terrain and environmental constraints, it has gradually been used in river basin surveying and three-dimensional modeling. However, existing drone management methods for river basins cannot meet the requirements for modeling speed and accuracy in emergency rescue operations such as geological disasters.

[0004] Therefore, it is urgent to develop a watershed emergency drone management method and platform. Summary of the Invention

[0005] In response to the above-mentioned problems in the prior art, the present invention provides a watershed emergency drone management method and platform, which can meet the requirements for modeling speed and accuracy in emergency rescue.

[0006] In order to achieve the above-mentioned object of the invention, the technical solution adopted by the present invention is as follows:

[0007] In one aspect, a watershed emergency drone management method includes:

[0008] Determine the monitoring area and monitoring points based on the location of the hydropower station and historical geological disasters, formulate a flight route with the minimum time cost as the goal, and generate a flight mission taking into account the monitoring area and monitoring points;

[0009] Acquiring flight data during the execution of a flight mission, wherein the flight data includes video data and corresponding posture data;

[0010] Preprocessing the video data to obtain a target image, extracting feature points of the target image, and updating the three-dimensional map in combination with the pose data;

[0011] Display the updated 3D map.

[0012] Preferably, the monitoring area and monitoring points are determined based on the location of the hydropower station and historical geological disasters, including:

[0013] Obtain high-altitude remote sensing images, mark the location of hydropower stations and the location and severity of historical geological disasters on the high-altitude remote sensing images, and determine the monitoring areas and the importance of each monitoring area;

[0014] According to the different terrains of each monitoring area, determine whether there are terrain feature points;

[0015] If it exists, the terrain feature point is used as a monitoring point; if it does not exist, the monitoring points are evenly distributed in the monitoring area according to the importance corresponding to the monitoring area.

[0016] Preferably, considering the monitoring area and monitoring points, a flight route is formulated with the minimum time cost as the goal, and a flight mission is generated, including:

[0017] Formulate basic flight information, including mission name, cycle, number of executions, single mission execution time, and mission deadline;

[0018] A flight route is developed based on the monitoring area, monitoring points and basic flight information.

[0019] Preferably, a flight route is formulated based on the monitoring area, monitoring points and basic flight information, including:

[0020] Construct an empty route, scan the monitoring points with the UAV control platform as the ray point, add the scanned monitoring points to the empty route, and construct a new empty route when the number of added monitoring points reaches a preset number;

[0021] Traverse all monitoring points, add all monitoring points to the route, and obtain the initial flight route;

[0022] Determining a cost function using the initial flight path as an optimization target, and iterating the cost function using a global optimal brainstorming algorithm to determine an optimal flight path within the maximum number of iterations;

[0023] Based on the greedy algorithm and with time cost as the constraint, two or more optimal flight routes are spliced ​​together to obtain the flight route.

[0024] Preferably, obtaining flight data during the execution of the flight mission includes:

[0025] Acquire multiple video streams sent by the drone, where the multiple video streams are encrypted video data collected by the drone at different angles in the same scene, and the time difference between the initial frame collection time of each video stream meets the preset time requirement;

[0026] The multiple video streams are parsed and converted into a specified video format to obtain video data.

[0027] Preferably, preprocessing the video data to obtain the target image includes:

[0028] Extracting multiple frames of images from the video data;

[0029] Performing wavelet decomposition on the image to obtain a low-frequency image and three high-frequency images;

[0030] Adaptive Gamma correction is performed on the low-frequency image to obtain a low-frequency image with enhanced contrast;

[0031] Denoising the three high-frequency images respectively to obtain denoised high-frequency images;

[0032] Performing inverse wavelet transform on the obtained contrast-enhanced low-frequency image and denoised high-frequency image to obtain an enhanced image;

[0033] Select a suitable correction model for the enhanced image to perform image registration;

[0034] Perform error check on image registration results;

[0035] When the error meets the requirement, the registered image is obtained and normalized to obtain the target image.

[0036] Preferably, extracting target image feature points and combining them with pose data to update the three-dimensional map includes:

[0037] Perform similarity matching on multiple target images based on the extracted target image feature points to obtain several matching pairs;

[0038] Performing feature point matching on all matching pairs using a feature matching algorithm, and obtaining the pose parameters of the target image using a global motion recovery algorithm;

[0039] Correcting the pose data according to the position parameters to obtain target pose data;

[0040] generating a three-dimensional terrain according to the target posture data and the target image, and performing texture mapping on the three-dimensional terrain to obtain a new three-dimensional map;

[0041] Overlaying the new three-dimensional map onto the corresponding area on the historical three-dimensional map to update the historical three-dimensional map;

[0042] Obtain an updated three-dimensional map and mark the position and trajectory of the drone on the updated three-dimensional map.

