Unmanned aerial vehicle autonomous inspection method and device based on space perception field and unmanned aerial vehicle

Abstract

The invention relates to an unmanned aerial vehicle autonomous inspection method and device based on a space perception field and an unmanned aerial vehicle. In the process that the unmanned aerial vehicle flies along the initial inspection route, first space radar data and first space image data collected by the unmanned aerial vehicle at the initial inspection position are obtained for each to-be-inspected object, and target space point cloud data corresponding to the to-be-inspected object are determined according to the first space radar data and the first space image data; determining second space image data according to the current tour-inspection position of the unmanned aerial vehicle and second space radar data acquired by the unmanned aerial vehicle at the current tour-inspection position, generating a space sensing field of the unmanned aerial vehicle according to the target space point cloud data and the second space image data, and adjusting the initial tour-inspection route according to the space sensing field, and obtaining a target inspection route, and performing inspection according to the target inspection route and the target inspection position. According to the embodiment of the invention, the flexibility of the unmanned aerial vehicle in the inspection process and the accuracy during shooting are improved.

Description

Technical Field

[0001] This application relates to the field of unmanned aerial vehicle (UAV) inspection technology, and in particular to an autonomous inspection method, device, and UAV based on a spatial perception field. Background Technology

[0002] With the development of science and technology, the field of drones has also seen tremendous technological advancements. Drones are now used in various industries, replacing humans in completing a variety of complex and dangerous tasks. Among these applications, drones are most widely used in power grid inspection.

[0003] Currently, when drones are used to inspect power grids, pilots need to control the drones to fly in semi-automatic mode and then switch to automatic mode. In automatic mode, the drones fly along pre-planned routes, resulting in poor flexibility during inspections. Summary of the Invention

[0004] Therefore, it is necessary to address the aforementioned technical problems by providing a method, device, and drone for autonomous drone inspection based on a spatial perception field, which can improve the flexibility of drone inspection during drone inspections.

[0005] Firstly, this application provides an autonomous inspection method for unmanned aerial vehicles (UAVs) based on a spatially perceived field, including:

[0006] During the flight of the UAV along the initial inspection route, for each object to be inspected, the first spatial radar data and the first spatial image data collected by the UAV at the initial inspection position are acquired.

[0007] Based on the first spatial radar data and the first spatial image data, determine the target spatial point cloud data corresponding to the object to be inspected;

[0008] Based on the current inspection location of the UAV and the second spatial radar data acquired by the UAV at the current inspection location, the second spatial image data is determined;

[0009] The spatial perception field of the UAV is generated based on the target spatial point cloud data and the second spatial image data;

[0010] The initial inspection route is adjusted according to the spatial sensing field to obtain the target inspection route, and inspection is carried out according to the target inspection route and the target inspection position; the target inspection position is determined based on the initial inspection position.

[0011] In one embodiment, determining the target spatial point cloud data corresponding to the object to be inspected at the initial inspection location based on the first spatial radar data and the first spatial image data includes:

[0012] Based on the first spatial radar data, determine the first distance between each collected object and the UAV at the initial inspection location;

[0013] The object to be inspected is determined to be the first distance, which is other than the first distance that is greater than the preset sensing distance.

[0014] First spatial point cloud data is obtained based on the first spatial radar data corresponding to the object to be inspected. Second spatial point cloud data is obtained by preprocessing the first spatial point cloud data.

[0015] The correlation evaluation is performed between the second spatial point cloud data and the object to be inspected to obtain a correlation evaluation value;

[0016] The second spatial point cloud data corresponding to the association evaluation value that is not less than the preset association evaluation value is determined as the target spatial point cloud data.

[0017] In one embodiment, determining the second spatial image data that meets the quality requirements based on the current inspection location of the UAV and the second spatial radar data acquired by the UAV at the current inspection location includes:

[0018] Based on the current inspection location and the second spatial radar data, the second distance between the UAV and the object to be inspected is obtained;

[0019] The second spatial image data is determined based on the preset distance and the second distance.

[0020] In one embodiment, generating the UAV's spatial perception field based on the target spatial point cloud data and the second spatial image data includes:

[0021] Obtain the color distribution information of the object to be inspected in the second spatial image data;

[0022] The target space point cloud data is voxelized to obtain a three-dimensional voxel model of the object to be inspected.

[0023] The three-dimensional voxel model is rendered based on the color distribution information to generate the spatial perception field of the UAV.

[0024] In one embodiment, adjusting the initial inspection route based on the spatial sensing field to obtain the target inspection route includes:

[0025] The spatial perception field is divided according to the initial inspection route to obtain the target perception area;

[0026] The target perception area is identified to obtain obstacle data and meteorological data;

[0027] The initial inspection route is adjusted based on the obstacle data and meteorological data to obtain the target inspection route.

[0028] In one embodiment, adjusting the initial inspection route based on the obstacle data and meteorological data to obtain the target inspection route includes:

[0029] The initial inspection route is adjusted based on the obstacle data and meteorological data to obtain the first flight route;

[0030] Acquire ultra-distance image data located outside a preset sensing distance from the second spatial image data;

[0031] A second flight path is generated based on the aforementioned long-range image data;

[0032] The target inspection route is obtained based on the first flight route and the second flight route.

[0033] In one embodiment, the method further includes:

[0034] The spatial position and shape information of the object to be inspected are obtained, and the relative direction between the UAV and the object to be inspected is determined based on the initial inspection position and the spatial position.

[0035] Based on the initial inspection position and the relative direction, obtain the third spatial image data of the object to be inspected;

[0036] The initial inspection position is evaluated based on the third spatial image data and the shape information to obtain an evaluation value for the initial inspection position;

[0037] The target inspection location is determined based on the evaluation value and the preset evaluation value.

[0038] In one embodiment, determining the target inspection location based on the evaluation value and a preset evaluation value includes:

[0039] If the evaluation value is less than the preset evaluation value, a three-dimensional inspection view of the object to be inspected is determined based on the spatial location and shape information of the object to be inspected.

[0040] Based on the ideal spatial image data of the object to be inspected and the three-dimensional inspection view, the hovering direction of the UAV is determined;

[0041] The target inspection position of the UAV is determined based on the collected parameters and the hovering direction of the UAV.

