A method and system for correcting the route of a UAV on a power transmission line based on point cloud recognition

Abstract

The present invention proposes a method for automatic verification of transmission line UAV routes based on point cloud. By identifying and segmenting the point cloud data of the transmission line, the transmission line data is obtained, the preset shooting points and target points are safety verified, and the shooting points are regenerated according to the verification results, so as to improve the safety and effectiveness of the shooting points, realize the optimization of automatic planning of the UAV waypoints, and finally generate a new UAV automatic positioning route, so as to realize the automatic inspection of the UAV on the transmission line. The invention saves labor costs and provides a more accurate route by using the automatic verification method.

Description

Technical Field

[0001] The present invention relates to the technical field of transmission line UAV route correction, and includes a route automatic correction technical method, device, electronic device, and computer-readable storage medium for realizing automatic planning of safety detection and automatic verification of transmission line target points. Background Art

[0002] Transmission lines are composed of line towers, conductors, insulators, ground wires, etc., and are erected above the ground. They play an important role in the power transmission system. Regular inspections of transmission lines can help prevent power system failures and reduce the risk of fires and tripping. Early detection and repair of problems in transmission lines can reduce energy waste and environmental pollution. Drones are equipped with high-resolution cameras and sensors that can collect detailed images of transmission lines and other data. By regularly using drones for inspections, power companies can promptly detect potential failures, damage or deterioration of transmission lines, thereby reducing the risk of tripping and improving the reliability of the power system.

[0003] At present, the inspection of power transmission lines mainly relies on two methods: manual inspection by professional pilots and automatic inspection by preset routes. Manual inspection by professional pilots relies on the experience and knowledge of professional pilots. It requires long-term training for pilots to enable them to master relevant content such as flight regulations, equipment operation, emergency situation handling, and power professional knowledge to ensure the safety and effectiveness of drone inspections. However, this method has problems such as strong subjectivity, low degree of automation, and low work efficiency. With the rapid development of power grids, manual inspection technology is increasingly difficult to adapt to modern needs. Therefore, the automatic positioning inspection and data return function of drones has become one of the important development directions of transmission line inspection. Automatic positioning inspection by drones is an inspection mode that uses automatically generated inspection routes based on shooting specifications for inspection work. However, the automatically generated inspection routes currently have problems such as poor versatility, low safety, and inaccurate routes.

[0004] In order to solve the above problems, the present invention proposes a method for automatic verification of transmission line UAV routes based on point cloud. By identifying and segmenting the point cloud data of the transmission line, the transmission line data is obtained and summarized, the preset shooting points and target points are safety verified, and the shooting points are regenerated according to the verification results, so as to improve the safety and effectiveness of the shooting points, realize the optimization of automatic planning of the UAV waypoints, and finally generate a new UAV automatic positioning route to realize automatic inspection of the transmission line by UAV. The invention saves labor costs and provides safer, more accurate and more effective shooting images by using the automatic verification method. Summary of the invention

[0005] The purpose of the present invention is to provide a transmission line UAV route correction method, system, electronic device, and computer-readable storage medium based on point cloud recognition. The present invention uses point cloud segmentation technology to segment different point cloud categories in the transmission line, uses a waypoint verification algorithm to complete waypoint route optimization, and finally generates a UAV automatic positioning route to improve the safety of UAV automatic inspection.

[0006] To achieve the above object, the present invention adopts the following technical solutions:

[0007] In a first aspect, the present invention provides a method for correcting a transmission line UAV route based on point cloud recognition, which specifically comprises the following steps:

[0008] S1. Obtain point cloud data of power transmission lines through drones.

[0009] Furthermore, the step S1 specifically includes: using a drone equipped with a laser scanning device to obtain laser point cloud data of the transmission channel by scanning the transmission channel in accordance with the transmission line scanning specification.

[0010] S2. Using point cloud segmentation technology to identify and segment the point cloud data.

[0011] Furthermore, the step S2 specifically includes: using a deep learning algorithm to identify and segment the laser point cloud data of the power transmission channel.