[0043] On the other hand, a watershed emergency drone management platform includes:

[0044] The mission planning module is used to determine the monitoring area and monitoring points based on the location of the hydropower station and historical geological disasters, formulate a flight route with the minimum time cost as the goal, and generate a flight mission taking into account the monitoring area and monitoring points;

[0045] A data acquisition module is used to acquire flight data during the execution of the flight mission, wherein the flight data includes video data and corresponding posture data;

[0046] A data processing module is used to pre-process the video data to obtain a target image, extract feature points of the target image, and update the three-dimensional map in combination with the posture data;

[0047] The map display module is used to display the updated three-dimensional map.

[0048] The beneficial effects of the present invention are:

[0049] The present invention provides a watershed emergency drone management method and platform, which can evaluate the location of hydropower stations and historical geological disasters, determine the monitoring area and its corresponding monitoring points, and construct a flight route based on the characteristics of the monitoring points and the characteristics of the drone, generate and execute flight missions, and quickly complete inspections of all monitoring points to ensure that no points are missed, making the collected data more accurate. The data format can be quickly encrypted and transmitted over the network through mature technology. The collected data can quickly and accurately construct a new three-dimensional map and update the historical three-dimensional map, making the three-dimensional map more comprehensive, detailed and accurate during use. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] Figure 1 A flow chart of a watershed emergency drone management method provided by an embodiment of the present invention;

[0051] Figure 2 This is a flowchart of a watershed emergency drone management method provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0052] The specific embodiments of the present invention are described below to facilitate understanding of the present invention by those skilled in the art. However, it should be clear that the present invention is not limited to the scope of the specific embodiments. For those skilled in the art, as long as various changes are within the spirit and scope of the present invention as defined and determined by the appended claims, these changes are obvious, and all inventions and creations utilizing the concepts of the present invention are protected.

[0053] Example 1

[0054] like Figure 1 As shown, Figure 1 A flow chart of a watershed emergency drone management method provided by an embodiment of the present invention, the method comprising:

[0055] Step 1: Determine the monitoring area and monitoring points based on the location of the hydropower station and historical geological disasters, formulate a flight route with the minimum time cost as the goal, and generate a flight mission taking into account the monitoring area and monitoring points;

[0056] In an embodiment of the present invention, the monitoring area and monitoring points are determined based on the location of the hydropower station and the historical geological disaster situation, including: obtaining a high-altitude remote sensing image, marking the location of the hydropower station and the location of the historical geological disaster and its severity on the high-altitude remote sensing image, and determining the monitoring area and the importance of each monitoring area; based on the different terrain of each monitoring area, determining whether there are terrain feature points; if so, using the terrain feature points as monitoring points; if not, evenly distributing monitoring points in the monitoring area according to the importance corresponding to the monitoring area.

[0057] Monitoring points should be geographically significant, and their selection should adhere to the following principles: To ensure adequate monitoring accuracy in all parts of each monitoring area, monitoring points should be roughly evenly distributed across the map (though a completely even grid-like distribution is not required). Furthermore, monitoring points should be located at the edges and corners of the monitoring area. Depending on the actual rendering of the three-dimensional map, the following adjustments can be made: In areas where geological disasters frequently occur, the number of monitoring points can be appropriately increased; in areas where geological disasters are less common, the number of monitoring points can be appropriately reduced. Furthermore, monitoring points should be appropriately increased at the boundaries between different monitoring areas.

[0058] Specifically, high-altitude remote sensing images are obtained through drones; high-altitude remote sensing technology can detect larger areas and obtain data on the area at a macro level; ground conditions will not limit remote sensing technology. In some areas with harsh conditions such as deserts and swamps, remote sensing technology is used to replace humans in collecting and detecting important data. Therefore, the use of remote sensing technology within the flow area can actively understand the geological conditions and easily find the scope of geological disasters; at the same time, remote sensing technology can be used to dynamically monitor the changing climate and promptly remind people in areas prone to geological disasters to take preventive measures as soon as possible.