[0042] Secondly, this application also provides an autonomous inspection device for unmanned aerial vehicles based on a spatial perception field, comprising:

[0043] The first acquisition module is used to acquire first spatial radar data and first spatial image data collected by the UAV at the initial inspection position for each object to be inspected during the flight of the UAV along the initial inspection route.

[0044] The first determining module is used to determine the target spatial point cloud data corresponding to the object to be inspected based on the first spatial radar data and the first spatial image data.

[0045] The second determining module is used to determine the second spatial image data based on the current inspection position of the UAV and the second spatial radar data obtained by the UAV at the current inspection position.

[0046] The generation module is used to generate the spatial perception field of the UAV based on the target spatial point cloud data and the second spatial image data;

[0047] The adjustment module is used to adjust the initial inspection route according to the spatial perception field to obtain the target inspection route, and to perform inspection according to the target inspection route and the target inspection position; the target inspection position is determined based on the initial inspection position.

[0048] Thirdly, this application also provides an unmanned aerial vehicle (UAV) including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the method steps provided in the first aspect.

[0049] Fourthly, this application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the method steps provided in the first aspect.

[0050] Fifthly, this application also provides a computer program product, including a computer program that, when executed by a processor, implements the method steps provided in the first aspect.

[0051] The aforementioned UAV autonomous inspection method, device, and UAV based on a spatial perception field, during the flight of the UAV along the initial inspection route, acquire first spatial radar data and first spatial image data collected by the UAV at the initial inspection position for each object to be inspected. Based on the first spatial radar data and first spatial image data, the target spatial point cloud data corresponding to the object to be inspected is determined. Based on the current inspection position of the UAV and the second spatial radar data acquired by the UAV at the current inspection position, second spatial image data is determined. Based on the target spatial point cloud data and the second spatial image data, the UAV's spatial perception field is generated. The initial inspection route is adjusted according to the spatial perception field to obtain the target inspection route, and inspection is carried out based on the target inspection route and the target inspection position; the target inspection position is determined based on the initial inspection position. In the process of inspection based on the initial inspection route, the embodiments of this application improve the first spatial image data based on the first spatial radar data and the first spatial image data to obtain second spatial image data with higher image quality. The initial inspection route is then adjusted based on the perception field generated by the second spatial image data, which improves the flexibility of the UAV in the inspection process and the accuracy of shooting, while reducing the consumption of human resources and the consumption of UAV computing resources. Attached Figure Description

[0052] To more clearly illustrate the technical solutions in the embodiments of this application or related technologies, the drawings used in the description of the embodiments of this application or related technologies will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0053] Figure 1 This is an application environment diagram of an unmanned aerial vehicle (UAV) autonomous inspection method based on a spatial perception field in one embodiment;

[0054] Figure 2 This is a flowchart illustrating an autonomous inspection method for unmanned aerial vehicles (UAVs) based on a spatially perceived field, as shown in one embodiment.

[0055] Figure 3 This is a flowchart illustrating a method for determining target spatial point cloud data in one embodiment;

[0056] Figure 4 This is a flowchart illustrating a method for determining second spatial image data in one embodiment;

[0057] Figure 5 This is a flowchart illustrating a spatial sensing field determination method in one embodiment;

[0058] Figure 6 This is a flowchart illustrating a method for determining a target inspection route in one embodiment;

[0059] Figure 7 This is a flowchart illustrating the target inspection route determination method in another embodiment;

[0060] Figure 8 This is a flowchart illustrating a method for determining the target inspection location in one embodiment;

[0061] Figure 9 This is a flowchart illustrating the target inspection location determination method in another embodiment;

[0062] Figure 10 This is a structural block diagram of an unmanned aerial vehicle (UAV) autonomous inspection device based on a spatial perception field in one embodiment. Detailed Implementation

[0063] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0064] It should be noted that the terms "first," "second," etc., used in this application can be used to describe various elements, but these elements are not limited by these terms. These terms are only used to distinguish the first element from the second element. The terms "comprising" and "having," and any variations thereof, used in this application, are intended to cover non-exclusive inclusion. The term "multiple" used in this application refers to two or more. The term "and / or" used in this application refers to one of the embodiments, or any combination of multiple embodiments.

[0065] The UAV autonomous inspection method based on spatial perception field provided in this application embodiment can be applied to, for example... Figure 1 The application environment is shown below. This application environment includes a drone, which can be a server, and its internal structure diagram can be as follows. Figure 1As shown, the UAV includes a processor, memory, input / output interfaces (I / O), and a communication interface. The processor, memory, and I / O interfaces are connected via a system bus, and the communication interface is also connected to the system bus via the I / O interfaces. The UAV's processor provides computing and control capabilities. The UAV's memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and a database. The internal memory provides the environment for the operating system and computer programs in the non-volatile storage media to run. The UAV's database stores relevant data for autonomous inspection. The UAV's I / O interfaces are used for information exchange between the processor and external devices. The UAV's communication interface is used for communication with external terminals via a network connection. When the computer program is executed by the processor, it implements an autonomous inspection method for UAVs based on a spatial perception field.

[0066] Those skilled in the art will understand that Figure 1 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the drone to which the present application is applied. A specific drone may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0067] In one exemplary embodiment, such as Figure 2 As shown, an autonomous inspection method for unmanned aerial vehicles (UAVs) based on a spatial perception field is provided, and this method is applied to... Figure 1 Taking drones as an example, the following steps are taken: S201 to S205. Among them:

[0068] S201: During the flight of the UAV along the initial inspection route, for each object to be inspected, the UAV acquires the first spatial radar data and the first spatial image data collected at the initial inspection position.

[0069] The initial inspection position is the initial position corresponding to the object to be inspected. For example, if the initial inspection route includes 10 transmission towers, and each tower corresponds to an object to be inspected, then each object to be inspected corresponds to an initial inspection position. If the object to be inspected is the first object to be inspected on the initial inspection route, then the initial inspection position can also be the starting position of the initial inspection route.

[0070] In this embodiment, power grid construction data is acquired, including transmission tower height data, transmission tower location data, and power grid distribution data. An initial two-dimensional route is generated based on the transmission tower location data and the power grid distribution data. This initial two-dimensional route is then processed into a three-dimensional model based on the transmission tower height data to obtain the initial inspection route for the UAV.

