GIS-based unmanned aerial vehicle collaborative inspection scheduling method and system

Abstract

The invention relates to the field of unmanned aerial vehicles, in particular to a GIS-based unmanned aerial vehicle collaborative inspection scheduling method and system, which comprises a task management module, a GIS data management module, an intelligent scheduling module, an unmanned aerial vehicle control module, a data acquisition module and an early warning module, the task management module is used for issuing initial task information or emergency task information to the intelligent scheduling module, the GIS data management module is used for storing related data of each inspection position, and the intelligent scheduling module is used for generating an inspection path of the inspection unmanned aerial vehicle and generating an emergency inspection route of the emergency unmanned aerial vehicle. The unmanned aerial vehicle control module is used for controlling the inspection unmanned aerial vehicle or the emergency unmanned aerial vehicle, the data acquisition module is used for shooting various images required by tasks, and the early warning module is used for obtaining risk indexes of all inspection positions. According to the scheme, the regional risk is measured by integrating factors in all aspects, the judgment accuracy degree is improved, and the intelligent degree of the system is improved.

Description

Technical Field

[0001] The present invention relates to the field of unmanned aerial vehicles (UAVs), and in particular to a GIS-based UAV collaborative inspection and scheduling method and system. Background Art

[0002] Currently, drones are increasingly being used to inspect target areas in areas such as power lines, oil pipelines, forest fire prevention, and traffic monitoring to improve operational efficiency and reduce labor costs.

[0003] For example, the prior art disclosed in CN117572806A discloses an automatic homing control system for a GIS geographic monitoring drone, which relates to the field of geographic monitoring drone technology. The GIS geographic monitoring drone automatic homing control system includes a flight control module, an automatic battery replacement and charging module, an automatic repair module, an image acquisition module, a positioning module, a data collection module, a server terminal module, and a data processing module. The flight control module is connected to the drone body by signal, the automatic battery replacement and charging module is located inside the drone's nest, the data collection module and the data processing module form a server terminal module, the positioning module and the image acquisition module are both installed on the drone body, and the image acquisition module is wirelessly connected to the data collection module and the data processing module. Inspection points are determined based on the area of the geographic location being inspected, and the drone is flown to each inspection point one by one.

[0004] Another typical example is a GIS-based UAV plant protection system and method disclosed in the prior art CN108717301B. The GIS-based UAV plant protection system includes a UAV and a monitoring terminal. The UAV is provided with a positioning unit, a depth camera, and a flight execution unit. The monitoring terminal includes a GIS processing unit, a manual control unit, an automatic control unit, and a control mode switching unit. The depth camera collects depth point cloud data of the farmland and sends it to the GIS processing unit. The GIS processing unit analyzes the depth point cloud data and displays the areas where crops are planted in the farmland on the GIS map.

[0005] Next, let's look at a GIS-based UAV sea area monitoring method disclosed in the prior art CN112218051A. The method includes the following steps: obtaining UAV measurement and control related parameter fields in the form of a configuration file on the GIS system; for a specific UAV, receiving the UAV's real-time video data and telemetry data generated and stored by the flight control system;

[0006] For video data, a grayscale projection-based method is used to achieve rapid motion estimation of video frames to form video motion coding; for telemetry data, the UAV payload status data and payload status data analyzed from the telemetry data are used to encode the motion patterns of the UAV and payload with time as the axis to obtain telemetry motion coding; and synchronization of telemetry data and video data is achieved.

[0007] Existing drone scheduling often lacks intelligent support and cannot be flexibly adjusted according to dynamic factors such as regional risks. In order to solve the common problems in this field, the present invention was made. Summary of the Invention

[0008] The purpose of the present invention is to address the current deficiencies and propose a GIS-based UAV collaborative inspection scheduling method and system.

[0009] In order to overcome the deficiencies of the prior art, the present invention adopts the following technical solutions:

[0010] A GIS-based UAV collaborative inspection and scheduling system is characterized by including a task management module, a GIS data management module, an intelligent scheduling module, a UAV control module, a data acquisition module and an early warning module. The task management module is used to send initial task information or emergency task information to the intelligent scheduling module, the GIS data management module is used to store relevant data of each inspection location, the intelligent scheduling module generates an inspection path for the inspection UAV based on the information provided by the GIS data management module and the initial task information, or generates an emergency inspection route for the emergency UAV based on the emergency task information, the UAV control module is used to control the inspection UAV according to the inspection path generated by the intelligent scheduling module or control the emergency UAV according to the emergency inspection route, the data acquisition module is used to capture various images required for the task during the flight of the inspection UAV, the early warning module is used to obtain the risk index of each inspection location based on the data collected by the data acquisition module, and judge whether it is necessary to send an early warning signal to the task management module based on the risk index; the early warning signal is used for the task management module to select and send the emergency task information to the intelligent scheduling module.

