Accident scene intelligent identification positioning and processing method and system based on unmanned aerial vehicle library

Abstract

The invention discloses an accident scene intelligent identification positioning and processing method and system based on an unmanned aerial vehicle library, and relates to the technical field of unmanned aerial vehicles, and the method comprises the steps: constructing an unmanned aerial vehicle library according to a management and control region, obtaining the basic information of unmanned aerial vehicles, determining a no-load state performance index according to the types of the unmanned aerial vehicles, library function sub-regions and environment factors, and formulating a deployment list to start the unmanned aerial vehicle, calculating an initial state performance index in combination with hardware, electrical and environmental factors, and planning a reconnaissance route to execute a task according to characteristics of a management and control area. The accident type and the fixed-focus unmanned aerial vehicle position are analyzed from the investigation video, and initial accident information and report concurrent emergency response signals are generated. And determining the position of the cooperative unmanned aerial vehicle according to the position of the focus unmanned aerial vehicle and the signal, and determining standard accident report information according to the position of the cooperative unmanned aerial vehicle and the video information, thereby realizing unmanned aerial vehicle cooperative operation scheduling, accurately processing an accident scene, and solving the defects of an existing method in unmanned aerial vehicle performance evaluation, accident monitoring and positioning and cooperative operation scheduling.

Description

Technical Field

[0001] The present invention relates to the technical field of unmanned aerial vehicles (UAVs), and in particular to an intelligent identification and positioning and processing method and system for accident scenes based on a UAV library. Background Art

[0002] Traditional accident handling methods have limitations such as delayed response and incomplete information acquisition. The accident handling system based on the drone library has emerged. With the flexible and wide field of view of drones, it can quickly reach the accident scene, accurately identify and locate accident details, and efficiently generate reports and emergency plans, greatly improving the timeliness and accuracy of accident handling, effectively protecting the safety of life and property, and reducing accident losses and adverse social impacts.

[0003] The existing accident scene intelligent identification and positioning, processing methods and systems based on the drone library often consider the drone performance factors in isolation, lacking a comprehensive assessment of the empty, start-up and other states combined with the environment. In accident monitoring, the image analysis is simple and the position information processing is rough, making it difficult to achieve accurate drone performance evaluation, accurate accident monitoring and positioning, and reasonable scheduling of collaborative operations. Therefore, it is necessary to provide an accident scene intelligent identification and positioning, processing method and system based on the drone library to solve the above-mentioned problems. Summary of the invention

[0004] In order to solve the above technical problems, an accident scene intelligent identification and positioning, processing method and system based on a drone library are provided. This technical solution solves the problem that some accident scene intelligent identification and positioning, processing method and system based on a drone library proposed in the above background technology are difficult to achieve accurate evaluation of drone performance, accurate monitoring and positioning of accidents and reasonable scheduling of collaborative operations.

[0005] In order to achieve the above purpose, the technical solution adopted by the present invention is: Intelligent identification, positioning and processing methods of accident scenes based on the drone library include: Determine the control area information and build a drone library based on the control area information; Obtain basic information of drones in the drone library, wherein the basic information of drones includes drone type information, drone number information, and drone parking location information; Based on the basic information of the drone, obtain the drone's empty state information and determine the drone's empty state performance index; Determine the drone deployment list information based on the drone basic information and drone empty state performance index; Based on the drone deployment list information, start the drones in sequence, and obtain the drone initial state performance index when the drone is started. At the same time, combined with the control area information, determine the drone reconnaissance route information; Using the drone reconnaissance route information, the drone reconnaissance mission is performed, and the drone reconnaissance video information is obtained at the same time. Based on the drone reconnaissance video information, it is determined whether an accident has occurred. If not, the reconnaissance mission is completed based on the drone reconnaissance route information. If so, the accident type information is obtained based on the drone reconnaissance video information, and the focus drone location information is determined at the same time; Determine the initial accident information based on the accident type information, the focus drone location information, and the drone reconnaissance video information, and send the initial accident report information and emergency response signal; Determine the location information of the cooperative drone based on the location information of the focus drone and the emergency response signal; Determine standard accident report information based on the focus drone location information, the coordinated drone location information, and the drone reconnaissance video information; Send standard accident report information, determine emergency handling information, and guide drone information to properly handle the accident scene.

[0006] Preferably, the acquiring of the drone's empty state information based on the drone's basic information and determining the drone's empty state performance index specifically includes: According to the basic information of the drone, the drone type information is determined, and the drone library is initially divided into regions according to the drone type information to obtain the functional sub-region information of the drone library; Based on the functional sub-area information of the drone library, obtain the functional sub-area environment information; Perform secondary regional division on the functional sub-region based on the environmental information of the functional sub-region to obtain the environmental factor sub-region information; Determine the basic parameters of the drone type and the basic parameters of the environmental factors based on the drone type information, the functional sub-area information, the environmental factor sub-area information and the functional sub-area environmental information; Based on the basic parameters of the drone type and the basic parameters of the environmental factors, determine the drone type influencing sub-coefficient and the environmental factor influencing sub-coefficient; The sum of the UAV type influencing sub-coefficient and the environmental factor influencing sub-coefficient is taken as the UAV no-load state influencing coefficient; According to the drone's empty state information, determine the drone's mechanical performance information, drone's electrical performance information and environmental adaptability information; Determine the initial performance index of the drone in the empty state based on the drone's mechanical performance information, drone's electrical performance information and environmental adaptability information; Based on the UAV no-load state influence coefficient and the UAV no-load state initial performance index, the UAV no-load state performance index is determined.

[0007] Preferably, based on the drone deployment list information, the drones are started in sequence, and the drone initial state performance index is obtained when the drone is started, and the drone reconnaissance route information is determined in combination with the control area information, specifically including: Obtain information related to the hardware status, electrical system, and environmental factors when the drone is started; Determine the hardware status adjustment coefficient, the electrical system adjustment coefficient and the environmental factor adjustment coefficient based on the hardware status related information, the electrical system related information and the environmental factor related information; Obtain the product value of the drone's no-load state performance index, hardware state adjustment coefficient, electrical system adjustment coefficient, and environmental factor adjustment coefficient to determine the initial state performance index; Based on the initial state performance index and drone type information, the drones are preliminarily classified into three groups of high, medium and low performance drones; According to the information of the controlled area, obtain the topographic and geomorphic feature information within the controlled area; Based on the topographical features in the control area, a spatial coordinate system of the performance attenuation area is constructed; According to the initial state performance index and the spatial coordinate system of the performance decay region, the initial UAV reconnaissance routes are constructed for the three groups of UAVs with high, medium and low performance; Based on the initial reconnaissance route of the UAV, the process of three groups of UAVs with high, medium and low performance performing reconnaissance tasks was simulated respectively, and the real-time state performance prediction index of the UAVs when performing reconnaissance tasks was obtained; Determine the coordinate values ​​of the drone in the spatial coordinate system of the performance degradation area and determine the location information of the nearest drone hangar; The drone reconnaissance residual energy index is set based on the location information of the nearest drone hangar and the coordinate values ​​of the drone in the spatial coordinate system of the performance attenuation area; If the real-time performance prediction index of the drone when performing the reconnaissance mission is greater than the drone's reconnaissance residual energy index, the drone will continue to perform the reconnaissance mission; If the real-time state performance prediction index of the drone when performing the reconnaissance mission is less than the drone reconnaissance residual energy index, the drone's initial reconnaissance route is marked as abnormal, and the drone's initial reconnaissance route is reconstructed; If the real-time state performance prediction index of the UAV when performing the reconnaissance mission is equal to the UAV reconnaissance residual energy index, the initial reconnaissance route of the UAV is used as the UAV reconnaissance route information.