[0071] Taking a power grid in a certain region that has been in use for two years as an example, the power grid inspection requirements and power grid construction data for that region were obtained. The power grid construction data includes specific information on 15 transmission towers in the area, with each tower ranging in height from 50 to 80 meters. The tower locations were recorded using Global Positioning System (GPS) coordinates, and the average distance between adjacent towers was 100 meters (i.e., power grid distribution data). An initial two-dimensional route was generated based on the tower location data and the power grid distribution data. This initial two-dimensional route is a polyline connecting all the coordinate points of the transmission towers. Then, based on the tower height data, the two-dimensional route was processed into a three-dimensional model to generate an initial inspection route at a height of 60 meters above the ground.

[0072] In one possible implementation, the drone is initially located at the initial inspection position, and the drone uses radar acquisition devices and image acquisition devices to acquire first spatial radar data and first spatial image data.

[0073] In another possible implementation, if the initial inspection position is the starting point of the initial inspection route, the UAV is not located at the initial inspection position. Therefore, the route of the initial inspection route is extracted, and the starting point is captured in reverse according to the route to obtain the initial inspection position; the current docking position of the UAV is obtained, and based on the docking position and the initial inspection position, the UAV's positioning flight path is generated, and the UAV is controlled to move to the initial inspection position according to the positioning flight path. When the UAV is located at the initial inspection position, the UAV performs radar detection and image capture in the surrounding space to obtain first spatial radar data and first spatial image data.

[0074] Taking a power grid in a certain region that has been in use for two years as an example, the initial inspection route is extracted as a west-to-east flight path. The initial inspection location is then captured in reverse, at the coordinates of the westernmost transmission tower No. 1 (118.5°E, 32.8°N). The drone is currently docked at a maintenance warehouse (118.3°E, 32.7°N). The generated positioning flight path requires the drone to first climb to an altitude of 80 meters, avoiding communication towers along the way. After flying along this positioning route for 12 minutes, the drone reaches the initial inspection location, activates the millimeter-wave radar for a 360-degree scan, and obtains first-space radar data for 57 objects within a 300-meter radius. Simultaneously, a high-definition camera is activated to capture first-space image data including 3 transmission towers, 2 residential buildings, and 1 river. Optionally, the millimeter-wave radar scanning frequency is set to 20Hz, continuously scanning for 3 minutes to generate a dynamic spatial radar dataset. The image acquisition interval is 2 seconds, accumulating 90 panoramic photos to form a spatial image dataset.

[0075] S202, Based on the first space radar data and the first space image data, determine the target space point cloud data corresponding to the object to be inspected.

[0076] Optionally, the object to be inspected can be one or multiple. For example, power grid inspection requirements clearly state that insulators, connecting hardware, and conductor joints on transmission towers need to be photographed. The resulting inspection requirement table includes three photographic targets: the surface condition of the insulators, the corrosion status of the hardware, and abnormal joint temperatures. This table is then converted into photographic command parameters that the drone can recognize, including a 45-degree downward shooting angle, a 4K shooting resolution, and three photos of each target taken from different angles, forming the initial photographic requirements.

[0077] In this embodiment, the first distance between each collected object and the UAV at the initial inspection position can be determined based on the first spatial radar data. The collected object corresponding to the first distance other than the first distance greater than the preset perception distance is identified as the object to be inspected. The first spatial point cloud data is obtained based on the first spatial radar data corresponding to the object to be inspected. The first spatial point cloud data is preprocessed to obtain the second spatial point cloud data. The correlation between the second spatial point cloud data and the object to be inspected is evaluated to obtain the correlation evaluation value. The second spatial point cloud data corresponding to the correlation evaluation value not less than the preset correlation evaluation value is identified as the target spatial point cloud data.

[0078] In one possible implementation, the second spatial point cloud data obtained above can also be used as the target spatial point cloud data.

[0079] S203, determine the second space image data based on the current inspection position of the UAV and the second space radar data acquired by the UAV at the current inspection position.

[0080] In this embodiment of the application, the second distance between the UAV and the object to be inspected is obtained based on the current inspection position and the second spatial radar data, and the second spatial image data is determined based on the preset distance and the second distance.

[0081] In one possible implementation, the second spatial radar data and the first spatial radar data can be directly compared. If the difference between the second spatial radar data and the first spatial radar data meets the preset difference, the first spatial image data is used as the second spatial image data. If the difference between the second spatial radar data and the first spatial radar data does not meet the preset difference, the acquisition parameters of the UAV's acquisition device are adjusted, and the second spatial image data is obtained based on the adjusted acquisition parameters.

[0082] S204 generates the UAV's spatial perception field based on target spatial point cloud data and second spatial image data.

[0083] In this embodiment of the application, the color distribution information of the object to be inspected in the second spatial image data is obtained, the target spatial point cloud data is voxelized to obtain a three-dimensional voxel model of the object to be inspected, and the three-dimensional voxel model is rendered according to the color distribution information to generate the spatial perception field of the UAV.

[0084] In one possible implementation, color distribution information can be fused with the target spatial point cloud. This involves projecting the point cloud onto second-space image data and assigning color information to each point; alternatively, the image pixels of the object to be inspected in the second-space image data can be correlated with the point cloud to add texture and color. An environmental model is then constructed using the fused data, and a spatial perception field is generated based on this model.

[0085] S205, adjust the initial inspection route according to the spatial perception field to obtain the target inspection route, and perform inspection according to the target inspection route and the target inspection position; the target inspection position is determined based on the initial inspection position.

[0086] In this embodiment of the application, the spatial sensing field can be divided according to the initial inspection route to obtain the target sensing area, the target sensing area can be identified to obtain obstacle data and meteorological data, and the initial inspection route can be adjusted according to the obstacle data and meteorological data to obtain the target inspection route.

[0087] In one possible implementation, meteorological data can be received in real time, and the initial inspection route can be adjusted using the meteorological data to obtain the target inspection route.

[0088] The target inspection location is determined based on the initial inspection location. This can be either using the initial inspection location as the target inspection location, or evaluating the initial inspection location and adjusting it if it does not meet the requirements to obtain the target inspection location.