[0011] Furthermore, the task management module includes an input unit, a display unit and a task configuration unit, wherein the input unit is used to receive user instructions, the display unit is used to display the analysis results of the early warning module, and the task configuration unit is used to set the target area, task type, time requirement, etc. of the inspection task;

[0012] Furthermore, the GIS data management module includes a geographic information storage unit and a network connection unit. The geographic information storage unit is used to store and manage remote sensing images and topographic maps. The network connection unit is used to receive third-party data such as real-time meteorological information and traffic information and mark and update remote sensing images and topographic maps based on the third-party data.

[0013] Furthermore, the intelligent scheduling module includes a drone library, a drone resource management unit and a path planning unit. The drone library is used to store drones, the drone resource management unit is used to manage the power of each drone and its position in the drone library, and the path planning unit is used to generate an inspection path for each drone based on the information stored in the GIS data management module and the initial task information.

[0014] Furthermore, the drone control module includes an instruction parsing unit, a mobile control unit and a posture adjustment unit. The instruction parsing unit is used to parse the training path obtained by the path planning unit and send corresponding control signals to the mobile control unit and the posture adjustment unit. The mobile control unit is used to control the drone to move according to the control signal. The posture adjustment unit is used to control the drone to take pictures at a designated inspection position according to the control signal. The data acquisition module includes a camera, which is set on the drone and is used to take images of the inspection position.

[0015] Furthermore, the early warning module includes an image preprocessing unit, an image recognition unit, a risk judgment model, a calculation unit and an alarm unit. The image preprocessing unit is used to amplify and reduce noise on the captured image. The image recognition unit is used to identify the object type and size of each part of the preprocessed image. The risk judgment model is used to obtain corresponding risk parameters based on the recognized image. The calculation unit is used to calculate the risk index based on the recognition result of the image recognition unit, the risk parameters obtained by the risk judgment model and past records. When the risk index is greater than the risk index threshold, the alarm unit sends a warning signal to the task management module.

[0016] A GIS-based UAV collaborative inspection scheduling method includes the following steps:

[0017] S1, the task management module sends the initial task information to the intelligent scheduling module;

[0018] S2, the intelligent scheduling module generates the inspection route of the inspection drone based on the initial task information and the information provided by the GIS data management module;

[0019] S3, the drone control module controls the inspection drone to perform inspections according to the generated inspection path, and the data acquisition module captures the target image at the designated inspection location;

[0020] S4, the early warning module obtains the risk index based on the data collected by the data collection module, and determines whether the risk index is greater than the risk index threshold. If not, the process ends; otherwise, the region corresponding to the risk index is considered a high-risk region and the next step is executed;

[0021] S5, the inspection drone continues to inspect along the original inspection route, the early warning module sends a signal to the task management module, the task management module sends the emergency task information to the intelligent scheduling module, the intelligent scheduling module generates an emergency inspection route based on the high-risk areas obtained by the early warning module, the drone control module controls the emergency drone to reach the high-risk area according to the emergency inspection route, and the emergency drone collects various data in the high-risk area and feeds back to the user.

[0022] Furthermore, the early warning module obtains risk indicators by the following steps:

[0023] S41, the image preprocessing unit amplifies and reduces noise on the image captured by the data acquisition module;

[0024] S42, the image recognition unit recognizes the type and size of the object in each part of the preprocessed image;

[0025] S43, the risk judgment model obtains risk parameters based on the recognized image;

[0026] S44, the calculation unit calculates the risk index according to the risk parameter, the recognition result of the image recognition unit, the risk parameter and the past records.

[0027] The beneficial effects achieved by the present invention are: 1. By combining the recognition results of the image recognition unit, the judgment results of the risk judgment model and past data to calculate the risk index, it is beneficial to comprehensively measure regional risks based on various factors, which is beneficial to improve the accuracy of judgment and the intelligence of the system.