[0008] Preferably, the method of obtaining accident type information based on drone reconnaissance video information and determining the focus drone location information specifically includes: Based on the drone reconnaissance video information, obtain the drone reconnaissance frame image; Analyze the drone reconnaissance frame images to obtain abnormal reconnaissance frame images; Obtain abnormal type information based on abnormal detection frame images; Determine the detection timestamp information of the abnormal detection frame image through the abnormal type information; Determine whether the reconnaissance timestamp information is empty. If so, traverse the drone reconnaissance frame images, and obtain the abnormal reconnaissance frame associated images, and at the same time associate and integrate the abnormal reconnaissance frame associated images with the abnormal reconnaissance frame images to obtain the accident type judgment image set and the first emergency response information. If not, obtain the abnormal reconnaissance frame difference image based on the reconnaissance frame timestamp information, and at the same time associate and integrate the abnormal reconnaissance frame difference image with the abnormal reconnaissance frame image to obtain the accident type judgment image set; Determine the accident type information through the accident type judgment image set, and obtain the accident geographical location information in the accident type judgment image set; Determine whether the accident geographic location information is empty. If so, determine the image set based on the accident type, obtain the accident geographic feature information, obtain the accident geographic location area information, and determine the accident geographic location prediction information and the second emergency response information. If not, use the accident geographic location information as the accident geographic location prediction information; The accident geographic location prediction information is used as the focus UAV location information.

[0009] Preferably, determining the initial accident information according to the accident type information, the focus drone location information and the drone reconnaissance video information, and sending the initial accident report information and the emergency response signal specifically includes: Determine the accident impact level information based on the accident type information and drone reconnaissance video information; Determine the first accident coding information based on the accident impact level information and the accident type information; Determine the second accident coding information based on the focus drone location information; Using drones to detect video information, determine accident feature information, and perform data processing on the accident feature information to obtain third accident coding information; Based on the accident impact level information and the accident type information, communication channel information, communication method information and communication protocol information are determined, and the first accident coding information, the second accident coding information and the third accident coding information are digitally signed to obtain initial accident coding information; Construct a preset report template, and input the accident type information, the focus drone location information, the accident impact level information and the accident feature information into the preset report template to obtain the initial accident text information; Integrate the initial accident code information and the initial accident text information to obtain the initial accident information, and determine and send the initial accident report information; Based on the initial accident coding information, the first emergency response information and the second emergency response information, an emergency response signal is set, and the initial accident report information is data processed to obtain initial accident report backup information; Based on the accident impact level information, emergency response scope information is determined, and based on the emergency response scope information, initial accident report backup information and an emergency response signal are sent.

[0010] Preferably, the determining the cooperative drone position information based on the focus drone position information and the emergency response signal specifically includes: Based on the focus UAV position information, determine the focus UAV attitude information; According to the focus UAV attitude information, determine the focus UAV tilt angle information and heading angle information; Obtain the sending timestamp information of the emergency response signal, and determine the straight-line distance information between the focus UAV and the coordinated UAV by combining the focus UAV position information, focus UAV tilt angle information and heading angle information; Based on the focus UAV position information and the straight-line distance information between the focus UAV and the cooperative UAV, the response position information of the cooperative UAV is determined; The coordinates of the cooperative UAV in the spatial coordinate system of the performance attenuation area are determined based on the response position information of the cooperative UAV, and the corresponding real-time state performance prediction index of the cooperative UAV is obtained at the same time; Based on the real-time status performance prediction index corresponding to the cooperative UAV, the cooperative UAV is assigned a cooperative task, and the cooperative task information is determined, and the cooperative residual energy index is set; The location information of the collaborative UAV is determined according to the real-time status performance prediction index, collaborative residual energy index and collaborative task information corresponding to the collaborative UAV.

[0011] Preferably, the determining of standard accident report information according to the focus drone position information, the coordinated drone position information and the drone reconnaissance video information specifically includes: Determine the distribution strategy information according to the focus UAV location information and the coordinated UAV location information; Taking the focus drone position as the origin, construct the plane coordinate system of the accident area; Based on the distribution strategy information, the focus UAV location information and the cooperative UAV location information, the accident standard area information is determined; Use drones to detect video information to determine the accident object information, accident scene environment information and accident scene scope information; Conduct trend prediction through accident scene environment information and accident scene range information to obtain accident impact range information; According to the accident standard area information, the accident standard area is shadowed in the accident area plane coordinate system to obtain accident shadow graphic information; Determine the geometric center of the accident shadow figure, and obtain the coordinate information of the geometric center of the accident shadow figure as the accident center position information; Determine standard accident report information based on accident type information, detection timestamp information, accident center location information and accident object information.

[0012] Furthermore, an accident scene intelligent identification, positioning and processing system based on a drone library is proposed to implement any of the positioning and processing methods described above, including: A drone resource management and information integration module, which is used to deeply analyze the geographic information of the control area, build a drone library, plan the layout of the drone library according to the type, performance and maintenance requirements of the drones, obtain the basic information of the drones in the drone library, and perform preliminary detection and sorting of the drone empty state information; UAV performance evaluation and mission planning module, which is used to calculate the UAV no-load state performance index based on the UAV type and environmental factors, determine the UAV initial state performance index based on the hardware, electrical and environmental data at startup, formulate a UAV deployment list based on the UAV performance index and the characteristics of the control area, plan the UAV reconnaissance route, and set the residual energy index and exception handling mechanism; The accident monitoring and emergency response module is used to receive drone reconnaissance videos, use image analysis technology to identify anomalies, determine the type of accident, handle the lack of accident location information, determine the location of the focus drone, and then determine the location information of the accident center, generate and send an initial accident report and emergency response signal based on the accident information, and determine the location information of the coordinated drone based on the focus drone position and signal and dispatch it; The accident handling and report generation module is used to determine the emergency handling plan based on the accident type and on-site conditions, dispatch and guide drones to assist in rescue operations, integrate various types of accident information, generate standard accident reports, store and distribute accident reports, and provide data support for accident analysis and subsequent decision-making.

[0013] Preferably, the accident monitoring and emergency response module includes: An accident monitoring and positioning unit, which is used to receive drone reconnaissance videos, use image analysis technology to identify anomalies, determine the type of accident, handle the lack of accident location information, determine the focus drone location, and then determine the accident center location information; The emergency response unit is used to generate and send an initial accident report and an emergency response signal based on the accident information, and determine the location information of the coordinated drone and dispatch it according to the location and signal of the focus drone.

[0014] Preferably, the accident handling and report generation module includes: An accident handling unit, which is used to determine an emergency handling plan based on the type of accident and the on-site situation, and dispatch and guide drones to assist in rescue operations; The report generation unit is used to integrate various types of accident information, generate standard accident reports, store and distribute accident reports, and provide data support for accident analysis and subsequent decision-making.

[0015] Compared with the prior art, the present invention has the following beneficial effects: The proposed scheme is based on the UAV library of accident scene intelligent identification and positioning, processing method and system, through the UAV reconnaissance video information, extract UAV reconnaissance frame images, analyze and obtain abnormal reconnaissance frame images, determine abnormal type information and corresponding reconnaissance timestamp information, and process and integrate the accident type judgment image set according to different timestamp information, and then determine the accident type information, and combine the accident geographical location information obtained or further inferred from the image set to determine the location information of the focus UAV, so as to comprehensively consider the UAV's empty state, hardware, electrical and environmental factors at startup, comprehensively and accurately evaluate the performance status of the UAV at different stages, and reasonably classify the UAV according to performance, so as to solve the problem of considering the performance factors of the UAV in isolation and lacking the comprehensive evaluation of the empty state, startup and other states combined with the environment, and realize the accurate evaluation of the UAV performance, and provide a basis for the reasonable arrangement of subsequent tasks; The intelligent identification, positioning, and processing method and system of the accident scene based on the drone library proposed in this scheme determines the posture information of the focus drone based on its position information, and determines the straight-line distance information between the focus drone and the cooperative drone in combination with the timestamp information of the emergency response signal, thereby determining the response position information of the cooperative drone, and then considering the real-time state performance prediction index of the cooperative drone to allocate the cooperative task, set the cooperative residual energy index, and finally determine the position information of the cooperative drone; determine the distribution strategy information based on the position information of the focus drone and the position information of the cooperative drone, construct a plane coordinate system of the accident area, and determine various accident-related information in combination with the drone reconnaissance video information, and then determine the standard accident report information, comprehensively consider various factors to reasonably determine the position, task and related operation parameters of the cooperative drone, and sort out the standard accident report information based on the overall situation of the accident, so as to ensure that the cooperative operation is carried out in an orderly and reasonable manner, solve the problem of difficulty in realizing reasonable scheduling of cooperative operations, and enable different drones to cooperate at the accident scene according to their own conditions and overall needs to effectively handle the accident scene situation; The scheme proposes an intelligent identification and positioning, processing method and system for accident scenes based on a drone library. Based on drone reconnaissance video information, drone reconnaissance frame images are extracted, abnormal reconnaissance frame images are analyzed to obtain abnormal type information and corresponding reconnaissance timestamp information, and the accident type information is obtained by processing and integrating the accident type judgment image set according to different timestamp information. The accident type information is then determined, and the location information of the focus drone is determined in combination with the accident geographic location information (including geographic location prediction information) obtained or further inferred from the image set. The image information in the drone reconnaissance video is carefully analyzed, and multi-dimensional data such as image features and timestamps are fully utilized to accurately judge the type of accident, and relatively accurately determine the geographical location of the accident, clarify the location of the focus drone, and solve the problem of simple image analysis and rough location information processing in accident monitoring, which makes it difficult to accurately monitor and locate accidents, thereby improving the accuracy of accident monitoring and positioning. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 This is a flow chart of the accident scene intelligent identification, positioning and processing method based on the drone library proposed by the present invention; Figure 2 This is a flow chart for obtaining the unmanned aerial vehicle no-load state performance index in the present invention; Figure 3 A flow chart for determining the UAV reconnaissance route information in the present invention; Figure 4 This is a system framework diagram of the accident scene intelligent identification, positioning and processing system based on the drone library proposed in the present invention. DETAILED DESCRIPTION