[0089] In the aforementioned UAV autonomous inspection method based on a spatial perception field, during the UAV's flight along the initial inspection route, for each object to be inspected, first spatial radar data and first spatial image data collected by the UAV at the initial inspection position are acquired. Based on the first spatial radar data and first spatial image data, target spatial point cloud data corresponding to the object to be inspected is determined. Based on the UAV's current inspection position and second spatial radar data acquired by the UAV at the current inspection position, second spatial image data is determined. Based on the target spatial point cloud data and second spatial image data, a spatial perception field for the UAV is generated. The initial inspection route is adjusted based on the spatial perception field to obtain the target inspection route, and inspection is performed based on the target inspection route and target inspection position; the target inspection position is determined based on the initial inspection position. In the embodiment of this application, during the inspection process based on the initial inspection route, the first spatial image data is improved based on the first spatial radar data and first spatial image data to obtain second spatial image data with higher image quality. Thus, the initial inspection route is adjusted based on the spatial perception field generated from the second spatial image data, improving the flexibility of the UAV during the inspection process and the accuracy during shooting, while reducing the consumption of human resources and UAV computing resources.

[0090] Figure 3 This is a flowchart illustrating a method for determining target spatial point cloud data in one embodiment, such as... Figure 3 As shown, this application embodiment relates to a possible implementation of how to determine the target spatial point cloud data corresponding to the object to be inspected at the initial inspection position based on first spatial radar data and first spatial image data, including the following steps:

[0091] S301 determines the first distance between each data collection object and the UAV at the initial inspection location based on the first spatial radar data.

[0092] In this embodiment of the application, the first spatial radar data is filtered and denoised, and a network model is used to segment different acquisition objects (e.g., power towers, wires, insulator strings, etc.) and calculate the first distance between each acquisition object and the UAV.

[0093] S302, determine the object to be inspected as the first distance, excluding the first distance that is greater than the preset sensing distance.

[0094] In this embodiment of the application, taking a power grid in a certain area that has been in use for two years as an example, when processing the first space radar data, the collection objects with a distance of more than 200 meters are first filtered out, and 23 valid targets to be inspected are retained.

[0095] S303: Based on the first spatial radar data corresponding to the object to be inspected, first spatial point cloud data is obtained; the first spatial point cloud data is preprocessed to obtain second spatial point cloud data.

[0096] In this embodiment, target shape data of the object to be inspected is obtained based on the first spatial radar data corresponding to the object to be inspected, and first spatial point cloud data is generated based on the target shape data. The first spatial point cloud data is then processed by denoising, clustering, etc., to obtain second spatial point cloud data.

[0097] S304. The correlation evaluation between the second spatial point cloud data and the object to be inspected is performed to obtain the correlation evaluation value.

[0098] In this embodiment, a correlation evaluation value is obtained by evaluating the correlation between the second spatial point cloud data and the object to be inspected. For example, the correlation evaluation value can be obtained by comparing the geometric characteristics of the second spatial point cloud data and the object to be inspected, such as volume, surface area, density, and eccentricity. Alternatively, the reference point cloud of the object to be inspected can be registered with the second spatial point cloud data, and then the registration error can be calculated. The smaller the error, the higher the correlation evaluation value.

[0099] S305, determine the second spatial point cloud data corresponding to the association evaluation value that is not less than the preset association evaluation value as the target spatial point cloud data.

[0100] In this embodiment of the application, taking a power grid that has been in use for two years in a certain region as an example, when processing the first spatial point cloud data of the transmission tower, a clustering algorithm is first used to remove discrete noise points, reducing the original 350,000 point clouds to 280,000 valid points (i.e., the second spatial point cloud data). By calculating the consistency between the second spatial point cloud data and the object to be inspected, abnormal point clouds with an association evaluation value lower than 0.7 are deleted, and finally, the target spatial point cloud data consisting of 240,000 points is retained.

[0101] In this embodiment, the first distance between each collected object and the UAV at the initial inspection position is determined based on first spatial radar data; the collected objects corresponding to the first distance other than the first distance greater than the preset sensing distance are identified as objects to be inspected; first spatial point cloud data is obtained based on the first spatial radar data corresponding to the objects to be inspected; the first spatial point cloud data is preprocessed to obtain second spatial point cloud data; the correlation between the second spatial point cloud data and the objects to be inspected is evaluated to obtain a correlation evaluation value; the second spatial point cloud data corresponding to a correlation evaluation value not less than the preset correlation evaluation value is identified as target spatial point cloud data. This embodiment reduces the amount of data and improves the processing efficiency of point cloud data by removing data from the first spatial point cloud data through preprocessing and correlation evaluation, thereby improving the inspection efficiency of the UAV.

[0102] Figure 4 This is a flowchart illustrating a method for determining second spatial image data in one embodiment, as shown below. Figure 4As shown, this application embodiment relates to a possible implementation of how to determine second spatial image data that meets quality requirements based on the current inspection position of the UAV and the second spatial radar data acquired by the UAV at the current inspection position, including the following steps:

[0103] S401, based on the current inspection position and second space radar data, obtains the second distance between the UAV and the object to be inspected.

[0104] In this embodiment of the application, the centroid of the second spatial radar data is determined, and the distance from the current inspection position of the UAV to the centroid is used as the second distance between the UAV and the object to be inspected.

[0105] In one possible implementation, the distance between the current inspection location and each spatial radar data in the second spatial radar data is determined, and the minimum distance is taken as the second distance between the UAV and the object to be inspected.

[0106] S402, determine the second spatial image data based on the preset distance and the second distance.

[0107] In this embodiment of the application, when the second distance is less than the preset distance, the acquisition parameters of the acquisition device of the UAV are adjusted, that is, when the UAV acquires the second spatial image data, the acquisition image pixel is automatically reduced; when the second distance is not less than the preset distance, the UAV acquires the second spatial image data, and the pixel is automatically increased, so that the pixel quality of the second spatial image data is greater than the pixel quality of the first spatial image data.

[0108] In another possible implementation, the first spatial image data can be post-processed according to a preset distance and a second distance to obtain the second spatial image data.