[0028] 2. Two types of drones are used for work. Inspection drones are used for inspection, which is conducive to making preliminary judgments on regional risks through images. It can be applied to most work tasks. When the risk is high, emergency drones are dispatched for detailed data collection, which is conducive to staff combining further collected data with the original image data to make detailed judgments on regional risks and confirm the possibility of actual risk occurrence. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] The present invention can be further understood from the following description in conjunction with the accompanying drawings. The components in the figures are not necessarily drawn to scale, but rather the emphasis is placed on illustrating the principles of the embodiments. In different views, the same reference numerals designate the same parts.

[0030] Figure 1 It is a structural schematic diagram of the present invention.

[0031] Figure 2 It is the workflow diagram of the present invention.

[0032] Figure 3 This is a flow chart of the early warning module of the present invention for obtaining risk indicators.

[0033] Figure 4 This is a graph showing the relationship between historical data parameters, the total number of evaluation results, and the number of reasonable evaluation results. DETAILED DESCRIPTION

[0034] The following is an explanation of the embodiments of the present invention through specific embodiments. Those skilled in the art can understand the advantages and effects of the present invention from the contents disclosed in this specification. The present invention can be implemented or applied through other different specific embodiments, and the details in this specification can also be modified and changed in various ways based on different viewpoints and applications without departing from the spirit of the present invention. In addition, the drawings of the present invention are only simple schematic illustrations and are not depicted in actual size. It is stated in advance. The following embodiments will further explain the relevant technical content of the present invention in detail, but the disclosed content is not intended to limit the scope of protection of the present invention.

[0035] Example 1: According to Figure 1 、 Figure 2 、 Figure 3 and Figure 4, this embodiment provides a GIS-based UAV collaborative inspection and scheduling system, including a task management module, a GIS data management module, an intelligent scheduling module, a UAV control module, a data acquisition module and an early warning module, the task management module is used to send initial task information or emergency task information to the intelligent scheduling module, the GIS data management module is used to store relevant data of each inspection location, the intelligent scheduling module generates an inspection path for the inspection UAV according to the information provided by the GIS data management module and the initial task information, or generates an emergency inspection route for the emergency UAV according to the emergency task information, the UAV control module is used to control the inspection UAV according to the inspection path generated by the intelligent scheduling module or control the emergency UAV according to the emergency inspection route, the data acquisition module is used to capture various images required for the task during the flight of the inspection UAV, the early warning module is used to obtain the risk index of each inspection location based on the data collected by the data acquisition module, and judge whether it is necessary to send an early warning signal to the task management module based on the risk index; the early warning signal is used for the task management module to select and send the emergency task information to the intelligent scheduling module.

[0036] Furthermore, the task management module includes an input unit, a display unit and a task configuration unit, wherein the input unit is used to receive user instructions, the display unit is used to display the analysis results of the early warning module, and the task configuration unit is used to set the target area, task type, time requirement, etc. of the inspection task;

[0037] Furthermore, the GIS data management module includes a geographic information storage unit and a network connection unit. The geographic information storage unit is used to store and manage remote sensing images and topographic maps. The network connection unit is used to receive third-party data such as real-time meteorological information and traffic information and mark and update remote sensing images and topographic maps based on the third-party data.

[0038] Furthermore, the intelligent scheduling module includes a drone library, a drone resource management unit and a path planning unit. The drone library is used to store drones, the drone resource management unit is used to manage the power of each drone and its position in the drone library, and the path planning unit is used to generate an inspection path for each drone based on the information stored in the GIS data management module and the initial task information.

[0039] Specifically, the path planning unit generates an inspection path for the UAV through an existing path planning algorithm (TSP, A-star, etc.).

[0040] Furthermore, the drone control module includes an instruction parsing unit, a mobile control unit and a posture adjustment unit. The instruction parsing unit is used to parse the training path obtained by the path planning unit and send corresponding control signals to the mobile control unit and the posture adjustment unit. The mobile control unit is used to control the drone to move according to the control signal. The posture adjustment unit is used to control the drone to take pictures at a designated inspection position according to the control signal. The data acquisition module includes a camera, which is set on the drone and is used to take images of the inspection position.

[0041] Furthermore, the early warning module includes an image preprocessing unit, an image recognition unit, a risk judgment model, a calculation unit and an alarm unit. The image preprocessing unit is used to amplify and reduce noise on the captured image. The image recognition unit is used to identify the object type and size of each part of the preprocessed image. The risk judgment model is used to obtain corresponding risk parameters based on the recognized image. The calculation unit is used to calculate the risk index based on the recognition result of the image recognition unit, the risk parameters obtained by the risk judgment model and past records. When the risk index is greater than the risk index threshold, the alarm unit sends a warning signal to the task management module.