[0017] The following description is used to disclose the present invention so that those skilled in the art can implement the present invention. The preferred embodiments described below are only examples, and those skilled in the art may think of other obvious variations.

[0018] Reference Figure 1 - Figure 4 As shown, the accident scene intelligent identification, positioning and processing method based on the drone library includes: Determine the control area information and build a drone library based on the control area information; Obtain basic information about drones in the drone library, including drone type information, drone number information, and drone parking location information; Based on the basic information of the drone, obtain the drone's empty state information and determine the drone's empty state performance index; Determine the drone deployment list information based on the drone basic information and drone empty state performance index; Based on the drone deployment list information, start the drones in sequence, and obtain the drone initial state performance index when the drone is started. At the same time, combined with the control area information, determine the drone reconnaissance route information; Using the drone reconnaissance route information, the drone reconnaissance mission is performed, and the drone reconnaissance video information is obtained at the same time. Based on the drone reconnaissance video information, it is determined whether an accident has occurred. If not, the reconnaissance mission is completed based on the drone reconnaissance route information. If so, the accident type information is obtained based on the drone reconnaissance video information, and the focus drone location information is determined at the same time; Determine the initial accident information based on the accident type information, the focus drone location information, and the drone reconnaissance video information, and send the initial accident report information and emergency response signal; Determine the location information of the cooperative drone based on the location information of the focus drone and the emergency response signal; Determine standard accident report information based on the focus drone location information, the coordinated drone location information, and the drone reconnaissance video information; Send standard accident report information, determine emergency handling information, and guide drone information to properly handle the accident scene.

[0019] Specifically, the drone library is constructed and basic information is obtained. The drone library is constructed after the control area is clarified, and basic information including the type, number, and parking location of the drone is obtained, which is used as the basis for subsequent operations. Determining the no-load state performance index involves dividing the library into regions according to the type of drone, dividing the region again considering environmental factors, determining the relevant basic parameters and influencing sub-coefficients, and finally calculating the no-load state performance index in combination with the mechanical, electrical, and environmental adaptability information of the drone, which is used to evaluate the initial performance state of the drone. Determining the reconnaissance route information is to calculate the initial state performance index in combination with the hardware, electrical, and environmental factor-related information when starting the drone, classify the drone, construct a coordinate system based on the topography of the control area, and simulate the reconnaissance mission to determine the real-time state performance prediction index, etc., so as to determine a reasonable reconnaissance route and ensure the smooth implementation of the reconnaissance mission. Determine and process the accident-related information, analyze the accident type and the location information of the focus drone from the drone reconnaissance video information, and then determine the initial accident information, send reports and signals, and determine the location information of the coordinated drone and the standard accident report information, so as to facilitate accurate response and subsequent processing of the accident.

[0020] It is understandable that the control area information refers to the description of the specific geographical area, topography, environmental characteristics, etc. that need to be monitored and managed. It is the basic basis for building a drone library, planning drone reconnaissance routes, and evaluating the performance of drones in performing tasks in the area, which limits the scope of the entire accident monitoring and handling work. The basic information of drones includes drone type information, drone number information, and drone parking location information. The drone type information describes the specific model, specifications, and other classifications of drones. Different types of drones have different performance characteristics. The number information, that is, the number of drones in the drone library, is related to the amount of resources that can be deployed for reconnaissance and other tasks. The parking location information refers to the specific parking coordinates and other locations of each drone in the drone library, which is convenient for accurate startup and scheduling. The basic information of drones is used to grasp the status of drone resources and is the basic data support for the subsequent determination of the empty state performance index and the formulation of deployment plans.

[0021] Furthermore, standard accident report information is sent, and emergency handling information is determined. At the same time, the guidance drone information is determined to reasonably handle the accident scene. The standard accident report information includes basic information about the accident, on-site conditions, etc. The basic information includes the time and location of the accident. On-site emergency measures include alarming and seeking help. The guidance drone information is based on the real-time status performance prediction index to determine the drone with the best performance and the best status. The drone is used as a command drone to ensure that the command center can keep in touch with the drone, adjust the mission in a timely manner, optimize the flight route, and promptly transmit on-site information to relevant departments, and update the emergency handling status.

[0022] Furthermore, based on the basic information of the drone, the drone's empty state information is obtained, and the drone's empty state performance index is determined, including: According to the basic information of the drone, the drone type information is determined, and the drone library is initially divided into regions according to the drone type information to obtain the functional sub-region information of the drone library; Based on the functional sub-area information of the drone library, obtain the functional sub-area environment information; Perform secondary regional division on the functional sub-region based on the environmental information of the functional sub-region to obtain the environmental factor sub-region information; Determine the basic parameters of the drone type and the basic parameters of the environmental factors based on the drone type information, the functional sub-area information, the environmental factor sub-area information and the functional sub-area environmental information; Based on the basic parameters of the drone type and the basic parameters of the environmental factors, determine the drone type influencing sub-coefficient and the environmental factor influencing sub-coefficient; The sum of the UAV type influencing sub-coefficient and the environmental factor influencing sub-coefficient is taken as the UAV no-load state influencing coefficient; According to the drone's empty state information, determine the drone's mechanical performance information, drone's electrical performance information and environmental adaptability information; Determine the initial performance index of the drone in the empty state based on the drone's mechanical performance information, drone's electrical performance information and environmental adaptability information; Based on the UAV no-load state influence coefficient and the UAV no-load state initial performance index, the UAV no-load state performance index is determined.

[0023] Specifically, the no-load state information covers the mechanical properties, electrical properties, environmental adaptability and other aspects of the UAV when it is not carrying mission equipment and flying without load. The UAV no-load state performance index is an indicator that reflects the comprehensive performance of the UAV in the no-load state and is quantified by comprehensively analyzing these conditions. It evaluates the capability level of the UAV in the initial state, and serves as an important reference for subsequently determining its initial state performance index when actually performing reconnaissance and other tasks, and assists in the reasonable arrangement of UAV task allocation.