[0059] In an embodiment of the present invention, the monitoring area and monitoring points are taken into consideration, a flight route is formulated with the minimum time cost as the goal, and a flight mission is generated, including: formulating basic flight information, the basic flight information including the mission name, cycle period, number of executions, single mission execution time, and mission deadline; and formulating a flight route based on the monitoring area, monitoring points and basic flight information.

[0060] Furthermore, a flight route is formulated based on the monitoring area, monitoring points and basic flight information, including: constructing an empty route, scanning the monitoring points with the UAV control platform as the ray point, adding the scanned monitoring points to the empty route, and when the number of added monitoring points reaches a preset number, constructing a new empty route to traverse all monitoring points, adding all monitoring points to the route to obtain an initial flight route; determining a cost function using the initial flight path as the optimization target, and iterating the cost function using the global optimal brainstorming algorithm to determine the optimal flight path within the maximum number of iterations; based on a greedy algorithm, with time cost as a constraint, splicing two or more optimal flight routes to obtain a flight route.

[0061] Since there may be multiple drone control platforms in a watershed, when obtaining the initial flight route, the monitoring points need to be scanned with a fixed radius. When the scanned monitoring point overlaps with the flight route of another drone monitoring platform, there is no need to add the monitoring point to the flight route to avoid the same monitoring point being inspected multiple times and to avoid wasting inspection time.

[0062] Adding the scanned monitoring points to the empty route includes: adding route information such as the single-point acquisition time, single-point acquisition range, maximum flight curvature, maximum flight climb, and cumulative acquisition time of the monitoring points to the empty route.

[0063] Specifically, the cost function is as follows:

[0064]

[0065] Among them, c i represents the i-th category route information, λ i represents the weight of the i-th category route information.

[0066] Because only monitoring point information and the maximum monitoring point capacity of each flight route are considered when determining the optimal flight route, an embodiment of the present invention also proposes a greedy-based path splicing method that can splice multiple routes to meet the maximum single flight time of the drone, thereby improving drone utilization. Specifically, a route with a cumulative flight time less than the maximum flight time is selected from multiple optimal routes. Combinations of any two optimal routes are enumerated, and the flight time from the last monitoring point of the first optimal route to the first monitoring point of the second optimal route is calculated. The optimal flight route with the shortest flight time is then selected for splicing to arrive at the flight route.

[0067] The above method can reduce the number of times the UAV travels back and forth to the UAV control platform, and a single flight can perform multiple flight missions, which is beneficial to improving the utilization rate of the UAV, improving the efficiency of UAV data collection, and shortening the time for 3D map generation.

[0068] Step 2: Acquire flight data during the execution of the flight mission, wherein the flight data includes video data and corresponding posture data;

[0069] In an embodiment of the present invention, obtaining flight data during the execution of a flight mission includes: obtaining multiple video streams sent by a drone, wherein the multiple video streams are encrypted video data collected by the drone at different angles in the same scene, and the time difference between the initial frame collection time of each video stream meets a preset time requirement; parsing the multiple video streams, converting them into a specified video format, and obtaining video data.

[0070] It should also be noted that the drone uploads the collected video stream and the position data of the video stream, encrypts it according to a preset communication protocol, and sends it to the base station. Finally, it sends it to the server through the base station, and the server performs other processing such as parsing the multiple video streams. The preset communication protocol includes but is not limited to RTMP, RTSP, and HLS, and the base station includes but is not limited to a 5G base station.

[0071] By adopting the above method, the encrypted transmission of multi-channel video streams of UAVs is realized, ensuring the security and timeliness of data transmission.

[0072] Step 3: pre-process the video data to obtain a target image, extract feature points of the target image, and update the three-dimensional map in combination with the pose data;

[0073] In an embodiment of the present invention, preprocessing video data to obtain a target image includes: extracting multiple frames of images from the video data; performing wavelet decomposition on the images to obtain a low-frequency image and three high-frequency images; performing adaptive gamma correction on the low-frequency image to obtain a contrast-enhanced low-frequency image; performing denoising on the three high-frequency images to obtain denoised high-frequency images; performing inverse wavelet transform on the obtained contrast-enhanced low-frequency image and denoised high-frequency image to obtain an enhanced image; selecting a suitable correction model for the enhanced image to perform image registration; performing error checking on the image registration result; and when the error meets the requirement, obtaining a registered image and performing normalization on the registered image to obtain the target image.