[0042] Specifically, the image recognition unit uses existing image recognition models (such as VGG, SSD and U-Net, etc.) to identify the types of objects in the image, such as mountains, vegetation, humans, etc.; specifically, the recognition content of the risk judgment model (SVR, CNN and XGBoost, etc.) is set by those skilled in the art according to the tasks they want the drone to perform. The risk judgment model is trained by those skilled in the art according to the tasks they want the drone to perform. For example, if they want to analyze the risk of landslides in various places, the recognition content is mountains. By taking pictures of mountains, the model uses the image after image recognition as input and the risk parameter as output. The value range of the risk parameter is 100%. The range is 0 to 5. The larger the parameter, the higher the probability of the risk considered by the model to be. If one wants to analyze the risk of landslides in various places, the training set is images taken before the landslides occurred in places obtained from various sources, as well as ordinary mountain images. The technical personnel in this field assign scores to the images in these training sets based on their experience (0 to 5, the larger the value, the more likely a landslide is to occur), thereby obtaining the corresponding risk judgment model. For example, if one wants to analyze the risk of insect pests, the identification content is trees, and the risk parameter also ranges from 0 to 5. The technical personnel in this field assign scores to the images in the training set of trees containing insect pests (0 to 5, the larger the value, the more serious the insect pest situation).

[0043] Specifically, the risk index threshold is set by technical personnel in this field according to the severity of disasters corresponding to different images in the training set. For example, if the analysis content is insect pests, images with scores greater than 3 in the training images are generally considered by the staff to be serious, while images with scores less than or equal to 3 are generally considered by the staff to be mild, then the threshold can be set to 3. For example, if the analysis content is wildfires, images with scores greater than 4 in the training images are generally considered by the staff to be prone to wildfires, while images with scores less than or equal to 4 are generally considered by the staff to be unlikely to be wildfires, then the threshold can be set to 4. The risk index threshold is the boundary value for accidents. When the risk index is greater than the risk index threshold, it is considered that the inspection location is more likely to have a corresponding accident. Otherwise, it is considered less likely.

[0044] A GIS-based UAV collaborative inspection scheduling method includes the following steps:

[0045] S1, the task management module sends the initial task information to the intelligent scheduling module;

[0046] S2, the intelligent scheduling module generates the inspection route of the inspection drone based on the initial task information and the information provided by the GIS data management module;

[0047] S3, the drone control module controls the inspection drone to perform inspections according to the generated inspection path, and the data acquisition module captures the target image at the designated inspection location;

[0048] S4, the early warning module obtains the risk index based on the data collected by the data collection module, and determines whether the risk index is greater than the risk index threshold. If not, the process ends; otherwise, the region corresponding to the risk index is considered a high-risk region and the next step is executed;

[0049] S5, the inspection drone continues to inspect along the original inspection route, the early warning module sends a signal to the task management module, the task management module sends the emergency task information to the intelligent scheduling module, the intelligent scheduling module generates an emergency inspection route based on the high-risk areas obtained by the early warning module, the drone control module controls the emergency drone to reach the high-risk area according to the emergency inspection route, and the emergency drone collects various data in the high-risk area and feeds back to the user.

[0050] Specifically, by dispatching emergency drones to collect data from high-risk areas without changing the original inspection drones' inspection routes, the inspection drones can continue to inspect other areas. When there are multiple high-risk areas, relying on a single inspection drone cannot fully collect data from each area (such as soil samples). It is beneficial to use multiple emergency drones to collect data from high-risk areas separately, increasing the amount of data that can be collected. The intervention of emergency drones also improves the data collection and risk warning capabilities of the entire system.

[0051] Specifically, the main function of the inspection drone is image capture. By taking images of each inspection location, it can be used for various risk identifications, such as landslides, wildfires, etc., and has high versatility. The main function of the emergency drone is to collect detailed data required for a certain risk judgment. For example, landslides require the collection of corresponding soil, insect pests require the collection of corresponding leaves, and wildfires require the detection of smoke concentration in the air.

