Pole tower element numbering method based on unmanned aerial vehicle and detection system
Through drone flight and target detection technology, the latitude and longitude coordinates of power pole tower components are calculated and numbered, which solves the problem that maintenance personnel have difficulty locating faulty components and lacks unified standards, and realizes automatic numbering and precise positioning of pole tower components, improving maintenance efficiency.
Patent Information
- Application Number
- CN202510407770.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-04-02
- Publication Date
- 2025-05-02
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
During the maintenance and inspection of power towers, it is difficult for maintenance personnel to determine the exact location of the faulty components as soon as possible, and the number of power towers and components lacks unified standards, resulting in confusion of information.
The tower component numbering method based on the drone is adopted to obtain the tower cross-load elevation information and component pixel positions through drone flight, and calculate the latitude and longitude coordinates of the components and number them in combination with the target detection method. The system includes a computing module, a storage module, a display module and an operation and maintenance module to realize the automatic numbering and precise positioning of tower components.
The automatic numbering and precise positioning of tower components is realized, the rapid positioning ability of maintenance personnel to the faulty components is improved, information chaos is reduced, and the management and maintenance efficiency of power towers is improved.
Smart Images

Figure CN119919839A_ABST
Abstract
Description
Technical Field
[0001] The invention relates to the field of power pole towers, and in particular to a pole tower component numbering method and a detection system based on an unmanned aerial vehicle. Background Art
[0002] Power inspection is an indispensable work in modern industry and an important task to ensure the safety and reliability of industrial applications. At present, the power inspection technology based on drones is developing in full swing. It is an emerging technology that has emerged with the continuous development of drone technology and the needs of the power industry. Power inspection based on drones usually focuses on the fault detection of power line components, thereby replacing the inspectors to discover power line problems and further repair the problems. However, with the detection of faulty components, maintenance orders are sent to maintenance personnel. It is usually difficult for maintenance personnel to know the exact location of the faulty components in the first place. At the same time, when managing power towers, power towers and power components are usually numbered, but if the relevant numbers are not represented by an absolute value, the relevant number representation may be confusing due to the lack of a unified standard. Summary of the invention
[0003] In order to solve the above-mentioned problem of accurate positioning of power towers and power components, the present invention provides a tower component numbering method and detection system based on drones.
[0004] The present invention discloses a tower element numbering method based on an unmanned aerial vehicle. The method obtains the tower crossarm elevation information Hhi, the longitude and latitude coordinates P1 (Lo1, La1) and P2 (Lo2, La2) of the points on both sides of the longest crossarm, wherein i starts from 1 and represents the i-th crossarm from bottom to top. Meanwhile, the unmanned aerial vehicle flies around the tower and adopts a target detection method to obtain the pixel positions of the tower crossarm and the tower elements on the crossarm, wherein the pixel length PW_0 of the longest crossarm and the pixel distance PW_ij from the center of the element to the end point P1 of the longest crossarm, wherein j starts from 1 and represents different elements from left to right on the i-th crossarm, calculates the longitude and latitude coordinate positions of the tower elements and numbers them T_ij (x, y, z), wherein T_ij represents the j-th element of category T on the i-th crossarm, z=Hhi represents the elevation of the crossarm where the element is located, and x and y are calculated by a group of equations: ;
[0005] in, , indicating the latitude and longitude coordinate length of the longest cross arm.
[0006] In particular, the specific steps to obtain Hhi, P1, and P2 are: S1. The drone takes off vertically from the ground. Before taking off, the drone altitude information H0 is recorded, and target detection is started. When the cross arm is detected to appear at the center height of the image field of view, the drone altitude information Hhi is recorded; S2. The drone flies to the top of the tower, with the camera pointing vertically downward, and flies horizontally along the direction of the longest crossarm. Target detection is started. When one side of the longest crossarm is detected in the center of the image field of view, the current drone position P1 (Lo1, La1) is recorded. When the concrete pole of the tower is detected in the center of the image field of view, the current drone position P0 (Lo0, La0) is recorded. When the other side of the longest crossarm is detected in the center of the image field of view, the current drone position P2 (Lo2, La2) is recorded.
[0007] Among them, when there are multiple crossarms with the same length as the longest crossarm, the lowest crossarm is selected as the longest crossarm for reference. When there are multiple crossarms and they are not on the same line in the top view, the method is applied separately to the crossarms in different directions to establish multiple crossarm equations and perform calculations.
[0008] In particular, the data collected and saved are picture data taken by drones or the like at time intervals.
[0009] Preferably, the drone-based tower element numbering method can be used to establish a tower space model in combination with tower types.