[0109] For example, using the target spatial point cloud data consisting of the aforementioned 240,000 points, a 3D reconstruction is performed using a 5cm voxel size to generate a power transmission tower model with a clear angle steel connection structure. When the drone is 150 meters away from the power transmission tower (below the preset 200-meter threshold), the camera resolution is automatically reduced from 4K to 1080P to save storage space. If the power transmission tower is more than 500 meters away, an 8K oversampling mode is enabled and optical zoom is increased.

[0110] In this embodiment, a second distance between the UAV and the object to be inspected is obtained based on the current inspection location and the second spatial radar data; second spatial image data is determined based on the preset distance and the second distance, so that the obtained second spatial image data better meets the inspection requirements and improves the accuracy of the second spatial image data.

[0111] Figure 5 This is a flowchart illustrating a spatial sensing field determination method in one embodiment, as shown below. Figure 5 As shown, this application embodiment relates to a possible implementation of how to generate a UAV's spatial perception field based on target spatial point cloud data and second spatial image data, including the following steps:

[0112] S501, Obtain the color distribution information of the object to be inspected in the second spatial image data.

[0113] In this embodiment, the second spatial image data is converted from RGB to other color spaces (such as HSV, LAB, etc.) to better analyze color features. Based on other color spaces, color histograms are statistically analyzed or color matrices are calculated to obtain color distribution information. For example, the angular features of the steel frame structure of the tower are extracted, and the gray paint features of the corresponding transmission tower are matched in the second spatial image data. For the red tile roof of the adjacent house, the main hue of RGB (192, 64, 48) is extracted through color histogram analysis.

[0114] S502, the target space point cloud data is voxelized to obtain a three-dimensional voxel model of the object to be inspected.

[0115] In this embodiment, a voxelization method (such as voxel mesh) is used to convert the target spatial point cloud data into a three-dimensional voxel model. For example, the target spatial point cloud of the transmission tower is fused with a real-world image to construct a three-dimensional voxel model containing accurate dimensions (tower height 68 meters, base width 8 meters) and true colors.

[0116] S503 renders the 3D voxel model based on color distribution information to generate the UAV's spatial perception field.

[0117] In this embodiment, the obtained color distribution information is mapped onto a three-dimensional voxel model. That is, the three-dimensional voxel model is colored according to the color distribution information to generate the spatial perception field of the UAV. For example, a semi-transparent blue is used to render the river area, and a spatial perception field containing terrain, buildings, vegetation distribution, and power facilities is finally generated. The damaged part of the insulator string is marked (located at the 6th crossarm of the tower), and the unique dark brown corrosion feature of this area is extracted. The RGB (88, 52, 32) color values ​​are mapped to the corresponding positions in the three-dimensional voxel model.

[0118] In this embodiment, the color distribution information of the object to be inspected in the second spatial image data is obtained, the target spatial point cloud data is voxelized to obtain a three-dimensional voxel model of the object to be inspected, the three-dimensional voxel model is rendered according to the color distribution information, and the spatial perception field of the UAV is generated, which lays the foundation for subsequent adjustment of the initial inspection route based on the UAV's perception field to obtain the target inspection route.

[0119] Figure 6This is a flowchart illustrating a method for determining a target inspection route in one embodiment, as shown below. Figure 6 As shown, this application embodiment relates to a possible implementation of how to adjust the initial inspection route based on the spatial perception field to obtain the target inspection route, including the following steps:

[0120] S601 divides the spatial perception field according to the initial inspection route to obtain the target perception area.

[0121] In this embodiment of the application, the inspection direction of the UAV is obtained according to the initial flight path, and the spatial perception field of the UAV is divided according to the inspection direction to obtain the target perception area.

[0122] S602 identifies the target perception area and obtains obstacle data and meteorological data.

[0123] In this embodiment of the application, obstacle data includes dynamic obstacles and static obstacles. Based on all grids marked as "occupied" within the target perception area, these typically correspond to static obstacles such as buildings, trees, and utility poles.

[0124] In one possible implementation, the target perception area includes marked areas where wind speeds exceed a safety threshold, areas with low visibility, turbulent areas, etc.; wind direction information is used to calculate its impact on the drone's heading, and areas with low visibility may require the drone to change altitude or detour.

[0125] S603 adjusts the initial inspection route based on obstacle data and meteorological data to obtain the target inspection route.

[0126] In this embodiment of the application, the initial inspection route can be adjusted based on obstacle data and meteorological data to obtain a first flight route, and the first flight heading can be used as the target inspection route.

[0127] In one possible implementation, it is also possible to acquire ultra-distance image data located outside the preset sensing distance in the second spatial image data, generate a second flight path based on the ultra-distance image data, and obtain the target inspection path based on the first and second flight paths.

[0128] In this embodiment, the spatial perception field is divided according to the initial inspection route to obtain the target perception area. The target perception area is identified to obtain obstacle data and meteorological data. The initial inspection route is adjusted according to the obstacle data and meteorological data to obtain the target inspection route. In this embodiment, the accuracy of UAV inspection based on the target inspection route is improved by adjusting the initial inspection route with obstacle data and meteorological data.

[0129] Figure 7This is a flowchart illustrating the target inspection route determination method in another embodiment, as shown below. Figure 7 As shown, this application embodiment relates to a possible implementation of how to adjust the initial inspection route based on obstacle data and meteorological data to obtain the target inspection route, including the following steps:

[0130] S701 adjusts the initial inspection route based on obstacle data and meteorological data to obtain the first flight route.

[0131] S702, acquire ultra-distance image data located outside the preset sensing distance in the second spatial image data.

[0132] S703 generates a second flight path based on over-the-horizon image data.

[0133] S704, based on the first and second flight paths, obtains the target inspection route.

[0134] In this embodiment, an avoidance route is generated based on obstacle data, and an optimized route is generated based on meteorological data. The initial inspection route is adjusted by combining the avoidance route and the optimized route, and a short-distance precise route (i.e., the first flight route) is generated in real time. Based on the second spatial image data, long-distance image data located beyond a preset sensing distance is extracted, and a long-distance approximate route (i.e., the second flight route) is generated based on the long-distance image data. The target inspection heading is then determined in real time by combining the short-distance precise route and the long-distance approximate route.