[0012] S3. Determine data information in the transmission corridor based on the data after point cloud segmentation, wherein the data information includes data related to poles, towers, conductors, insulators, and ground wires.

[0013] Furthermore, the step S3 specifically includes: performing point cloud segmentation on data sets of different object categories in the transmission corridor by using a deep learning algorithm, and obtaining relevant point cloud data of towers, conductors, insulators, and ground wires.

[0014] S4. Verify the UAV waypoints using the data information in the transmission corridor.

[0015] S5. Correct the waypoint of the UAV according to the verification result.

[0016] Furthermore, the deep learning algorithm in step S2 is a PointNet++ neural network algorithm.

[0017] Furthermore, the step S2 specifically includes the following steps:

[0018] Through interpolation, concat and PointNet++ upsampling, the number of point clouds is obtained and the point cloud features are extracted. Each point is classified according to the point cloud features to obtain the segmented data information.

[0019] Furthermore, the step S2 specifically includes the following steps: the interpolation upsampling operation is performed through an interpolate structure.

[0020] Furthermore, the step S4 specifically includes: performing safety verification on the preset shooting points and target points, wherein the waypoint information is verified using a preset algorithm to obtain a verification result.

[0021] Further, the step S5 specifically includes: determining whether the waypoint has passed the safety check, if it has passed the safety check, not performing the waypoint correction; if it has not passed the safety check, using a preset optimization algorithm to perform the waypoint correction on the waypoint that has not passed the safety check.

[0022] In a second aspect, the present invention also provides a transmission line UAV route correction system based on point cloud recognition, which specifically includes:

[0023] A point cloud acquisition module is used to acquire point cloud data of the power transmission line through the UAV; the point cloud acquisition module comprises: using the UAV equipped with a laser scanning device to obtain the laser point cloud data of the power transmission channel by scanning the power transmission channel according to the power transmission line scanning specification;

[0024] An identification and segmentation module is used to identify and segment the point cloud data using point cloud segmentation technology; the identification and segmentation module specifically includes: using a deep learning algorithm to identify and segment the transmission channel laser point cloud data;

[0025] A data information determination module is used to determine the data information in the transmission corridor according to the data after point cloud segmentation, wherein the data information includes data related to poles, wires, insulators, and ground wires; the data information determination module specifically includes: performing point cloud segmentation on data sets of different object categories in the transmission corridor by using a deep learning algorithm, and obtaining relevant point cloud data of poles, wires, insulators, and ground wires;

[0026] A verification module, used to verify the UAV waypoints using the data information in the transmission corridor;

[0027] The correction module corrects the waypoints of the drone according to the verification results.

[0028] Furthermore, the correction module specifically includes: determining whether the waypoint has passed the safety check, if it has passed the safety check, not performing the waypoint correction; if it has not passed the safety check, using a preset optimization algorithm to perform the waypoint correction on the waypoint that has not passed the safety check.

[0029] In a third aspect, the present invention further provides an electronic device, comprising:

[0030] at least one processor, and a memory communicatively coupled to at least one of the processors;

[0031] Among them, the memory stores instructions that can be executed by at least one of the processors, and the instructions are executed by at least one of the processors so that at least one of the processors can execute the above-mentioned transmission line drone route correction method based on point cloud recognition.

[0032] In a fourth aspect, the present invention further provides a computer-readable storage medium having computer instructions stored thereon, wherein the computer instructions are used to enable the computer to execute the above-mentioned method for transmission line drone route correction based on point cloud recognition.