[0024] It is understandable that the initial regional division is based on the drone type information. First, the drone hangar is divided into initial regions based on the drone type information in the drone basic information. Different types of drones often have different performance characteristics, size specifications, functional uses, etc. For example, some drones are suitable for long-distance reconnaissance, some are better at low-altitude close-range observation, and some focus on carrying different types of monitoring equipment. By dividing by type, the drone hangar can be divided into different functional sub-areas, so that the drones stored in each area are relatively consistent in functional characteristics, which is convenient for subsequent management, maintenance, and targeted performance evaluation. The functional sub-area information of the drone hangar obtained in this way lays the foundation for subsequent operations. Obtain the environmental information of the functional sub-area. After clarifying the functional sub-area of ​​the drone hangar, collect the relevant environmental information for each functional sub-area. The environmental information here covers many aspects, such as the temperature, humidity, ventilation, electromagnetic environment, etc. of the area. Because these environmental factors will affect the storage status of the drone and the performance after subsequent takeoff, understanding this information will help to further refine the analysis of the degree to which the drone performance is affected by the environment, and provide a basis for the next secondary regional division. Perform secondary area division to obtain environmental factor sub-area information, and use the obtained functional sub-area environmental information to further subdivide the previously divided functional sub-area to obtain environmental factor sub-area information. For example, for a functional sub-area, if the temperature of one part of the area is relatively stable and the temperature of the other part fluctuates greatly, it can be further divided according to the environmental factor of temperature, so that the environmental factors in the divided sub-areas are more homogeneous, which is more convenient for accurately considering the effect of the environment on the performance of the drone, and also prepares for the subsequent determination of relevant parameters and coefficients. Determine the basic parameters of the drone type and the basic parameters of environmental factors, and comprehensively consider the drone type information, functional sub-area information, environmental factor sub-area information, and functional sub-area environmental information. The basic parameters of the drone type will be determined based on the specific model and function settings of the drone, such as its maximum flight speed, maximum range, and the weight of the equipment that can be carried, which reflect its own performance characteristics; while the basic parameters of environmental factors are quantitative reference values ​​of the impact of environmental factors such as temperature and humidity on the performance of the drone from the perspective of the environment, such as the corresponding parameters of the impact of different temperature ranges on the battery life of the drone. The determination of these basic parameters is a key pre-step for the subsequent calculation of the influencing sub-coefficients. Determine the sub-coefficients of drone type and environmental factors. Based on the basic parameters of the drone type obtained above, analyze the influence of the performance characteristics of the type of drone on the overall performance through specific calculation methods or empirical models, and thus determine the sub-coefficients of drone type. Similarly, based on the basic parameters of environmental factors, measure the proportion of environmental factors in the performance of drones and obtain the sub-coefficients of environmental factors.For example, if the maximum range parameter of a certain type of drone accounts for a large proportion in the comprehensive evaluation, then the corresponding type influence sub-coefficient will also be relatively high; if the humidity in a certain environmental factor sub-area has a significant impact on the electrical performance of the drone, then the environmental factor influence sub-coefficient corresponding to the humidity will be more prominent. Calculate the drone's no-load state influence coefficient, add the drone type influence sub-coefficient and the environmental factor influence sub-coefficient, and the sum is the drone no-load state influence coefficient. This coefficient comprehensively reflects the degree of influence of the drone's own type characteristics and the environmental factors on its performance in the no-load state, and is the key intermediate quantity connecting the basic performance-related parameters and the final no-load state performance index. Determine the drone's mechanical performance information, drone electrical performance information, and environmental adaptability information, and analyze based on the drone's no-load state information. The mechanical performance information includes the drone's fuselage structural strength, flight stability, and operational flexibility of each mechanical component; the electrical performance information involves battery life, circuit system stability, and the working status of each electronic device; the environmental adaptability information reflects whether the drone can work normally under different environmental conditions (such as high temperature, low temperature, humidity, etc.) and the degree to which the performance is affected. This information describes the actual performance of the drone in an empty state from different dimensions, providing specific data support for determining the initial performance index of the empty state. Determine the performance index of the drone in an empty state. Finally, based on the calculated drone empty state impact coefficient and the drone empty state initial performance index, the drone empty state performance index is finally determined through a specific calculation relationship between the two (such as multiplication, etc., depending on the set evaluation model). This index comprehensively and comprehensively reflects the actual performance of the drone in an empty state after considering the type of drone itself and the environmental factors in which it is located, and provides an important performance reference for subsequent judgments on whether the drone is suitable for performing tasks and how to reasonably arrange tasks.

[0025] Determine the UAV type influencing coefficient. Assume that there are m UAVs, each with three basic parameters of UAV type (flight speed ,voyage , Weight of equipment that can be carried ), construct an m-by-3 data matrix X. Standardize the data, and set the standardized data matrix Z. For each element in the matrix X (No. The first drone parameters), the standardized formula is ,in For the The mean of the parameters, For the The standard deviation of the parameters; Calculate the covariance matrix C of the standardized data matrix Z. The elements of the covariance matrix are The calculation formula is , calculate the eigenvalues ​​of the covariance matrix C , , and the corresponding eigenvector , , ; Let the principal component Y=ZV, where , V is a matrix composed of eigenvectors, and the first principal component is + , the second principal component is + , the third principal component is + ,in The flight speed The corresponding standardized data is For the voyage The corresponding standardized data is Weight of the equipment that can be carried The corresponding standardized data; The variance contribution of the first principal component is , then the variance contribution rate of the first principal component is the value of one of the UAV type influencing sub-coefficients.

[0026] Furthermore, based on the drone deployment list information, the drones are started in sequence, and the drone initial state performance index is obtained when the drone is started. At the same time, combined with the control area information, the drone reconnaissance route information is determined, including: Obtain information related to the hardware status, electrical system, and environmental factors when the drone is started; Determine the hardware status adjustment coefficient, the electrical system adjustment coefficient and the environmental factor adjustment coefficient based on the hardware status related information, the electrical system related information and the environmental factor related information; Obtain the product value of the drone's no-load state performance index, hardware state adjustment coefficient, electrical system adjustment coefficient, and environmental factor adjustment coefficient to determine the initial state performance index; Based on the initial state performance index and drone type information, the drones are preliminarily classified into three groups of high, medium and low performance drones; According to the information of the controlled area, obtain the topographic and geomorphic feature information within the controlled area; Based on the topographical features in the control area, a spatial coordinate system of the performance attenuation area is constructed; According to the initial state performance index and the spatial coordinate system of the performance decay region, the initial UAV reconnaissance routes are constructed for the three groups of UAVs with high, medium and low performance; Based on the initial reconnaissance route of the UAV, the process of three groups of UAVs with high, medium and low performance performing reconnaissance tasks was simulated respectively, and the real-time state performance prediction index of the UAVs when performing reconnaissance tasks was obtained; Determine the coordinate values ​​of the drone in the spatial coordinate system of the performance degradation area and determine the location information of the nearest drone hangar; The drone reconnaissance residual energy index is set based on the location information of the nearest drone hangar and the coordinate values ​​of the drone in the spatial coordinate system of the performance attenuation area; If the real-time performance prediction index of the drone when performing the reconnaissance mission is greater than the drone's reconnaissance residual energy index, the drone will continue to perform the reconnaissance mission; If the real-time state performance prediction index of the drone when performing the reconnaissance mission is less than the drone reconnaissance residual energy index, the drone's initial reconnaissance route is marked as abnormal, and the drone's initial reconnaissance route is reconstructed; If the real-time state performance prediction index of the UAV when performing the reconnaissance mission is equal to the UAV reconnaissance residual energy index, the initial reconnaissance route of the UAV is used as the UAV reconnaissance route information.

[0027] Specifically, hardware status related information: This involves the status of each hardware component of the drone when it is started, such as the start-up status of the engine, the integrity of the mechanical structure such as the wing, whether the sensor is working properly, etc. These hardware statuses will directly affect the drone's ability to perform subsequent tasks. Different hardware conditions correspond to different levels of performance, so it is necessary to collect this information to further analyze its impact on the overall performance. Electrical system related information: It includes battery power, whether the circuit connection is stable, and the working status of various electrical equipment (such as flight control system, communication equipment, etc.). The electrical system is the key guarantee for the normal operation of the drone. Whether the power is sufficient or not, whether the circuit is normal, etc. will cause the actual performance of the drone to fluctuate, so this information is very important for evaluating the performance of the drone after it is started. Environmental factor related information: It refers to the relevant conditions of the external environment when the drone is started, such as the temperature, humidity, wind speed, air pressure, and the surrounding electromagnetic environment at that time. The external environment will have a significant impact on the flight performance of the drone. For example, high temperature affects battery life and strong winds interfere with flight stability, so this information should be obtained to comprehensively consider the actual performance level that the drone can achieve.

[0028] Determine each adjustment coefficient, hardware status adjustment coefficient: by analyzing the hardware status related information, using specific evaluation methods or referring to previous experience, experimental data, etc., determine a coefficient to quantify the impact of the hardware status on the overall performance of the drone. For example, if the hardware components are in good condition, the coefficient is close to 1, indicating that the impact on performance is small; if there are hidden dangers of some hardware failures, the coefficient will be correspondingly smaller, which means that the overall performance will be discounted. Electrical system adjustment coefficient: based on the electrical system related information, according to similar evaluation logic, determine the impact coefficient of the electrical system on the performance of the drone. For example, when the battery is fully charged and the electrical equipment is running stably, the coefficient is large; if the battery is insufficient or there is a risk of circuit failure, the coefficient will be reduced, reflecting the performance degradation caused by electrical system problems. Environmental factor adjustment coefficient: considering the environmental factor related information, measure the effect of the external environment on the performance of the drone and determine the coefficient. For example, in a good environment with suitable temperature and breeze, the coefficient is high; while in bad weather and complex electromagnetic interference environment, the coefficient will be small, reflecting the adverse impact of the environment on the performance of the drone.