[0074] Furthermore, the calculation formula for denoising is as follows:

[0075]

[0076] in, The high-frequency image obtained by denoising, P(X, Y) is the low-frequency image contaminated by noise, (x, y) is a pixel point of the high-frequency image obtained by denoising, (X, Y) is a pixel point of the low-frequency image contaminated by noise, a×b is a template in the form of a 4-neighborhood or 8-neighborhood, and Q is the area of ​​the template;

[0077] The calculation formula for the normalization process is as follows:

[0078]

[0079] Among them, r is the value before normalization, F (r) is the normalized grayscale value, and R is the registered image to be processed.

[0080] In an embodiment of the present invention, feature points of a target image are extracted and combined with the posture data to update a three-dimensional map, including: performing similarity matching on multiple target images based on the extracted feature points of the target image to obtain a number of matching pairs; performing feature point matching on all the matching pairs through a feature matching algorithm, and obtaining the posture parameters of the target image using a global motion recovery algorithm; correcting the posture data according to the map parameters to obtain target posture data; generating a three-dimensional terrain based on the target posture data and the target image, and performing texture mapping on the three-dimensional terrain to obtain a new three-dimensional map; superimposing the new three-dimensional map on the corresponding area on the historical three-dimensional map to update the historical three-dimensional map; obtaining an updated three-dimensional map, and marking the position and trajectory of the drone on the updated three-dimensional map.

[0081] The construction of a three-dimensional model requires not only two-dimensional images but also the parameters of the camera to obtain a depth image. However, existing camera parameters are generally obtained through sensor equipment mounted on drones, and the accuracy cannot be guaranteed. Therefore, an embodiment of the present invention provides a method for correcting posture data to improve the accuracy of three-dimensional maps.

[0082] By using the above method, the three-dimensional map of each monitoring area in the watershed can be updated in real time, and the mission execution status of the drone can be displayed on the map, including but not limited to the location, flight trajectory, etc.

[0083] Step 4: Display the updated 3D map.

[0084] In summary, the embodiment of the present invention provides a method for managing emergency drones in a watershed, which can evaluate the location of a hydropower station and historical geological disasters, determine the monitoring area and its corresponding monitoring points, and construct a flight route based on the characteristics of the monitoring points and the characteristics of the drone, generate and execute flight missions, and quickly complete inspections of all monitoring points to ensure that there are no omissions, making the collected data more accurate, and its data format can be quickly encrypted and transmitted over the network through mature technology. The collected data can quickly and accurately construct a new three-dimensional map and update the historical three-dimensional map, making the three-dimensional map more comprehensive, detailed and accurate during use.

[0085] Example 2

[0086] like Figure 2 As shown, Figure 2 A schematic diagram of the structure of a river basin emergency drone management platform provided by an embodiment of the present invention includes: a mission planning module for determining monitoring areas and monitoring points based on the location of the hydropower station and historical geological disasters, taking into account the monitoring areas and monitoring points, formulating flight routes with the goal of minimizing time cost, and generating flight missions;

[0087] The data acquisition module is used to obtain flight data during the execution of the flight mission, and the flight data includes video data and corresponding posture data; the data processing module is used to pre-process the video data to obtain the target image, extract the target image feature points, and update the three-dimensional map in combination with the posture data; the map display module is used to display the updated three-dimensional map.

[0088] It should be understood that the river basin emergency drone management platform provided in the embodiment of the present invention and the river basin emergency drone management method provided in the above embodiment are based on the same inventive concept. For more specific working principles of each module in the embodiment of the present invention, please refer to the above embodiment and will not be repeated in the embodiment of the present invention.

[0089] It will be understood by those skilled in the art that although preferred embodiments of the present invention have been described, further changes and modifications may be made to these embodiments once those skilled in the art are aware of the underlying inventive concepts. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present invention. Obviously, those skilled in the art may make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, the present invention is intended to include such changes and modifications as fall within the scope of the claims and equivalents of the present invention.