[0033] The beneficial effects of the present invention are as follows:

[0034] The present invention proposes a method for automatic verification of transmission line UAV routes based on point cloud, which obtains transmission line data by identifying and segmenting the point cloud data of the transmission line, performs safety verification on preset shooting points and target points, and regenerates shooting points based on the verification results, thereby improving the safety and effectiveness of the shooting points. By adjusting the captured image, the shooting factor can be adaptively adjusted to ensure that the UAV captures complete and effective images, so as to facilitate the subsequent analysis process, thereby improving the accuracy and effectiveness of the captured images. The present invention saves labor costs, provides a more accurate route by using an automatic verification method, realizes the optimization of automatic planning of waypoints by UAVs, and finally generates a new automatic positioning route for UAVs, thereby realizing automatic inspection of transmission lines by UAVs. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] In order to more clearly illustrate the specific implementation methods of the present invention or the technical solutions in the prior art, the drawings required for use in the specific implementation methods or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some implementation methods of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0036] Figure 1 The invention provides a method for automatically verifying the route of a UAV for a power transmission line based on point cloud;

[0037] Figure 2 The laser point cloud image data of the power transmission corridor in the present invention;

[0038] Figure 3 This is the PointNet++ network structure diagram in the present invention;

[0039] Figure 4 This is an example diagram of an abstract projection in the present invention. DETAILED DESCRIPTION

[0040] The following is a description of exemplary embodiments of the present application in conjunction with the accompanying drawings, including various details of the embodiments of the present application to facilitate understanding, which should be considered as merely exemplary. Therefore, it should be recognized by those of ordinary skill in the art that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of the present application. Similarly, for the sake of clarity and conciseness, the description of well-known functions and structures is omitted in the following description.

[0041] The core idea of ​​the present invention is a method for automatic correction of waypoints in transmission line route planning based on point cloud recognition. First, the point cloud data of the transmission line is obtained by a drone, and the point cloud segmentation technology is used to identify and segment it, and the data of towers, conductors, insulators, ground wires, etc. in the transmission corridor are obtained. The optimization of the drone waypoints is realized through the verification algorithm and the correction algorithm. The specific implementation is as follows:

[0042] The first step is to obtain point cloud data of the transmission line corridor.

[0043] Using drones equipped with laser scanning equipment, according to the transmission line scanning specifications, the transmission channel laser point cloud data is obtained. Figure 2 shown.

[0044] In the second step, the transmission channel semantic segmentation is performed to extract key data separation.

[0045] The present invention uses the PointNet++ neural network in deep learning to segment the transmission channel point cloud. PointNet++ is an extension of PointNet, which adds a multi-level structure to enable the network to extract higher-level features and cover areas of different scales. Its design idea is to process point cloud information of different scales through a hierarchical structure to better capture the details and global features of the target and achieve segmentation of different categories in point cloud data. The network structure is as follows Figure 3 shown.

[0046] The network structure mainly consists of three parts:

[0047] 1. Set Abstraction (SA module): It is used to iteratively sample the local structure of the point cloud and continuously increase the receptive field to extract the features of the point cloud. It includes three sub-parts:

[0048] Sampling layer: samples the input points and selects several center points.

[0049] Grouping layer: Use the sampled center point to divide the point set into several areas, set different circular radii and the maximum number of points in each circular area.

[0050] PointNet layer: Encode each region, obtain the feature vector, and superimpose the feature vector in each region on the feature channel of the center point of the region.

[0051] 2. Fully connected layer (FC): Classify objects based on the collected point cloud features. The network includes three fully connected layers, and the feature dimension changes from 1024>512>256>num_classs. Among them, num_classs is the number of categories to be classified.

[0052] 3. Upsampling layer (FP): Segment the object based on the collected point cloud features. Upsampling is performed through interpolation, concat and PointNet to obtain the number of point clouds and extract point cloud features. The interpolation upsampling operation is performed through the interpolate structure, and finally softmax classification is performed on each point.

[0053] The third step is to determine the data information in the transmission corridor based on the data after point cloud segmentation.

[0054] By using the PointNet++ algorithm to perform point cloud segmentation on the data set of different object categories in the transmission corridor annotated with labelme, point cloud data of the ground, transmission lines, different types of towers, and insulators can be obtained. By using the deep learning algorithm to perform point cloud segmentation on the data set of different object categories in the transmission corridor, relevant point cloud data of towers, conductors, insulators, and ground wires can be obtained.

[0055] The fourth step is to verify the safety of transmission line waypoints.