[0029] Determine the initial state performance index, multiply the drone's no-load state performance index with the hardware state adjustment coefficient, electrical system adjustment coefficient, and environmental factor adjustment coefficient. The product value is the initial state performance index. This index comprehensively considers the drone's own no-load performance and the influence of hardware, electrical, and environmental factors at startup. It can more accurately reflect the actual performance status of the drone at the moment it is about to perform a mission, providing a basis for subsequent classification and mission planning.

[0030] The drones are preliminarily classified according to the calculated initial state performance index and drone type information. Different types of drones have different performance bases. Combined with the initial state performance index, they can be roughly divided into three groups: high, medium, and low performance. For example, for reconnaissance drones, if their initial state performance index is high and their type has advantages in reconnaissance functions, they will be classified as high-performance groups; conversely, if the index is low and the performance is relatively weak, they will be classified as low-performance groups. This classification makes it easier to arrange reconnaissance tasks according to drones of different performance levels and give full play to the role of each drone.

[0031] Obtain terrain and geomorphic feature information and construct a coordinate system. Terrain and geomorphic feature information: extract terrain and geomorphic information such as mountains, rivers, plains, and building distribution from the control area information. These terrains and geomorphic features will affect the flight of the drone. For example, mountainous areas will cause signal obstruction and complex airflow, while plains will allow relatively smooth flight. Mastering this information will help plan suitable reconnaissance routes, improve reconnaissance efficiency, and ensure the safety of drones. Construct a spatial coordinate system for performance attenuation areas: Based on the terrain and geomorphic features within the control area, establish a spatial coordinate system to describe the performance attenuation of the drone when flying in the area. Different terrain and geomorphic areas can be mapped to different positions in the coordinate system, and corresponding coordinate values ​​can be assigned according to the degree of their impact on the performance of the drone, which will facilitate subsequent analysis of the performance changes of the drone at different positions and the planning of reasonable flight routes.

[0032] Construct the initial reconnaissance route of the drone. According to the initial state performance index obtained above and the constructed performance attenuation area spatial coordinate system, construct the initial reconnaissance routes for the three groups of drones with high, medium and low performance. For high-performance drones, they will be arranged with routes that have wider coverage and pass through more complex terrain areas to fully utilize their performance advantages to obtain more information; medium-performance drones will plan routes of relatively moderate difficulty; low-performance drones will avoid complex terrain and have relatively short distances, etc., to ensure that they can complete the reconnaissance mission within the range allowed by their own performance.

[0033] Simulate the execution of reconnaissance missions and obtain the real-time state performance prediction index. According to the constructed initial reconnaissance route of the drone, simulate the process of three groups of drones performing reconnaissance missions. In this simulation process, comprehensively consider the factors such as the change of the drone's position during flight (the change of coordinates in the spatial coordinate system of the performance attenuation area) and environmental changes, and calculate the real-time state performance prediction index of the drone at each stage through the corresponding performance prediction model or algorithm. This index can reflect the performance changes of the drone in the process of executing the mission in real time, so as to facilitate the timely discovery of potential problems and make adjustments.

[0034] Determine the coordinate value, the location information of the nearest drone hangar and set the reconnaissance residual energy index, determine the coordinate value: clarify the specific coordinate position of the drone in the spatial coordinate system of the performance attenuation area, which can intuitively know where the drone is in the control area, as well as the corresponding terrain and its impact on performance.

[0035] Determine the location of the nearest drone hangar: Find the location of the drone hangar closest to the current location of the drone. This is important for subsequent decisions such as whether the drone has enough energy to return, whether it needs to be replaced or repaired in time, etc.

[0036] Set the drone reconnaissance remaining energy index: Combine the location information of the nearest drone warehouse and the coordinate value of the drone in the coordinate system, and comprehensively consider factors such as the drone's remaining flight capability and the energy required for the return trip. Set an index to measure the drone's remaining capacity to continue to perform reconnaissance missions, which is the reconnaissance remaining energy index.

[0037] According to the real-time state performance prediction index and the reconnaissance residual energy index, the reconnaissance route is judged and adjusted. If the real-time state performance prediction index is greater than the reconnaissance residual energy index, it means that the drone currently has enough performance margin to continue to perform the reconnaissance mission, and there will be no situation that affects the completion of the mission, such as insufficient energy. Therefore, the mission can be continued according to the original plan, giving full play to its performance advantages to obtain more reconnaissance information. If the real-time state performance prediction index is less than the reconnaissance residual energy index, it means that if the drone continues to perform the mission in the current state, it will face the risk of energy exhaustion, insufficient performance, and even inability to return safely. At this time, it is necessary to mark the initial reconnaissance route of the drone as abnormal, indicating that there is a problem with the route, and then rebuild a more suitable initial reconnaissance route to ensure that the drone can complete the reconnaissance mission safely and effectively. If the real-time state performance prediction index is equal to the reconnaissance residual energy index, it means that the drone is just in a critical state to complete the mission. At this time, the current initial reconnaissance route of the drone is used as the final drone reconnaissance route information, and the mission is continued according to this route, but its performance changes should also be closely monitored to prevent unexpected situations from affecting the execution of the mission.

[0038] Furthermore, based on the drone reconnaissance video information, the accident type information is obtained, and the focus drone location information is determined, including: Based on the drone reconnaissance video information, obtain the drone reconnaissance frame image; Analyze the drone reconnaissance frame images to obtain abnormal reconnaissance frame images; Obtain abnormal type information based on abnormal detection frame images; Determine the detection timestamp information of the abnormal detection frame image through the abnormal type information; Determine whether the reconnaissance timestamp information is empty. If so, traverse the drone reconnaissance frame images, and obtain the abnormal reconnaissance frame associated images, and at the same time associate and integrate the abnormal reconnaissance frame associated images with the abnormal reconnaissance frame images to obtain the accident type judgment image set and the first emergency response information. If not, obtain the abnormal reconnaissance frame difference image based on the reconnaissance frame timestamp information, and at the same time associate and integrate the abnormal reconnaissance frame difference image with the abnormal reconnaissance frame image to obtain the accident type judgment image set; Determine the accident type information through the accident type judgment image set, and obtain the accident geographical location information in the accident type judgment image set; Determine whether the accident geographic location information is empty. If so, determine the image set based on the accident type, obtain the accident geographic feature information, obtain the accident geographic location area information, and determine the accident geographic location prediction information and the second emergency response information. If not, use the accident geographic location information as the accident geographic location prediction information; The accident geographic location prediction information is used as the focus UAV location information.

[0039] Specifically, the drone reconnaissance frame images are obtained, and based on the drone reconnaissance video information, the video is decomposed into frames according to the video playback order and a certain time interval or frame number. These images are drone reconnaissance frame images. For example, a frame is extracted every 0.5 seconds to form a series of static images that can reflect the reconnaissance situation at different times. They carry visual information at different moments of the accident scene and are the basic materials for subsequent analysis and judgment.

[0040] Analyze and obtain abnormal reconnaissance frame images, and use image recognition technology and algorithms to analyze the extracted drone reconnaissance frame images one by one. For example, by comparing the differences between the shape, color, texture and other features of the objects in the image and the preset standard feature patterns in normal scenes, identify those images with abnormal conditions, such as smoke in the picture (compared with normal scenes without smoke), collapsed buildings (normal buildings should be intact), etc., which are abnormal reconnaissance frame images, and they indicate signs of accidents.

[0041] Obtain the abnormal type information, and further analyze the specific manifestations of the abnormalities in the detected abnormal detection frame images to determine the abnormal type information. For example, if flames and thick smoke appear in the image, the abnormal type can be determined to be a fire; if vehicles collide or deform on the road, it is determined to be a traffic accident; if part of the building structure collapses, it is a building collapse accident, etc. Abnormal type information helps to quickly determine the general nature of the accident and provides an important basis for more accurate handling of the accident later.