[0052] Furthermore, the early warning module obtains risk indicators by the following steps:

[0053] S41, the image preprocessing unit amplifies and reduces noise on the image captured by the data acquisition module;

[0054] S42, the image recognition unit recognizes the type and size of the object in each part of the preprocessed image;

[0055] S43, the risk judgment model obtains risk parameters based on the recognized image;

[0056] S44, the calculation unit calculates the risk index according to the risk parameter, the recognition result of the image recognition unit, the risk parameter and the past records.

[0057] Specifically, the risk index can be calculated according to the following formula:

[0058]

[0059] Among them, FX is the risk index, CS is the risk parameter obtained by the risk judgment model, Q is the image accuracy parameter, F is the historical data parameter, which is used to characterize the reliability of the historical assessment results of the risk assessment model, and w1 is the balance coefficient, which is used to adjust the weight of the image accuracy parameter in the calculation. The weight is set by those skilled in the art between 0.3 and 0.7 (the common weight value range is 0 to 1, and this solution sets the weight to 0.7 to 0.3 to avoid being too extreme) according to the image accuracy required for the risk type to be assessed. If the image quality is very critical in the risk assessment (such as high-resolution soil images are crucial in the judgment of landslides), then w1 is larger (such as w1 = 0.7), and the weight is assigned to the image accuracy. Q is given a higher weight if the image quality has little impact on the risk assessment (for example, when judging the risk of tree fall, it is easy to judge even if the image quality is poor); w2 is the floating range adjustment coefficient, which is used to control the uncertainty range of the risk indicator and reflect the reliability of the prediction. The value of this coefficient is 0.5 or 1. This coefficient can be set by those skilled in the art according to the type of risk to be assessed. If the assessed risk type is more harmful, such as wildfire, the coefficient is set to 1 to expand the floating range to cover extreme cases. If the assessed risk type is more harmful, such as insect pests, the coefficient is set to 0.5 to reduce the floating range and provide a more accurate prediction. ε is used to avoid the minimum value of the denominator being 0, and its value is generally 0.001;

[0060] e is a natural constant, ver is an averaging function, L is the standard deviation of the Laplace gradient of the recognized image (the larger the value, the better the image quality), which is obtained by the image recognition unit through OpenCV, l is the lower limit threshold of the gradient difference, which is set by those skilled in the art according to the required image accuracy, M is the number of pixels in each row of the processed image, N is the number of pixels in each column of the processed image, XS a is the gray value of the a-th pixel, xs a is the pixel value after convolution smoothing corresponding to the a-th pixel, and NUM is the number of pixels occupied by the unrecognizable object types in the recognition image;

[0061] Num is the total number of risk parameters obtained from past assessments of the risk judgment model. Those skilled in the art will regularly review and evaluate the risk parameters assessed by the risk judgment model. Reasonable risk parameters will be used as new training sets to train the model, thereby continuously improving the judgment ability of the model. The review frequency is preferably once a week and can also be adjusted by those skilled in the art according to actual application conditions. Right is the number of risk parameters that the staff considers reasonable among the num assessment results, and e is a natural constant.

[0062] The following is the procedure required to calculate the risk indicator:

[0063]

[0064]

[0065] return round(fx_base,3),round(fx_lower,3),round(fx_upper,3)

[0066] def_calculate_q(self,image_path,l=30):

[0067] """

[0068] Calculate the image quality parameter Q

[0069] :param image_path: image path

[0070] :paraml: Gradient difference lower limit threshold (default 30)

[0071] """

[0072] #1. Read the image and convert it to grayscale

[0073] img=cv2.imread(image_path,cv2.IMREAD_GRAYSCALE)

[0074] ifimgis None:

[0075] raise ValueError("Unable to read image file")

[0076] m,n=img.shape

[0077] #2. Calculate the Laplace gradient

[0078] laplacian=cv2.Laplacian(img,cv2.CV_64F)

[0079] L = np.std(laplacian) # gradient standard deviation

[0080] #3. Smoothing (Gaussian filtering)

[0081] smoothed=cv2.GaussianBlur(img,(5,5),0)

[0082] #4. Calculate pixel differences

[0083] diff=np.abs(img.astype(float)-smoothed.astype(float))

[0084] avg_diff=np.sum(diff) / (m*n*255)

[0085] #5. Number of unrecognizable pixels (assuming thresholding)

[0086] _,binary=cv2.threshold(img,127,255,cv2.THRESH_BINARY)

[0087] num_unrecognized=np.sum(binary==0)

[0088] #6. Comprehensive calculation of Q

[0089]

[0090] like Figure 4 As shown, Figure 4 This is a graph showing the relationship between historical data parameters, the total number of evaluation results (num), and the number of reasonable evaluation results (right).