Claims

1. A method for managing emergency drones in a watershed, characterized in that: include: Determine the monitoring area and monitoring points based on the location of the hydropower station and historical geological disasters, formulate a flight route with the minimum time cost as the goal, and generate a flight mission taking into account the monitoring area and monitoring points; Acquiring flight data during the execution of a flight mission, wherein the flight data includes video data and corresponding posture data; Preprocessing the video data to obtain a target image, extracting feature points of the target image, and updating the three-dimensional map in combination with the pose data; Display the updated 3D map; The extraction of target image feature points and the combination of pose data to update the 3D map include: Perform similarity matching on multiple target images based on the extracted target image feature points to obtain several matching pairs; Performing feature point matching on all matching pairs using a feature matching algorithm, and obtaining the pose parameters of the target image using a global motion recovery algorithm; Correcting the posture data according to the posture parameters to obtain target posture data; generating a three-dimensional terrain according to the target posture data and the target image, and performing texture mapping on the three-dimensional terrain to obtain a new three-dimensional map; Overlaying the new three-dimensional map onto the corresponding area on the historical three-dimensional map to update the historical three-dimensional map; Obtain an updated three-dimensional map and mark the position and trajectory of the drone on the updated three-dimensional map.

2. A watershed emergency drone management method according to claim 1, characterized in that: Based on the location of the hydropower station and historical geological disasters, the monitoring area and monitoring points are determined, including: Obtain high-altitude remote sensing images, mark the location of hydropower stations and the location and severity of historical geological disasters on the high-altitude remote sensing images, and determine the monitoring areas and the importance of each monitoring area; According to the different terrains of each monitoring area, determine whether there are terrain feature points; If it exists, the terrain feature point is used as a monitoring point; if it does not exist, the monitoring points are evenly distributed in the monitoring area according to the importance corresponding to the monitoring area.

3. A method for managing river basin emergency drones according to claim 1, characterized in that: Considering the monitoring area and monitoring points, a flight route is formulated with the minimum time cost as the goal, and a flight mission is generated, including: Formulate basic flight information, including mission name, cycle, number of executions, single mission execution time, and mission deadline; A flight route is developed based on the monitoring area, monitoring points and basic flight information.

4. A method for managing emergency drones in a watershed according to claim 3, characterized in that: Develop a flight route based on the monitoring area, monitoring points, and basic flight information, including: Construct an empty route, scan the monitoring points with the UAV control platform as the ray point, add the scanned monitoring points to the empty route, and construct a new empty route when the number of added monitoring points reaches a preset number; Traverse all monitoring points, add all monitoring points to the route, and obtain the initial flight route; Determining a cost function using the initial flight path as an optimization target, and iterating the cost function using a global optimal brainstorming algorithm to determine an optimal flight path within a maximum number of iterations; Based on the greedy algorithm and with time cost as the constraint, two or more optimal flight routes are spliced ​​together to obtain the flight route.

5. A watershed emergency drone management method according to claim 1, characterized in that: Obtain flight data during mission execution, including: Acquire multiple video streams sent by the drone, where the multiple video streams are encrypted video data collected by the drone at different angles in the same scene, and the time difference between the initial frame collection time of each video stream meets the preset time requirement; The multiple video streams are parsed and converted into a specified video format to obtain video data.

6. A method for managing emergency drones in a watershed according to claim 1, characterized in that: Preprocess the video data to obtain the target image including: Extracting multiple frames of images from the video data; Performing wavelet decomposition on the image to obtain a low-frequency image and three high-frequency images; Adaptive Gamma correction is performed on the low-frequency image to obtain a low-frequency image with enhanced contrast; Denoising the three high-frequency images respectively to obtain denoised high-frequency images; Performing inverse wavelet transform on the obtained contrast-enhanced low-frequency image and denoised high-frequency image to obtain an enhanced image; Select a suitable correction model for the enhanced image to perform image registration; Perform error check on image registration results; When the error meets the requirement, the registered image is obtained and normalized to obtain the target image.

7. A river basin emergency drone management platform, applicable to the method according to any one of claims 1 to 6, characterized in that: include: The mission planning module is used to determine the monitoring area and monitoring points based on the location of the hydropower station and historical geological disasters, formulate a flight route with the minimum time cost as the goal, and generate a flight mission taking into account the monitoring area and monitoring points; A data acquisition module is used to acquire flight data during the execution of the flight mission, wherein the flight data includes video data and corresponding posture data; A data processing module is used to pre-process the video data to obtain a target image, extract feature points of the target image, and update the three-dimensional map in combination with the posture data; The map display module is used to display the updated three-dimensional map.

Images