[0042] Determine the detection timestamp information of the abnormal detection frame image, and find the time stamp of the corresponding abnormal detection frame image in the entire detection video according to the abnormal type information, that is, the detection timestamp information. This timestamp records the specific time when the abnormal image appears, which can help sort out the time sequence of the accident and the subsequent correlation analysis of the image changes at different times. For example, determining when the accident began to show obvious signs is critical to grasping the development context of the accident.

[0043] Different processing is performed according to the situation of the detection timestamp information. When the detection timestamp information is empty: it means that due to some reasons (such as data transmission problems, video recording abnormalities, etc.), the accurate time stamp of the abnormal detection frame image cannot be obtained. At this time, it is necessary to traverse all the drone detection frame images to find other images that are associated with the current abnormal detection frame image. These associations are reflected in the similarity of scenes and the association of object positions. They are called abnormal detection frame associated images. Then these associated images are integrated with the abnormal detection frame images to form an accident type judgment image set, and the first emergency response information is generated at the same time. The first emergency response information contains some emergency measures that need to be taken immediately based on the preliminary judgment of the existing abnormal situation, and the accident type judgment image set provides a richer image data basis for the subsequent accurate judgment of the accident type. When the detection timestamp information is not empty: use the existing detection frame timestamp information to obtain the abnormal detection frame difference image. For example, we can select images within a certain time range before and after the timestamp, calculate the difference between them and the abnormal detection frame images in terms of pixels, features, etc., to form abnormal detection frame difference images, and then associate and integrate these difference images with the abnormal detection frame images to obtain the same accident type judgment image set. Such an image set can better show the dynamic change process before and after the accident, and help to more accurately analyze the accident type.

[0044] Determine the accident type information and the accident geographic location information, and conduct a comprehensive analysis of the accident type judgment image set obtained earlier, such as observing the comprehensive performance of various abnormal features in the image set, comparing the change trends between different images, etc., to finally determine the accident type information and clarify what specific type of accident it is. At the same time, try to extract information that can reflect the geographical location of the accident from the accident type judgment image set, such as landmark buildings, geographic coordinates, road names, etc. that appear in the picture, so as to obtain the accident geographical location information and provide clues for the subsequent location of the accident scene.

[0045] Further processing is performed based on the accident geographic location information. When the accident geographic location information is empty: Since it is impossible to directly obtain accurate geographic location information, we use the image set based on the accident type to analyze the geographical features presented in it, such as the topographic features (mountainous areas, plains, or riverside, etc.) and surrounding environmental features (whether there are factories, residential areas, and other landmarks) shown in the image, and infer the geographical location area information of the accident based on this, and determine the approximate area of ​​the accident. At the same time, it is necessary to further predict the specific geographical location of the accident, obtain the predicted geographical location information of the accident, and generate the second emergency response information, which involves some broad emergency search and rescue direction suggestions that should be taken for this kind of uncertain geographical location. Finally, the predicted geographical location information of the accident is used as the focus drone location information, which indicates the location of the drone closest to the inferred accident area, so as to facilitate the subsequent coordinated rescue work around this location.

[0046] When the accident geographic location information is not empty: the acquired accident geographic location information is directly used as the accident geographic location prediction information, and then used as the focus drone location information, which means that the specific location of the accident or a more accurate location range has been identified, and resources can be accurately dispatched, rescue and further accident handling work can be carried out based on this.

[0047] Furthermore, based on the accident type information, the focus drone location information and the drone reconnaissance video information, the initial accident information is determined, and the initial accident report information and emergency response signal are sent, including: Determine the accident impact level information based on the accident type information and drone reconnaissance video information; Determine the first accident coding information based on the accident impact level information and the accident type information; Determine the second accident coding information based on the focus drone location information; Using drones to detect video information, determine accident feature information, and perform data processing on the accident feature information to obtain third accident coding information; Based on the accident impact level information and the accident type information, communication channel information, communication method information and communication protocol information are determined, and the first accident coding information, the second accident coding information and the third accident coding information are digitally signed to obtain initial accident coding information; Construct a preset report template, and input the accident type information, the focus drone location information, the accident impact level information and the accident feature information into the preset report template to obtain the initial accident text information; Integrate the initial accident code information and the initial accident text information to obtain the initial accident information, and determine and send the initial accident report information; Based on the initial accident coding information, the first emergency response information and the second emergency response information, an emergency response signal is set, and the initial accident report information is data processed to obtain initial accident report backup information; Based on the accident impact level information, emergency response scope information is determined, and based on the emergency response scope information, initial accident report backup information and an emergency response signal are sent.

[0048] Specifically, the accident impact level information is determined by comprehensively analyzing the accident type information and drone reconnaissance video information. Different accident types have different potential hazards. For example, fires are affected by factors such as the scale of combustion and the nature of the burning materials; traffic accidents are affected by the number of vehicles and the degree of collision. The scope and severity of the impact. Combined with the actual pictures presented by the drone reconnaissance video, such as the scope of damage at the accident site, whether there are signs of casualties, and the degree of damage to the surrounding environment and facilities, according to the pre-set level classification standards, determine what level of impact the accident belongs to, such as being divided into extremely serious, serious, and large levels, so that corresponding measures can be taken according to its severity in the future. Determine the first accident coding information. After clarifying the accident impact level information and accident type information, use specific coding rules to generate the first accident coding information. This coding rule is usually pre-set within the system, aiming to express complex accident key information in a concise and unified coding form, facilitating the storage, transmission, and subsequent rapid identification and processing of information. For example, a digital code is used to represent different accident types (assuming that fire is 01, collapse is 02, etc.), and another number is used to represent the accident impact level (1 for extremely serious, 2 for serious, etc.). The combination of the two forms a string of numbers that is the first accident coding information, which can quickly convey the core characteristics of the accident. Determine the second accident coding information and encode it based on the focus drone location information. The focus drone location information contains geographic coordinate data such as longitude and latitude or relative location identifiers, and will convert these location information into corresponding code forms according to the established geographic coding method or location identifier coding rules, that is, the second accident coding information. In this way, in the entire accident-related information system, the approximate location of the accident can be accurately located through the code, providing a location reference for subsequent actions such as rescue. Determine the third accident coding information, and carefully analyze and extract the accident feature information based on the drone reconnaissance video information, such as some special identifications at the accident site, the status of key objects involved (such as the color of the leaked material, the damage of special equipment, etc.), and the characteristics of the accident scene (in mountainous areas, urban areas, etc.). Then, these accident feature information are processed, involving operations such as feature quantification and classification coding, and converted into third-party accident coding information in accordance with the corresponding coding specifications. Its role is to supplement more detailed accident scene characteristics and assist in a more comprehensive understanding of the overall picture of the accident. To determine the initial accident coding information, first determine the communication channel information (such as whether to transmit information through wireless private networks, public networks, etc.), communication method information (such as real-time video transmission, text message sending, etc.) and communication protocol information (following specific data transmission protocols, such as TCP / IP protocols, etc.) based on the accident impact level information and accident type information to ensure the standardization and effectiveness of subsequent information interaction.Next, the first accident coding information, the second accident coding information and the third accident coding information are processed using digital signature technology. Digital signatures can ensure the integrity, authenticity and non-repudiation of information. After being signed, these coded information are integrated to form the initial accident coding information, which is an important and secure identification content in the transmission and processing of the entire accident information. Determine the initial accident text information and build a preset report template. The template covers the standard filling format and content sections of various accident-related information, such as the time of the accident (which can be extracted from the drone reconnaissance video information or obtained according to the system time record, etc.), the accident type, the focus drone location, the accident impact level and the accident characteristic information determined earlier. After the corresponding information is accurately filled in these preset sections, the initial accident text information is generated. This is a report content presented in text form, which is relatively easy to understand and can more comprehensively reflect the accident situation. It can be directly read by relevant personnel to understand the accident overview.