[0135] Taking a power grid in a certain region that has been in use for two years as an example, when the UAV flies eastward along the initial inspection route, the spatial perception field divides the target perception area into 300m × 300m. A crane (45m high) in operation is detected 2km ahead of the initial inspection route, generating an avoidance route at an altitude of 70m. Simultaneously, a crosswind speed of 8m / s is detected, generating an optimized route that flies leeward. A first flight path with centimeter-level accuracy is generated within a 300m range. For transmission towers 5km away (preset perception distance), a long-distance approximate flight path maintaining a straight line is generated based on the ultra-long-range image data from the second spatial image data. By fusing the short-distance precise flight path with 50m accuracy and the long-distance approximate flight path with 500m accuracy, a composite inspection route that balances safety and efficiency is formed, i.e., the target inspection route, with an adjustment frequency of 3 times per second throughout the entire process.

[0136] In this embodiment, the initial inspection route is adjusted based on obstacle data and meteorological data to obtain a first flight route. Overrange image data located outside the preset perception distance is acquired from the second spatial image data. A second flight route is generated based on the overrange image data. The target inspection route is obtained based on the first and second flight routes. This allows the UAV to adjust the initial inspection route in real time based on the actual collected data during the inspection process to obtain the target inspection route. Thus, the inspection is carried out based on the target inspection route, which improves the flexibility and accuracy of UAV inspection.

[0137] Figure 8 This is a flowchart illustrating a method for determining the target inspection location in one embodiment, as shown below. Figure 8 As shown, it includes the following steps:

[0138] S801 acquires the spatial position and shape information of the object to be inspected, and determines the relative direction between the UAV and the object to be inspected based on the initial inspection position and spatial position.

[0139] In this embodiment, the power grid inspection requirements clearly specify the need to photograph the insulators, connecting hardware, and conductor joints of the transmission towers. This is summarized into an inspection requirement table containing three photographic targets for the objects to be inspected: the surface condition of the insulators, the corrosion status of the hardware, and abnormal joint temperatures. These photographic targets are then converted into photographic command parameters recognizable by the UAV, including a 45-degree downward shooting angle, a 4K shooting resolution, and the requirement to take three photos of each target from different angles, forming the initial photographic requirements. In other words, the initial photographic requirements are determined based on the power grid inspection requirements, and the objects to be inspected and the initial inspection locations are determined based on these initial photographic requirements.

[0140] Based on the object to be inspected, a drone is invoked to scan the object, obtaining its spatial position and shape information. Combining the initial inspection position and spatial position, the relative direction between the drone and the object is determined.

[0141] S802, based on the initial inspection position and relative direction, acquires the third-space image data of the object to be inspected.

[0142] S803 evaluates the initial inspection position based on the third space image data and shape information to obtain the evaluation value of the initial inspection position.

[0143] Optionally, the evaluation values ​​for the initial inspection position may include at least one of the following: visibility, occlusion level, image quality, distance, and angle. Visibility represents the proportion of the object to be inspected that can be seen from the initial inspection position; occlusion level represents the degree to which the object to be inspected is obscured by other objects; image quality represents the image clarity and lighting conditions of the captured third-space image data; distance represents whether the distance between the drone and the object to be inspected is suitable for inspection (too far and the details of the object will be unclear, too close and the full view of the object may not be captured); angle represents the angle between the shooting direction and the normal direction of the object's surface (frontal shooting is preferred).

[0144] In this embodiment of the application, visibility is evaluated by projecting a three-dimensional voxel model of the object to be inspected onto third-space image data, and then comparing the projected area with the area actually detected in the third-space image data.

[0145] Optionally, the evaluation value of the initial inspection position can be obtained by weighting the visibility, occlusion degree, image quality, distance, and angle.

[0146] S804 determines the target inspection location based on the evaluation value and the preset evaluation value.

[0147] In this embodiment of the application, if the evaluation value is greater than the preset evaluation value, the initial inspection position is taken as the target inspection position; if the evaluation value is greater than the preset evaluation value, the initial inspection position is adjusted to obtain the target inspection position.

[0148] Taking a power grid in a certain region that has been in use for two years as an example, the initial shooting requirements stipulated that the integrity of the insulator skirts must be captured. When the drone arrived at the initial inspection position of transmission tower No. 3 (30 meters southeast of the tower), laser scanning showed that the insulator string was located at a height of 42 meters on the tower. The current shooting angle was 60 degrees of elevation, and it was expected that the tower would obstruct the image. Third-space image data showed that 40% of the skirts were obscured by the crossarm structure. Based on the third-space image data, the insulator string was identified, and the overall evaluation of the initial inspection position, including the recognition rate (62%) and the visibility of key features (55%), was 68 points (lower than the preset evaluation value of 75 points). After adjustment, it was found that an unobstructed frontal view could be obtained at a position 15 meters northwest of the tower at an altitude of 50 meters, so this position was selected as the target inspection position.

[0149] In this embodiment, the spatial location and shape information of the object to be inspected are acquired, and the relative direction between the drone and the object is determined based on the initial inspection location and spatial location. Third-dimensional image data of the object is acquired based on the initial inspection location and relative direction. The initial inspection location is evaluated based on the third-dimensional image data and shape information to obtain an evaluation value. The target inspection location is determined based on the evaluation value and a preset evaluation value. This embodiment allows for adjustment of the preset initial inspection location during the inspection process, enabling the drone to capture inspection images at the target inspection location, thus improving the accuracy and safety of drone inspections.

[0150] Figure 9 This is a flowchart illustrating the target inspection location determination method in another embodiment, as shown below. Figure 9 As shown, this application embodiment relates to a possible implementation of how to determine the target inspection location based on the evaluation value and the preset evaluation value, including the following steps:

[0151] S901, if the evaluation value is less than the preset evaluation value, determine the three-dimensional inspection view of the object to be inspected based on the spatial location and shape information of the object to be inspected.

[0152] S902 determines the hovering direction of the UAV based on the ideal spatial image data and three-dimensional inspection view of the object to be inspected.

[0153] In this embodiment of the application, the pose of the UAV when shooting the ideal space image data is extracted from the ideal space image data. The optimal shooting angle of the UAV is obtained by using the three-dimensional inspection view and the UAV pose. The hovering direction of the UAV is determined based on the optimal shooting angle.