[0091] The beneficial effects of this solution are: 1. By combining the recognition results of the image recognition unit, the judgment results of the risk judgment model and past data to calculate the risk index, it is beneficial to comprehensively measure regional risks based on various factors, which is beneficial to improve the accuracy of judgment and the intelligence of the system.

[0092] 2. Two types of drones are used for work. Inspection drones are used for inspection, which is conducive to making preliminary judgments on regional risks through images. It can be applied to most work tasks. When the risk is high, emergency drones are dispatched for detailed data collection, which is conducive to staff combining further collected data with the original image data to make detailed judgments on regional risks and confirm the possibility of actual risk occurrence.

[0093] Example 2: This example should be understood to include all the features of any of the aforementioned examples, and further improve upon them, and also include a method for predicting future risks in various regions and adjusting inspection shifts based on future risks. This method is applicable to situations where the risk index is greater than the risk index threshold and the user receives the data collected by the emergency drone and determines that no manual intervention is temporarily required in the region (i.e., the situation is not particularly serious). The judgment mentioned here is a manual judgment; this method calculates the predicted risk change rate index of a region using the following formula:

[0094]

[0095] Among them, YC is the predicted risk change rate indicator, which is used to measure the predicted change of the risk indicator. The larger the indicator, the greater the predicted change. fx is the predicted risk indicator. I is the number of data collected by the emergency drone that requires the user to judge and analyze. i is the excess degree parameter of the i-th data, MIN i The lower limit of the recommended numerical range of the exceedance parameter of the i-th data, MAX i CS is the upper limit of the recommended numerical range of the exceedance parameter of the i-th data. i is the actual value of the i-th data. The recommended numerical range is set by those skilled in the art based on experience, known constants, and existing experimental data (for example, for sandy soil, it is known from the Code for Investigation of Geotechnical Engineering (GB 50021) that when the soil moisture content is greater than 30%, drainage reinforcement measures are required. At this time, it is considered that there is a high risk of landslide. The upper limit is 30% and the lower limit is 0%). FX is the currently obtained risk index, X is the number of risk indicators obtained in the past for the area, and day x The number of days between the risk index obtained in the past and the current one, e is a natural constant, FX x is the xth risk indicator obtained in the past.

[0096] Specifically, when the predicted risk change rate index is greater than or equal to 10% (set by technical personnel in this field according to the required judgment accuracy, when higher accuracy is required, the value can be appropriately reduced, and when the accuracy requirement is lower, the value can be appropriately increased), it is considered that the risk is likely to become serious, and the inspection shifts of the inspection drone need to be increased to deal with emergencies. When the predicted risk change rate is less than 10% and greater than or equal to 0, it is considered that the risk is unlikely to become serious, and there is no need to adjust the inspection shifts of the drone. When the predicted risk change rate is less than 0, it is considered that the risk is likely to decrease, and there is no need to adjust the inspection shifts of the drone.

[0097] Beneficial effects of this embodiment: By introducing the predicted risk change rate indicator, this solution can dynamically predict future risk trends based on historical risk data and multi-dimensional data fed back by emergency drones. When the predicted change rate exceeds a threshold (e.g., 10%), the inspection shifts are automatically increased to achieve forward-looking risk management and avoid the time lag problem of responding after a disaster occurs.

[0098] The above disclosure is only a preferred embodiment of the present invention and does not limit the scope of protection of the present invention. Therefore, any equivalent technical changes made by using the contents of the present invention and the drawings are included in the scope of protection of the present invention. In addition, as technology develops, the elements therein may be updated. The above units are only examples. Those skilled in the art can make different designs and adopt corresponding units according to actual needs when implementing this solution.

Claims

1. A GIS-based UAV collaborative inspection and dispatching system, characterized by: It includes a task management module, a GIS data management module, an intelligent scheduling module, a UAV control module, a data acquisition module and an early warning module. The task management module is used to send initial task information or emergency task information to the intelligent scheduling module. The GIS data management module is used to store relevant data of each inspection location. The intelligent scheduling module generates an inspection path for the inspection UAV according to the information provided by the GIS data management module and the initial task information, or generates an emergency inspection route for the emergency UAV according to the emergency task information. The UAV control module is used to control the inspection UAV according to the inspection path generated by the intelligent scheduling module or control the emergency UAV according to the emergency inspection route. The data acquisition module is used to capture various images required for the task during the flight of the inspection UAV. The early warning module is used to obtain the risk index of each inspection location based on the data collected by the data acquisition module, and determine whether it is necessary to send an early warning signal to the task management module based on the risk index; The warning signal is used by the task management module to select and send emergency task information to the intelligent scheduling module.