[0049] Further, based on the focus drone location information and the emergency response signal, the coordinated drone location information is determined, specifically including: Based on the focus UAV position information, determine the focus UAV attitude information; According to the focus UAV attitude information, determine the focus UAV tilt angle information and heading angle information; Obtain the sending timestamp information of the emergency response signal, and determine the straight-line distance information between the focus UAV and the coordinated UAV by combining the focus UAV position information, focus UAV tilt angle information and heading angle information; Based on the focus UAV position information and the straight-line distance information between the focus UAV and the cooperative UAV, the response position information of the cooperative UAV is determined; The coordinates of the cooperative UAV in the spatial coordinate system of the performance attenuation area are determined based on the response position information of the cooperative UAV, and the corresponding real-time state performance prediction index of the cooperative UAV is obtained at the same time; Based on the real-time status performance prediction index corresponding to the cooperative UAV, the cooperative UAV is assigned a cooperative task, and the cooperative task information is determined, and the cooperative residual energy index is set; The location information of the collaborative UAV is determined according to the real-time status performance prediction index, collaborative residual energy index and collaborative task information corresponding to the collaborative UAV.

[0050] Specifically, the collaborative residual energy index is the minimum real-time state performance prediction index that the collaborative UAV needs to retain when the collaborative focus UAV completes the accident location and processing tasks.

[0051] Furthermore, based on the focus drone location information, the coordinated drone location information and the drone reconnaissance video information, standard accident report information is determined, including: Determine the distribution strategy information according to the focus UAV location information and the cooperative UAV location information; Taking the focus drone position as the origin, construct the plane coordinate system of the accident area; Based on the distribution strategy information, the focus UAV location information and the cooperative UAV location information, the accident standard area information is determined; Use drones to detect video information to determine the accident object information, accident scene environment information and accident scene scope information; Conduct trend prediction through accident scene environment information and accident scene range information to obtain accident impact range information; According to the accident standard area information, the accident standard area is shadowed in the accident area plane coordinate system to obtain accident shadow graphic information; Determine the geometric center of the accident shadow figure, and obtain the coordinate information of the geometric center of the accident shadow figure as the accident center position information; Determine standard accident report information based on accident type information, detection timestamp information, accident center location information and accident object information.

[0052] Furthermore, an accident scene intelligent identification, positioning and processing system based on a drone library is proposed to implement any of the positioning and processing methods described above, including: The UAV resource management and information integration module is used to deeply analyze the geographic information of the control area, build a UAV library, plan the layout of the UAV library according to the type, performance and maintenance requirements of the UAV, obtain the basic information of the UAVs in the UAV library, and conduct preliminary detection and sorting of the UAV empty state information; UAV performance evaluation and mission planning module: The UAV performance evaluation and mission planning module is used to calculate the UAV no-load state performance index based on the UAV type and environmental factors, and determine the UAV initial state performance index based on the hardware, electrical and environmental data at startup. According to the UAV performance index and the characteristics of the control area, a UAV deployment list is formulated, the UAV reconnaissance route is planned, and the residual energy index and exception handling mechanism are set; The accident monitoring and emergency response module is used to receive drone reconnaissance videos, use image analysis technology to identify anomalies, determine the type of accident, handle the lack of accident location information, determine the location of the focus drone, and then determine the location information of the accident center. According to the accident information, the initial accident report and emergency response signal are generated and sent, and according to the location and signal of the focus drone, the location information of the coordinated drone is determined and dispatched; The accident handling and report generation module is used to determine the emergency handling plan based on the accident type and on-site conditions, dispatch and guide drones to assist in rescue operations, integrate various types of accident information, generate standard accident reports, store and distribute accident reports, and provide data support for accident analysis and subsequent decision-making.

[0053] Furthermore, the accident monitoring and emergency response module includes: The accident monitoring and positioning unit is used to receive drone reconnaissance videos, use image analysis technology to identify anomalies, determine the type of accident, handle the lack of accident location information, determine the location of the focus drone, and then determine the location information of the accident center; The emergency response unit is used to generate and send an initial accident report and emergency response signal based on the accident information, and determine the location information of the coordinated drone and dispatch it according to the location and signal of the focus drone.

[0054] Furthermore, the accident handling and report generation module includes: The accident handling unit is used to determine the emergency handling plan based on the accident type and on-site conditions, and dispatch and guide drones to assist in rescue operations; Report generation unit, the report generation unit is used to integrate various types of accident information, generate standard accident reports, store and distribute accident reports, and provide data support for accident analysis and subsequent decision-making.

[0055] The above shows and describes the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The above embodiments and descriptions only describe the principles of the present invention. The present invention may be subject to various changes and improvements without departing from the spirit and scope of the present invention. These changes and improvements fall within the scope of the present invention. The scope of protection claimed by the present invention is defined by the attached claims and their equivalents.

Claims

1. An intelligent identification, positioning and processing method for accident scenes based on a drone library, characterized in that: include: Determine the control area information and build a drone library based on the control area information; Obtain basic information of drones in the drone library, wherein the basic information of drones includes drone type information, drone number information, and drone parking location information; Based on the basic information of the drone, obtain the drone's empty state information and determine the drone's empty state performance index; Determine the drone deployment list information based on the drone basic information and the drone empty state performance index; Based on the drone deployment list information, start the drones in sequence, and obtain the drone initial state performance index when the drone is started. At the same time, combined with the control area information, determine the drone reconnaissance route information; Using the drone reconnaissance route information, the drone reconnaissance mission is performed, and the drone reconnaissance video information is obtained at the same time. Based on the drone reconnaissance video information, it is determined whether an accident has occurred. If not, the reconnaissance mission is completed based on the drone reconnaissance route information. If so, the accident type information is obtained based on the drone reconnaissance video information, and the focus drone location information is determined at the same time; Determine the initial accident information based on the accident type information, the focus drone location information, and the drone reconnaissance video information, and send the initial accident report information and emergency response signal; Determine the location information of the cooperative drone based on the location information of the focus drone and the emergency response signal; Determine standard accident report information based on the focus drone location information, the coordinated drone location information, and the drone reconnaissance video information; Send standard accident report information, determine emergency handling information, and guide drone information to properly handle the accident scene.

2. The method for intelligent identification, positioning and processing of accident scenes based on a drone library according to claim 1 is characterized in that: The method of obtaining the drone's empty state information based on the drone's basic information and determining the drone's empty state performance index specifically includes: According to the basic information of the drone, the drone type information is determined, and the drone library is initially divided into regions according to the drone type information to obtain the functional sub-region information of the drone library; Based on the functional sub-area information of the drone library, obtain the functional sub-area environment information; Perform secondary regional division on the functional sub-region based on the environmental information of the functional sub-region to obtain the environmental factor sub-region information; Determine the basic parameters of the drone type and the basic parameters of the environmental factors based on the drone type information, the functional sub-area information, the environmental factor sub-area information and the functional sub-area environmental information; Based on the basic parameters of the drone type and the basic parameters of the environmental factors, determine the drone type influencing sub-coefficient and the environmental factor influencing sub-coefficient; The sum of the UAV type influencing sub-coefficient and the environmental factor influencing sub-coefficient is taken as the UAV no-load state influencing coefficient; According to the drone's empty state information, determine the drone's mechanical performance information, drone's electrical performance information and environmental adaptability information; Determine the initial performance index of the drone in the empty state based on the drone's mechanical performance information, drone's electrical performance information and environmental adaptability information; Based on the UAV no-load state influence coefficient and the UAV no-load state initial performance index, the UAV no-load state performance index is determined.

3. The method for intelligent identification, positioning and processing of accident scenes based on a drone library according to claim 2 is characterized in that: The method starts the drones in sequence based on the drone deployment list information, obtains the drone initial state performance index when the drone is started, and determines the drone reconnaissance route information in combination with the control area information, specifically including: Obtain information related to the hardware status, electrical system, and environmental factors when the drone is started; Determine the hardware status adjustment coefficient, the electrical system adjustment coefficient and the environmental factor adjustment coefficient based on the hardware status related information, the electrical system related information and the environmental factor related information; Obtain the product value of the drone's no-load state performance index, hardware state adjustment coefficient, electrical system adjustment coefficient, and environmental factor adjustment coefficient to determine the initial state performance index; Based on the initial state performance index and drone type information, the drones are preliminarily classified into three groups of high, medium and low performance drones; According to the information of the controlled area, obtain the topographic and geomorphic feature information within the controlled area; Based on the topographical features in the control area, a spatial coordinate system of the performance attenuation area is constructed; According to the initial state performance index and the spatial coordinate system of the performance decay region, the initial UAV reconnaissance routes are constructed for the three groups of UAVs with high, medium and low performance; Based on the initial reconnaissance route of the UAV, the process of three groups of UAVs with high, medium and low performance performing reconnaissance tasks was simulated respectively, and the real-time state performance prediction index of the UAVs when performing reconnaissance tasks was obtained; Determine the coordinate values ​​of the drone in the spatial coordinate system of the performance degradation area and determine the location information of the nearest drone hangar; The drone reconnaissance residual energy index is set based on the location information of the nearest drone hangar and the coordinate values ​​of the drone in the spatial coordinate system of the performance attenuation area; If the real-time performance prediction index of the drone when performing the reconnaissance mission is greater than the drone's reconnaissance residual energy index, the drone will continue to perform the reconnaissance mission; If the real-time state performance prediction index of the drone when performing the reconnaissance mission is less than the drone reconnaissance residual energy index, the drone's initial reconnaissance route is marked as abnormal, and the drone's initial reconnaissance route is reconstructed; If the real-time state performance prediction index of the UAV when performing the reconnaissance mission is equal to the UAV reconnaissance residual energy index, the initial reconnaissance route of the UAV is used as the UAV reconnaissance route information.