[0154] The S903 determines the target inspection location of the UAV based on the collected parameters and hovering direction.

[0155] In this embodiment of the application, the acquisition parameters of the UAV are obtained, the optimal shooting distance of the UAV is obtained based on the acquisition parameters, the target inspection position is obtained based on the optimal shooting distance and the hovering direction, and the inspection object is photographed at the target inspection position to obtain the inspection image.

[0156] Taking a power grid in a certain region that has been in use for two years as an example, the initial inspection location was replanned. Based on the 84-degree viewing angle parameter, the optimal shooting distance for a 2-meter-long insulator string was determined to be 8-10 meters. Combining the 3D inspection view, it was determined that the drone should hover 15 meters northwest of the tower, at an altitude of 48 meters (118.72 degrees east longitude, 32.81 degrees north latitude). This location can completely cover the 12-piece umbrella skirt structure, and the lighting conditions meet the photosensitive requirements. The adjusted shooting position improved the imaging resolution from the original 31 million pixels to 45 million pixels, and the recognition rate of key features increased to 92%. The final inspection image clearly showed a 3cm crack defect in the 5th umbrella skirt.

[0157] In this embodiment, when the evaluation value is less than the preset evaluation value, a three-dimensional inspection view of the object to be inspected is determined based on the spatial location and shape information of the object to be inspected. The hovering direction of the UAV is determined based on the ideal spatial image data and the three-dimensional inspection view of the object to be inspected. The target inspection position of the UAV is determined based on the acquisition parameters and the hovering direction of the UAV. In this embodiment, the hovering direction of the UAV is determined by using ideal spatial image data and the three-dimensional inspection view, and the target inspection position of the UAV is determined based on the acquisition parameters and the hovering direction of the UAV, which improves the accuracy of the UAV inspection position determination.

[0158] In this embodiment of the application, when the evaluation value is less than the preset evaluation value, the three-dimensional inspection view of the object to be inspected is determined based on the spatial position and shape information of the object to be inspected. The hovering direction of the UAV is determined based on the ideal spatial image data and the three-dimensional inspection view of the object to be inspected. The target inspection position of the UAV is determined based on the acquisition parameters and hovering direction of the UAV.

[0159] It should be understood that although the steps in the flowcharts of the embodiments described above are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the embodiments described above may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages in other steps. It is understood that the steps in different embodiments can be freely combined as needed, and all non-contradictory solutions formed by such combinations are within the scope of protection of this application.

[0160] Based on the same inventive concept, this application also provides a UAV autonomous inspection device based on a spatially perceptual field for implementing the aforementioned UAV autonomous inspection method based on a spatially perceptual field. The solution provided by this device is similar to the solution described in the above method. Therefore, the specific limitations of one or more embodiments of the UAV autonomous inspection device based on a spatially perceptual field provided below can be found in the limitations of the UAV autonomous inspection method based on a spatially perceptual field described above, and will not be repeated here.

[0161] In one exemplary embodiment, such as Figure 10 As shown, an autonomous inspection device for unmanned aerial vehicles (UAVs) based on a spatial perception field is provided, comprising: a first acquisition module 11, a first determination module 12, a second determination module 13, a generation module 14, and an adjustment module 15, wherein:

[0162] The first acquisition module 11 is used to acquire, for each object to be inspected, the first spatial radar data and the first spatial image data collected by the UAV at the initial inspection position during the flight of the UAV along the initial inspection route.

[0163] The first determining module 12 is used to determine the target spatial point cloud data corresponding to the object to be inspected based on the first spatial radar data and the first spatial image data.

[0164] The second determining module 13 is used to determine the second spatial image data based on the current inspection position of the UAV and the second spatial radar data obtained by the UAV at the current inspection position.

[0165] The generation module 14 is used to generate the UAV's spatial perception field based on the target spatial point cloud data and the second spatial image data.

[0166] The adjustment module 15 is used to adjust the initial inspection route according to the spatial perception field to obtain the target inspection route, and to perform inspection according to the target inspection route and the target inspection position; the target inspection position is determined based on the initial inspection position.

[0167] In one embodiment, the first determining module 12 is specifically used to determine the first distance between each collection object at the initial inspection position and the UAV based on the first spatial radar data; determine the collection object corresponding to the first distance other than the first distance greater than the preset sensing distance as the object to be inspected; obtain the first spatial point cloud data based on the first spatial radar data corresponding to the object to be inspected; preprocess the first spatial point cloud data to obtain the second spatial point cloud data; perform a correlation evaluation on the second spatial point cloud data and the object to be inspected to obtain a correlation evaluation value; and determine the second spatial point cloud data corresponding to the correlation evaluation value not less than the preset correlation evaluation value as the target spatial point cloud data.

[0168] In one embodiment, the second determining module 13 is specifically used to obtain a second distance between the UAV and the object to be inspected based on the current inspection position and the second spatial radar data; and to determine the second spatial image data based on the preset distance and the second distance.

[0169] In one embodiment, the generation module 14 is specifically used to acquire the color distribution information of the object to be inspected in the second spatial image data; to voxelize the target spatial point cloud data to obtain a three-dimensional voxel model of the object to be inspected; and to render the three-dimensional voxel model according to the color distribution information to generate the spatial perception field of the UAV.

[0170] In one embodiment, the adjustment module 15 is specifically used to divide the spatial perception field according to the initial inspection route to obtain the target perception area; identify the target perception area to obtain obstacle data and meteorological data; and adjust the initial inspection route according to the obstacle data and meteorological data to obtain the target inspection route.

[0171] In one embodiment, the adjustment module 15 is specifically used to adjust the initial inspection route according to obstacle data and meteorological data to obtain a first flight route; acquire overrange image data located outside the preset sensing distance in the second spatial image data; generate a second flight route according to the overrange image data; and obtain a target inspection route according to the first flight route and the second flight route.

[0172] In one embodiment, the device further includes:

[0173] The third determining module is used to acquire the spatial position and shape information of the object to be inspected, and to determine the relative direction between the UAV and the object to be inspected based on the initial inspection position and spatial position.

[0174] The second acquisition module is used to acquire the third-space image data of the object to be inspected based on the initial inspection position and relative direction.