[0056] The data information in the transmission corridor is used to verify the waypoints of the drone. The preset shooting points and target points are safety verified, wherein the waypoint information is verified using a preset algorithm to obtain a verification result.

[0057] Read the waypoint information of the route, including the latitude and longitude coordinates, altitude, heading, gimbal angle, waypoint order, and whether to take photos. Fit a straight line between each group of adjacent waypoints to simulate the movement path of the drone between the waypoints. For example, assume that the coordinates of waypoint A are (x A ,y A , z A ), the coordinates of waypoint B are (x B ,y B , z B ), then the coordinates of the two points can be combined to calculate the projection equation of x on y in the x, y coordinate system and the projection equation of x on z in the x, z coordinate system:

[0058] y=ax+b

[0059] z=cx+d

[0060] Where a is the slope of x with respect to y in the x, y coordinate system; b is the intercept of x with respect to y in the x, y coordinate system; c is the slope of x with respect to z in the x, y coordinate system; d is the intercept of x with respect to z in the x, y coordinate system.

[0061] The above two equations can be used to fit the straight line equation between A and B. At this time, a set of average distributions at x is given. A , x B The x value is substituted into the equation to generate a set of data points. When there are enough data points, these data points can be approximated as a straight line. This set of data points is the point cloud data of the line connecting A and B. AB By calculating l AB The minimum distance d from the original point cloud min , if d min is less than the specified threshold d given safe_min , the security check fails.

[0062] In order to ensure the shooting effect and capture complete and effective images, so as to facilitate the subsequent analysis process and improve the accuracy and effectiveness of the captured images, the shooting factor is adjusted through an adaptive algorithm before the safety verification, including the following methods:

[0063] Set the size of the image taken by the drone camera to M*N, where M is the length of the image and N is the height of the image, and M and N are fixed values.

[0064] Since the height of the tower and the length of the wire will change according to the actual situation, the actual height of the tower is H j , the actual length of the wire is L j , j is the data of the jth segment on the route.

[0065] When determining the shooting data of the jth segment, the height of the tower shown in the image of the first shooting of this segment is h0, and the length of the wire is l0. Both h0 and l0 are values ​​in the shooting image, and h0≤N, l0≤M.

[0066] If h0>0.8*N, or l0>0.7*M, in order to ensure that a complete image is captured between every two towers and the data in the image is authentic and reliable, the shooting factor β needs to be adjusted. i ,

[0067] β i =β0-λi

[0068] Among them, β0 is the initial shooting factor, which is used to indicate the degree of zooming of the captured image, λ is the adjustment step size, 0.001≤λ≤0.005, λ can be adjusted according to actual needs, and i is the number of adjustments;

[0069] Adjusted beta i The following conditions (1) are met:

[0070]

[0071] Among them, h i is the adjusted β i The height of the tower pole in the newly captured image, l i is the adjusted β i The length of the corresponding wire in the newly captured image and meets the following condition (2):

[0072] 0.3*N≤h i ≤0.8*N

[0073] 0.4*M≤l i ≤0.7*N

[0074] If any of the above conditions (1) and (2) is not met, continue to adjust β i Until the above conditions (1) and (2) are met.

[0075] Through the above methods, it is possible to ensure that the drone can capture complete and effective images, and the quality of the captured transmission corridor images can be guaranteed to facilitate subsequent route correction, thereby improving the accuracy and effectiveness of the captured images.

[0076] Step 5: Transmission line waypoint correction.

[0077] Determine whether the waypoint has passed the safety check. If it has passed the safety check, no waypoint correction is performed. If it has not passed the safety check, a preset optimization algorithm is used to perform waypoint correction on the waypoint that has failed the safety check. This can be achieved in the following ways.

[0078] The waypoints that failed the safety check in the fourth step above are combined into a set {A i , B i}, where A i Represents the waypoint where the line starts, B i Indicates the waypoint where the connection ends. See B i Is it a shooting waypoint? If not, it is considered B. i It is a safe transition point between two shooting waypoints, referred to as a safety point.