[0010] At the same time, the present invention provides a pole tower component detection system based on drone, which includes a calculation module, a storage module, a display module and an operation and maintenance module, wherein the calculation module uses a pole tower component numbering method based on drone to complete the calculation of pole tower longitude and latitude and the calculation and numbering of pole tower component longitude and latitude, and the data is stored in the storage module. The display module displays the pole tower position information and component numbering information, and displays the inspection results of each component according to the operation and maintenance results. The operation and maintenance module is used for on-site inspection, on-site data collection and pole tower component fault detection, and the detection results are corresponded to the pole tower component numbers and stored.
[0011] Furthermore, the calculation module also selects data of flying around the tower at the same height with the tower element as the center of the field of view, obtains the position of the drone when the longest crossarm has the largest pixel size in the image, and records it as the forward position. Taking the forward position as the starting point, in the data of flying around the tower as the center, a point is selected as a typical angle every 60°, and each angle picture is obtained. The elements in the image are numbered according to the relative position of the tower element in the forward position and the current drone angle, and the corresponding pictures, element position pixel coordinates, element numbers, and pixel distances of the longest crossarm are saved in the storage module.
[0012] In particular, the calculation module also obtains the picture when the drone is in the P0 position and numbers the components in the image according to the relative positions of the tower components in the forward position, and the corresponding pictures, component position pixel coordinates, component numbers, and pixel distances of the longest crossarms are saved in the storage module.
[0013] In particular, the display module also displays data information of 7 typical angles, including 6 circumferential angles and the top center angle. The information includes at least the tower space model, images of each angle, the position of the element at each angle in the image, the element number at each angle and the operating status of the element, the image of the faulty element, and the faulty element information.
[0014] Furthermore, when the relevant information displayed by the display module is wrong, the system provides a modification interface.
[0015] Preferably, when the operation and maintenance module actually performs inspection calculations, it will extract image information from seven typical angles to perform component fault detection. The detection results will calculate the positions of all components in the image and compare them with the component positions in the corresponding typical angles. Each component will be numbered and the faulty component number will be output. The corresponding detection images and fault detection results will be stored in the storage module, and the display content of the display module will be updated.
[0016] In particular, the operation and maintenance module also includes faulty component navigation. When maintenance personnel need to conduct inspections, the operation and maintenance module will obtain the faulty component number, call out the component's spatial position, the tower base position P0, and obtain the person's current position to calculate the precise route for the maintenance personnel to the faulty component.
[0017] Furthermore, the operation and maintenance module will retrieve the tower-related information and calculate the precise route that at least includes the route to the tower and the climbing route on the tower.
[0018] In particular, the detection system also dynamically marks the relative position in the tower space model of the display module according to the longitude and latitude of the operation and maintenance personnel and the longitude and latitude of the faulty component to help the operation and maintenance personnel locate the faulty component.
[0019] In particular, operation and maintenance personnel can plan drone inspection routes in advance based on the precise longitude and latitude positions of tower components to achieve automatic drone inspections.
[0020] Beneficial effects: (1) The present invention collects data by first flying vertically to the tower and then flying along the longest cross arm of the tower. The longitude and latitude coordinates of the tower are calculated using the flight position data of the drone and the neural network target recognition technology. The longitude and latitude coordinates of the tower components are calculated by comprehensively utilizing the image similarity principle, the spatial angle conversion principle and the spatial linear equation. Finally, the components are numbered according to their categories and positions, thereby realizing automatic numbering of the tower components and associating them with the absolute longitude and latitude coordinates. Furthermore, the present invention saves the component numbers and relative positions of 7 typical angles. During subsequent inspections, it is only necessary to detect the relative positions of each component at the typical angle to quickly match the components with the corresponding numbers, thereby aggregating the data and realizing multi-dimensional management of the tower.
[0021] (2) The present invention obtains the precise latitude and longitude coordinate representation of each tower element. When the fault detection detects a fault in a certain element, by corresponding to the element number and thus the latitude and longitude of the element, the route for the maintenance personnel to reach the maintenance point can be quickly planned according to the maintenance personnel's own latitude and longitude position and the latitude and longitude position of the faulty element. In particular, by determining the tower type, the crawling route on the tower for the maintenance personnel to reach the location of the faulty element after arriving at the tower can be directly planned. The spatial model of the tower also renders the relative position of the maintenance personnel and the faulty element in real time, so that the maintenance personnel do not have to slowly search for the actual location of the faulty element. In addition, since the accurate latitude and longitude of the element is known, the operation and maintenance personnel can set the inspection route of the drone in advance and realize automatic drone inspection. In particular, when re-inspecting the faulty element, since the latitude and longitude coordinates of the element and the shooting angle when the faulty element is detected are known, the designated target can be found more quickly and the lens posture can be adjusted to obtain re-inspection data.