2. The GIS-based UAV collaborative inspection and dispatching system according to claim 1 is characterized in that: The task management module includes an input unit, a display unit and a task configuration unit. The input unit is used to receive user instructions, the display unit is used to display the analysis results of the early warning module, and the task configuration unit is used to set content including but not limited to the target area, task type, and time requirement of the inspection task.

3. The GIS-based UAV collaborative inspection and dispatching system according to claim 1 is characterized in that: The GIS data management module includes a geographic information storage unit and a network connection unit. The geographic information storage unit is used to store and manage remote sensing images and topographic maps. The network connection unit is used to receive third-party data such as real-time meteorological information and traffic information and mark and update remote sensing images and topographic maps based on the third-party data.

4. The GIS-based UAV collaborative inspection and dispatching system according to claim 1 is characterized in that: The intelligent scheduling module includes a drone library, a drone resource management unit and a path planning unit. The drone library is used to store drones, the drone resource management unit is used to manage the power of each drone and its position in the drone library, and the path planning unit is used to generate the inspection path of each drone based on the information stored in the GIS data management module and the initial task information.

5. The GIS-based UAV collaborative inspection and dispatching system according to claim 4 is characterized in that: The drone control module includes an instruction parsing unit, a mobile control unit and a posture adjustment unit. The instruction parsing unit is used to parse the training path obtained by the path planning unit and send corresponding control signals to the mobile control unit and the posture adjustment unit. The mobile control unit is used to control the drone to move according to the control signal. The posture adjustment unit is used to control the drone to take pictures at a designated inspection position according to the control signal. The data acquisition module includes a camera, which is set on the drone and is used to take images of the inspection position.

6. The GIS-based UAV collaborative inspection and dispatching system according to claim 1 is characterized in that: The early warning module includes an image preprocessing unit, an image recognition unit, a risk judgment model, a calculation unit and an alarm unit. The image preprocessing unit is used to amplify and reduce noise on the captured image. The image recognition unit is used to identify the object type and size of each part of the preprocessed image. The risk judgment model is used to obtain corresponding risk parameters based on the recognized image. The calculation unit is used to calculate the risk index based on the recognition result of the image recognition unit, the risk parameters obtained by the risk judgment model and past records. When the risk index is greater than the risk index threshold, the alarm unit sends a warning signal to the task management module.

7. A GIS-based UAV collaborative inspection scheduling method, which is applied to the GIS-based UAV collaborative inspection scheduling system according to claim 1, characterized in that: The method comprises the following steps: S1, the task management module sends the initial task information to the intelligent scheduling module; S2, the intelligent scheduling module generates the inspection route of the inspection drone based on the initial task information and the information provided by the GIS data management module; S3, the drone control module controls the inspection drone to perform inspections according to the generated inspection path, and the data acquisition module captures the target image at the designated inspection location; S4, the early warning module obtains the risk index based on the data collected by the data collection module, and determines whether the risk index is greater than the risk index threshold. If not, the process ends; otherwise, the region corresponding to the risk index is considered a high-risk region and the next step is executed; S5, the inspection drone continues to inspect along the original inspection route, the early warning module sends a signal to the task management module, the task management module sends the emergency task information to the intelligent scheduling module, the intelligent scheduling module generates an emergency inspection route based on the high-risk areas obtained by the early warning module, the drone control module controls the emergency drone to reach the high-risk area according to the emergency inspection route, and the emergency drone collects various data in the high-risk area and feeds back to the user.

8. The GIS-based UAV collaborative inspection and scheduling method according to claim 7 is characterized in that: The early warning module obtains risk indicators in the following steps: S41, the image preprocessing unit amplifies and reduces noise on the image captured by the data acquisition module; S42, the image recognition unit recognizes the type and size of each object in the pre-processed image; S43, the risk judgment model obtains risk parameters based on the recognized image; S44, the calculation unit calculates the risk index according to the risk parameter, the recognition result of the image recognition unit, the risk parameter and the past records.

Images