[0079] If B i is a safe point, then traverse B in sequence i until you find the waypoint you want to shoot. The target point of the shooting is Generate the minimum shooting distance d p_minThe sphere bounding box Sphere with a radius of and The connection l bp The intersection point with the sphere bounding box is the initial point, and the minimum safe distance d is generated with it as the center. safe_min A square S tangent to the sphere is generated with twice the side length, with S as the base, l bp Generate a rectangular bounding box RP0 for the axis. Project perpendicular to the surface S, S1 is the projection rectangle of RP0, o is the projection point of the target point, p is the projection point of the initial waypoint, p1 is the outer vertex of the projection circle C1 of the sphere of S1, angle α can be calculated through p, p1, o, rotate S1 clockwise by α degrees to obtain a new rectangle S2, the intersection of S2 and C1 is recorded as p2, p2, p1, o can calculate angle β, and α=β, such as Figure 4 As shown. This can generate rectangles, each rectangle S i An intersection point p with C1 can be obtained i . Similarly, perpendicular to the rectangle S i Projecting through the target point o, each rectangle can also obtain projection rectangles, which can be integrated to obtain Rectangles can be restored to the original coordinate system to obtain the corresponding Preset shooting points as well as A rectangular block {RP j}.

[0080] Traverse each preset shooting point in turn, calculate the shortest distance between each shooting point and the origin cloud, and extract the points with the shortest distance greater than d safe_min A collection of preset waypoints Traverse in sequence Data points within At the point With A i , a straight line is generated between Straight Line Perform a security check. If the security check passes, stop traversing. is the new waypoint, and A is removed i and A safe point between If all data points fail the safety check, a rectangle S perpendicular to the tower crossarm and the horizontal plane is generated. safe_A , S safe_A Generate rule S safe_A The minimum height is greater than the height of point A, and the maximum height is greater than the height of the tower pole, that is And at a height of A rectangular S parallel to the horizontal plane is generated above the tower safe_Z , S safe_Z With S safe_A Intersect and. At the same time S safe_A With A i The minimum distance of the connection line passes the safety check, S safe_A , S safe_Z Any point on the safe_A With A i The intersection of the lines with the minimum distance Set to A i The next destination, For safety reasons. safe_A Find out a little Make a straight line perpendicular to the horizontal plane, and Height equals S safe_A The maximum height, i.e. Similarly, we can generate the rectangle S for o safe_Z Intersecting rectangles S safe_o And generate waypoints in For S safe_o The intersection of the line with the minimum distance to o, and the intersection of this line and Sphere is the final shooting point of o Then A i and The order of waypoints between is A i ,

[0081] Through the above optimization algorithm, the waypoint data can be optimized to obtain the corrected waypoint data, and then the transmission line drone route correction based on point cloud recognition can be realized, the optimization of the automatic planning of waypoints by drones can be realized, and finally a new drone automatic positioning route can be generated, so as to realize the automatic inspection of transmission lines by drones, save labor costs, and provide a more accurate route by using automatic verification methods.

[0082] Another embodiment of the present invention further provides a transmission line UAV route correction system based on point cloud recognition, which specifically includes:

[0083] A point cloud acquisition module is used to acquire point cloud data of the power transmission line through the UAV; the point cloud acquisition module comprises: using the UAV equipped with a laser scanning device to obtain the laser point cloud data of the power transmission channel by scanning the power transmission channel according to the power transmission line scanning specification;

[0084] An identification and segmentation module is used to identify and segment the point cloud data using point cloud segmentation technology; the identification and segmentation module specifically includes: using a deep learning algorithm to identify and segment the transmission channel laser point cloud data;

[0085] A data information determination module is used to determine the data information in the transmission corridor according to the data after point cloud segmentation, wherein the data information includes data related to poles, wires, insulators, and ground wires; the data information determination module specifically includes: performing point cloud segmentation on data sets of different object categories in the transmission corridor by using a deep learning algorithm, and obtaining relevant point cloud data of poles, wires, insulators, and ground wires;

[0086] A verification module, used to verify the UAV waypoints using the data information in the transmission corridor;

[0087] The correction module corrects the waypoints of the drone according to the verification results.