4. The method for intelligent identification, positioning and processing of accident scenes based on a drone library according to claim 3 is characterized in that: The method of obtaining accident type information based on drone reconnaissance video information and determining the focus drone location information specifically includes: Based on the drone reconnaissance video information, obtain the drone reconnaissance frame image; Analyze the drone reconnaissance frame images to obtain abnormal reconnaissance frame images; Obtain abnormal type information based on abnormal detection frame images; Determine the detection timestamp information of the abnormal detection frame image through the abnormal type information; Determine whether the reconnaissance timestamp information is empty. If so, traverse the drone reconnaissance frame images, and obtain the abnormal reconnaissance frame associated images, and at the same time associate and integrate the abnormal reconnaissance frame associated images with the abnormal reconnaissance frame images to obtain the accident type judgment image set and the first emergency response information. If not, obtain the abnormal reconnaissance frame difference image based on the reconnaissance frame timestamp information, and at the same time associate and integrate the abnormal reconnaissance frame difference image with the abnormal reconnaissance frame image to obtain the accident type judgment image set; Determine the accident type information through the accident type judgment image set, and obtain the accident geographical location information in the accident type judgment image set; Determine whether the accident geographic location information is empty. If so, determine the image set based on the accident type, obtain the accident geographic feature information, obtain the accident geographic location area information, and determine the accident geographic location prediction information and the second emergency response information. If not, use the accident geographic location information as the accident geographic location prediction information; The accident geographic location prediction information is used as the focus UAV location information.

5. The method for intelligent identification, positioning and processing of accident scenes based on a drone library according to claim 4 is characterized in that: The method of determining the initial accident information based on the accident type information, the focus drone location information and the drone reconnaissance video information, and sending the initial accident report information and the emergency response signal specifically includes: Determine the accident impact level information based on the accident type information and drone reconnaissance video information; Determine the first accident coding information based on the accident impact level information and the accident type information; Determine the second accident coding information based on the focus drone location information; Using drones to detect video information, determine accident feature information, and perform data processing on the accident feature information to obtain third accident coding information; Based on the accident impact level information and the accident type information, communication channel information, communication method information and communication protocol information are determined, and the first accident coding information, the second accident coding information and the third accident coding information are digitally signed to obtain initial accident coding information; Construct a preset report template, and input the accident type information, the focus drone location information, the accident impact level information and the accident feature information into the preset report template to obtain the initial accident text information; Integrate the initial accident code information and the initial accident text information to obtain the initial accident information, and determine and send the initial accident report information; Based on the initial accident coding information, the first emergency response information and the second emergency response information, an emergency response signal is set, and the initial accident report information is data processed to obtain initial accident report backup information; Based on the accident impact level information, emergency response scope information is determined, and based on the emergency response scope information, initial accident report backup information and an emergency response signal are sent.

6. The method for intelligent identification, positioning and processing of accident scenes based on a drone library according to claim 5 is characterized in that: The determining of the cooperative drone position information based on the focus drone position information and the emergency response signal specifically includes: Based on the focus UAV position information, determine the focus UAV attitude information; According to the focus UAV attitude information, determine the focus UAV tilt angle information and heading angle information; Obtain the sending timestamp information of the emergency response signal, and determine the straight-line distance information between the focus UAV and the coordinated UAV by combining the focus UAV position information, focus UAV tilt angle information and heading angle information; Based on the focus UAV position information and the straight-line distance information between the focus UAV and the cooperative UAV, the response position information of the cooperative UAV is determined; The coordinates of the cooperative UAV in the spatial coordinate system of the performance attenuation area are determined based on the response position information of the cooperative UAV, and the corresponding real-time state performance prediction index of the cooperative UAV is obtained at the same time; Based on the real-time status performance prediction index corresponding to the cooperative UAV, the cooperative UAV is assigned a cooperative task, and the cooperative task information is determined, and the cooperative residual energy index is set; The location information of the collaborative UAV is determined according to the real-time status performance prediction index, collaborative residual energy index and collaborative task information corresponding to the collaborative UAV.

7. The method for intelligent identification, positioning and processing of accident scenes based on a drone library according to claim 6 is characterized in that: The standard accident report information is determined based on the focus drone position information, the coordinated drone position information and the drone reconnaissance video information, specifically including: Determine the distribution strategy information according to the focus UAV location information and the cooperative UAV location information; Taking the focus drone position as the origin, construct the plane coordinate system of the accident area; Based on the distribution strategy information, the focus UAV location information and the cooperative UAV location information, the accident standard area information is determined; Use drones to detect video information to determine the accident object information, accident scene environment information and accident scene scope information; Conduct trend prediction through accident scene environment information and accident scene range information to obtain accident impact range information; According to the accident standard area information, the accident standard area is shadowed in the accident area plane coordinate system to obtain accident shadow graphic information; Determine the geometric center of the accident shadow figure, and obtain the coordinate information of the geometric center of the accident shadow figure as the accident center position information; Determine standard accident report information based on accident type information, detection timestamp information, accident center location information and accident object information.

8. An accident scene intelligent identification, positioning and processing system based on a drone library, used to implement the positioning and processing method according to any one of claims 1 to 7, characterized in that: include: A drone resource management and information integration module, which is used to deeply analyze the geographic information of the control area, build a drone library, plan the layout of the drone library according to the type, performance and maintenance requirements of the drones, obtain the basic information of the drones in the drone library, and perform preliminary detection and sorting of the drone empty state information; UAV performance evaluation and mission planning module, which is used to calculate the UAV no-load state performance index based on the UAV type and environmental factors, determine the UAV initial state performance index based on the hardware, electrical and environmental data at startup, formulate a UAV deployment list based on the UAV performance index and the characteristics of the control area, plan the UAV reconnaissance route, and set the residual energy index and exception handling mechanism; The accident monitoring and emergency response module is used to receive drone reconnaissance videos, use image analysis technology to identify anomalies, determine the type of accident, handle the lack of accident location information, determine the location of the focus drone, and then determine the location information of the accident center, generate and send an initial accident report and emergency response signal based on the accident information, and determine the location information of the coordinated drone based on the focus drone position and signal and dispatch it; The accident handling and report generation module is used to determine the emergency handling plan based on the accident type and on-site conditions, dispatch and guide drones to assist in rescue operations, integrate various types of accident information, generate standard accident reports, store and distribute accident reports, and provide data support for accident analysis and subsequent decision-making.

9. The accident scene intelligent identification, positioning and processing system based on the drone library according to claim 8 is characterized in that: The accident monitoring and emergency response module includes: An accident monitoring and positioning unit, which is used to receive drone reconnaissance videos, use image analysis technology to identify anomalies, determine the type of accident, handle the lack of accident location information, determine the focus drone location, and then determine the accident center location information; The emergency response unit is used to generate and send an initial accident report and an emergency response signal based on the accident information, and determine the location information of the coordinated drone and dispatch it according to the location and signal of the focus drone.

10. The accident scene intelligent identification, positioning and processing system based on the drone library according to claim 8 is characterized in that: The accident handling and report generation module includes: An accident handling unit, which is used to determine an emergency handling plan based on the type of accident and the on-site situation, and dispatch and guide drones to assist in rescue operations; The report generation unit is used to integrate various types of accident information, generate standard accident reports, store and distribute accident reports, and provide data support for accident analysis and subsequent decision-making.

Images