[0175] The evaluation module is used to evaluate the initial inspection position based on the third space image data and shape information, and obtain the evaluation value of the initial inspection position.

[0176] The fourth determination module is used to determine the target inspection location based on the evaluation value and the preset evaluation value.

[0177] In one embodiment, the fourth determining module is specifically used to determine the three-dimensional inspection view of the object to be inspected based on the spatial position and shape information of the object to be inspected when the evaluation value is less than the preset evaluation value; determine the hovering direction of the UAV based on the ideal spatial image data and the three-dimensional inspection view of the object to be inspected; and determine the target inspection position of the UAV based on the acquisition parameters and hovering direction of the UAV.

[0178] The modules in the aforementioned UAV autonomous inspection device based on spatial perception fields can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in the UAV's processor in hardware form or independent of it, or stored in the UAV's memory in software form, so that the processor can call and execute the corresponding operations of each module.

[0179] In one exemplary embodiment, a drone is provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of any of the above method embodiments.

[0180] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements the steps of any of the above method embodiments.

[0181] In one embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements the steps of any of the above method embodiments.

[0182] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of the relevant data must comply with relevant regulations.

[0183] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile memory and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, artificial intelligence (AI) processors, etc., and are not limited to these.

[0184] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this application.

[0185] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A method for autonomous inspection of unmanned aerial vehicles (UAVs) based on a spatial perception field, characterized in that, The method includes: During the flight of the UAV along the initial inspection route, for each object to be inspected, the first spatial radar data and the first spatial image data collected by the UAV at the initial inspection position are acquired. Based on the first spatial radar data and the first spatial image data, determine the target spatial point cloud data corresponding to the object to be inspected; Based on the current inspection location of the UAV and the second spatial radar data acquired by the UAV at the current inspection location, the second spatial image data is determined; The spatial perception field of the UAV is generated based on the target spatial point cloud data and the second spatial image data; The initial inspection route is adjusted according to the spatial sensing field to obtain the target inspection route, and inspection is carried out according to the target inspection route and the target inspection position; the target inspection position is determined based on the initial inspection position.

2. The method according to claim 1, characterized in that, The step of determining the target spatial point cloud data corresponding to the object to be inspected at the initial inspection position based on the first spatial radar data and the first spatial image data includes: Based on the first spatial radar data, determine the first distance between each collected object and the UAV at the initial inspection location; The object to be inspected is determined to be the first distance, which is other than the first distance that is greater than the preset sensing distance. First spatial point cloud data is obtained based on the first spatial radar data corresponding to the object to be inspected. Second spatial point cloud data is obtained by preprocessing the first spatial point cloud data. The correlation evaluation is performed between the second spatial point cloud data and the object to be inspected to obtain a correlation evaluation value; The second spatial point cloud data corresponding to the association evaluation value that is not less than the preset association evaluation value is determined as the target spatial point cloud data.

3. The method according to claim 1, characterized in that, The step of determining the second spatial image data that meets the quality requirements based on the current inspection position of the UAV and the second spatial radar data acquired by the UAV at the current inspection position includes: Based on the current inspection location and the second spatial radar data, the second distance between the UAV and the object to be inspected is obtained; The second spatial image data is determined based on the preset distance and the second distance.

4. The method according to claim 1, characterized in that, The step of generating the UAV's spatial perception field based on the target spatial point cloud data and the second spatial image data includes: Obtain the color distribution information of the object to be inspected in the second spatial image data; The target space point cloud data is voxelized to obtain a three-dimensional voxel model of the object to be inspected. The three-dimensional voxel model is rendered based on the color distribution information to generate the spatial perception field of the UAV.

5. The method according to claim 1, characterized in that, The step of adjusting the initial inspection route based on the spatial sensing field to obtain the target inspection route includes: The spatial perception field is divided according to the initial inspection route to obtain the target perception area; The target perception area is identified to obtain obstacle data and meteorological data; The initial inspection route is adjusted based on the obstacle data and meteorological data to obtain the target inspection route.

6. The method according to claim 5, characterized in that, The step of adjusting the initial inspection route based on the obstacle data and meteorological data to obtain the target inspection route includes: The initial inspection route is adjusted based on the obstacle data and meteorological data to obtain the first flight route; Acquire ultra-distance image data located outside a preset sensing distance from the second spatial image data; A second flight path is generated based on the aforementioned long-range image data; The target inspection route is obtained based on the first flight route and the second flight route.

7. The method according to claim 1, characterized in that, The method further includes: The spatial position and shape information of the object to be inspected are obtained, and the relative direction between the UAV and the object to be inspected is determined based on the initial inspection position and the spatial position. Based on the initial inspection position and the relative direction, obtain the third spatial image data of the object to be inspected; The initial inspection position is evaluated based on the third spatial image data and the shape information to obtain an evaluation value for the initial inspection position; The target inspection location is determined based on the evaluation value and the preset evaluation value.

8. The method according to claim 7, characterized in that, Determining the target inspection location based on the evaluation value and the preset evaluation value includes: If the evaluation value is less than the preset evaluation value, a three-dimensional inspection view of the object to be inspected is determined based on the spatial location and shape information of the object to be inspected. Based on the ideal spatial image data of the object to be inspected and the three-dimensional inspection view, the hovering direction of the UAV is determined; The target inspection position of the UAV is determined based on the collected parameters and the hovering direction of the UAV.

9. An autonomous inspection device for unmanned aerial vehicles (UAVs) based on a spatial perception field, characterized in that, The device includes: The first acquisition module is used to acquire first spatial radar data and first spatial image data collected by the UAV at the initial inspection position for each object to be inspected during the flight of the UAV along the initial inspection route. The first determining module is used to determine the target spatial point cloud data corresponding to the object to be inspected based on the first spatial radar data and the first spatial image data. The second determining module is used to determine the second spatial image data based on the current inspection position of the UAV and the second spatial radar data obtained by the UAV at the current inspection position. The generation module is used to generate the spatial perception field of the UAV based on the target spatial point cloud data and the second spatial image data; The adjustment module is used to adjust the initial inspection route according to the spatial perception field to obtain the target inspection route, and to perform inspection according to the target inspection route and the target inspection position; the target inspection position is determined based on the initial inspection position.

10. A drone, comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1-8.

Images