[0088] The specific implementation of this system is the same as the above-mentioned transmission line drone route correction method based on point cloud recognition. Please refer to the method part of the above-mentioned embodiment and will not be repeated here.

[0089] Another embodiment of the present invention further provides an electronic device, comprising: at least one processor, and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the above-mentioned transmission line UAV route correction method based on point cloud recognition. If the specific technical details are not disclosed, please refer to the method part of the above-mentioned embodiment of the present invention.

[0090] Another embodiment of the present invention further provides a computer-readable storage medium, on which computer instructions are stored, and the computer instructions are used to enable the computer to execute the above-mentioned method for transmission line drone route correction based on point cloud recognition. If the specific technical details are not disclosed, please refer to the method part of the above-mentioned embodiment of the present invention.

[0091] The medium in the present invention may adopt any combination of one or more computer-readable media. The medium may be a computer-readable signal medium or a computer-readable storage medium. The medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination thereof. More specific examples of the medium (a non-exhaustive list) include: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof. In this document, the medium may be any tangible medium containing or storing a program that may be used by or in combination with an instruction execution system, device, or device.

[0092] The above specific implementations do not constitute a limitation on the protection scope of this application. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions and improvements made within the spirit and principles of this application should be included in the protection scope of this application.

Claims

1. A method for correcting the route of a UAV over a power transmission line based on point cloud recognition, comprising the following steps: S1. Obtaining point cloud data of a power transmission line by using a drone; the step S1 specifically includes: Using drones equipped with laser scanning equipment, according to the transmission line scanning specifications, the transmission channel is scanned to obtain the transmission channel laser point cloud data; S2. Using point cloud segmentation technology to identify and segment the point cloud data; the step S2 specifically includes: using a deep learning algorithm to identify and segment the transmission channel laser point cloud data; S3. Determine the data information in the transmission corridor according to the data after point cloud segmentation, wherein the data information includes data related to poles, wires, insulators, and ground wires; the step S3 specifically includes: performing point cloud segmentation on data sets of different object categories in the transmission corridor by using a deep learning algorithm, and obtaining relevant point cloud data of poles, wires, insulators, and ground wires; S4. Verify the UAV waypoints using the data information in the transmission corridor; Before the safety check, the shooting factor is adjusted through an adaptive algorithm, including the following methods: Set the size of the image taken by the drone camera to , where M is the length of the image and N is the height of the image, where M and N are fixed values; Since the height of the tower and the length of the wire will change according to the actual situation, the actual height of the tower is H. j , the actual length of the wire is L j , j is the data of the jth segment on the route; When determining the shooting data of the jth segment, the height of the tower shown in the first shooting of the segment is h0, and the length of the wire is l0. h0 and l0 are both values ​​in the shooting image, and h0 ,l0 ; like ,or In order to ensure that a complete image is captured between every two towers and the data in the image is authentic and reliable, the shooting factor needs to be adjusted. , in, is the initial shooting factor, which is used to indicate the degree of zooming of the captured image. is the adjustment step size, 0.001 , It can be adjusted according to actual needs. is the number of adjustments; After adjustment The following conditions (1) are met: ; in, It has been adjusted The height of the tower pole in the newly captured image is It has been adjusted The length of the corresponding wire in the newly captured image and meets the following conditions (2): ; If any of the above conditions (1) and (2) are not met, continue to adjust Until the above conditions (1) and (2) are met; Through the above methods, it is possible to ensure that the drone can capture complete and effective images, and the quality of the images of the transmission corridor can be guaranteed, so as to facilitate subsequent route correction, thereby improving the accuracy and effectiveness of the captured images; S5. Correct the waypoint of the UAV according to the verification result.

2. The method according to claim 1, characterized in that The deep learning algorithm in step S2 is the PointNet++ neural network algorithm.

3. The method according to claim 2, characterized in that The step S2 specifically includes the following steps: Through interpolation, concat and PointNet++ upsampling, the number of point clouds is obtained and the point cloud features are extracted. Each point is classified according to the point cloud features to obtain the segmented data information.

4. The method according to claim 3, characterized in that The step S2 specifically includes the following steps: the interpolation upsampling operation is performed through an interpolate structure.

5. The method according to claim 1, characterized in that The step S4 specifically includes: performing safety verification on the preset shooting points and target points, wherein the waypoint information is verified using a preset algorithm to obtain a verification result.

6. The method according to claim 1, characterized in that The step S5 specifically includes: determining whether the waypoint has passed the safety check, if it has passed the safety check, not performing the waypoint correction; if it has not passed the safety check, using a preset optimization algorithm to perform the waypoint correction on the waypoint that has not passed the safety check.

7. A transmission line UAV route correction system based on point cloud recognition, characterized in that: Specifically include: Point cloud acquisition module, used to obtain point cloud data of power transmission lines through drones; The point cloud acquisition module comprises: using a laser scanning device mounted on a drone to obtain laser point cloud data of a transmission channel by scanning the transmission channel in accordance with the transmission line scanning specification; An identification and segmentation module is used to identify and segment the point cloud data using point cloud segmentation technology; the identification and segmentation module specifically includes: using a deep learning algorithm to identify and segment the transmission channel laser point cloud data; A data information determination module is used to determine the data information in the transmission corridor according to the data after point cloud segmentation, wherein the data information includes data related to poles, wires, insulators, and ground wires; the data information determination module specifically includes: performing point cloud segmentation on data sets of different object categories in the transmission corridor by using a deep learning algorithm, and obtaining relevant point cloud data of poles, wires, insulators, and ground wires; A verification module, used to verify the UAV waypoints using the data information in the transmission corridor; Before the safety check, the shooting factor is adjusted through an adaptive algorithm, including the following methods: Set the size of the image taken by the drone camera to , where M is the length of the image and N is the height of the image, where M and N are fixed values; Since the height of the tower and the length of the wire will change according to the actual situation, the actual height of the tower is H. j , the actual length of the wire is L j , j is the data of the jth segment on the route; When determining the shooting data of the jth segment, the height of the tower shown in the first shooting of the segment is h0, and the length of the wire is l0. h0 and l0 are both values ​​in the shooting image, and h0 ,l0 ; like ,or In order to ensure that a complete image is captured between every two towers and the data in the image is authentic and reliable, the shooting factor needs to be adjusted. , in, is the initial shooting factor, which is used to indicate the degree of zooming of the captured image. is the adjustment step size, 0.001 , It can be adjusted according to actual needs. is the number of adjustments; After adjustment The following conditions (1) are met: ; in, It has been adjusted The height of the tower pole in the newly captured image is It has been adjusted The length of the corresponding wire in the newly captured image and meets the following conditions (2): ; If any of the above conditions (1) and (2) are not met, continue to adjust Until the above conditions (1) and (2) are met; Through the above methods, it is possible to ensure that the drone can capture complete and effective images, and the quality of the images of the transmission corridor can be guaranteed, so as to facilitate subsequent route correction, thereby improving the accuracy and effectiveness of the captured images; The correction module corrects the waypoints of the drone according to the verification results.

8. The system according to claim 7, characterized in that The correction module specifically includes: determining whether the waypoint has passed the safety check, if it has passed the safety check, not correcting the waypoint; if it has not passed the safety check, correcting the waypoint for the waypoint that has not passed the safety check using a preset optimization algorithm.

9. An electronic device, characterized in that: The electronic device comprises: at least one processor, and a memory communicatively coupled to at least one of the processors; The memory stores instructions executable by at least one of the processors, and the instructions are executed by at least one of the processors so that the at least one processor can execute the method described in any one of claims 1 to 6.

10. A computer-readable storage medium, characterized in that: The medium stores computer instructions, and the computer instructions are used to enable the computer to execute the method according to any one of claims 1 to 6.

Images