[0022] (3) The present invention establishes a tower space model. At the same time, since the longitude and latitude coordinates of the tower and its components can be calculated, the user can quickly view the conditions of the tower at various angles. On the one hand, this facilitates tower management. On the other hand, when the multi-angle real-time viewing function is provided to the operation and maintenance personnel, the operation and maintenance personnel can also more accurately judge the current tower condition. Finally, due to the determination of the component number, the inspection data can be saved in a more refined manner, which is convenient for data analysis. BRIEF DESCRIPTION OF THE DRAWINGS
[0023] The present invention is further described below in conjunction with the accompanying drawings and embodiments.
[0024] Figure 1 It is a schematic diagram of the logical structure of the method of the present invention; Figure 2 It is a schematic diagram of the system structure of the present invention. DETAILED DESCRIPTION
[0025] In order to make the technical means, creative features, objectives and effects achieved by the present invention easy to understand, the present invention is further explained below in conjunction with specific implementation methods.
[0026] Embodiment 1: The present invention discloses a pole tower element numbering method based on an unmanned aerial vehicle, the method comprising data collection, pole tower element position determination and numbering; Data collection: The collected data are image data taken by drones, including data of drone camera vertically flying from the ground along the pole tower, data of drone camera flying around the pole tower at the same height with the pole tower element as the center of the field of view, and data of drone camera flying vertically downward along the longest cross arm horizontally at the top of the pole tower; The tower element positions and numbers are determined and the calculation steps are as follows: S1. Select the data of the drone taking off vertically from the ground. Before taking off, record the drone elevation information H0, start target detection, and when the cross arm is detected to appear at the center height of the image field of view, record the drone elevation information Hhi, where i starts from 1, indicating the i-th cross arm from bottom to top. When the top of the tower is detected at the center height of the image field of view, record the drone height H1; S2. Select the data of the vertical downward horizontal flight of the lens on the top of the tower along the direction of the longest cross arm, start target detection, and when the side of the longest cross arm is detected at the center of the image field of view, record the current drone position and express it in longitude and latitude coordinates, recorded as P1 (Lo1, La1). When the concrete pole of the tower is detected at the center of the image field of view, record the current drone position P0 (Lo0, La0). When the other side of the longest cross arm is detected at the center of the image field of view, record the current drone position P2 (Lo2, La2), and establish the longitude and latitude straight line equation of the longest cross arm. , calculate the latitude and longitude coordinates of the longest cross arm ; S3. Select the flight data around the tower at the same height with the tower element as the center of the field of view, perform target detection on the flight image, obtain the category of the tower and the tower element and the position in the image, among which, the detected crossarms are sorted in the vertical direction and correspond to the crossarm label i in S1, and the corresponding height Hhi of the crossarm is obtained. According to the position of the drone and the longitude and latitude coordinate relationship of P1 and P2, the left and right directions of the crossarms in the image are confirmed and the pixel distance PW_0 of the longest crossarm is recorded. According to the crossarm where the element is located, the corresponding height Hhi of the element is determined and the pixel distance PW_ij from the center of the element to the end point P1 of the longest crossarm is calculated, where j starts from 1, indicating different elements from left to right on the i-th crossarm, and then the coordinates of the element are calculated and numbered T_ij(x, y,z), where z=Hhi, x and y are calculated according to the equation group: ;
[0027] Among them, when there are multiple crossarms with the same length as the longest crossarm, the lowest crossarm is selected as the longest crossarm for reference. When there are multiple crossarms and they are not on the same line in the top view, the method is applied separately to the crossarms in different directions to establish multiple crossarm equations and perform calculations.
[0028] In particular, the data collected and saved are picture data taken by the drone at time intervals, which can save drone storage resources and at the same time save the image data collected by the drone with higher quality and retain more image details.
[0029] Preferably, the drone-based tower element numbering method can be used to establish a tower space model in combination with tower types.
[0030] At the same time, the present invention provides a pole tower component detection system based on drone, which includes a calculation module, a storage module, a display module and an operation and maintenance module, wherein the calculation module uses a pole tower component numbering method based on drone to complete the calculation of pole tower longitude and latitude and the calculation and numbering of pole tower component longitude and latitude, and the data is stored in the storage module. The display module displays the pole tower position information and component numbering information, and displays the inspection results of each component according to the operation and maintenance results. The operation and maintenance module is used for on-site inspection, on-site data collection and pole tower component fault detection, and the detection results are corresponded to the pole tower component numbers and stored.
[0031] Furthermore, the calculation module also selects data of flying around the tower at the same height with the tower element as the center of the field of view, obtains the position of the drone when the longest crossarm has the largest pixel size in the image, and records it as the forward position. Taking the forward position as the starting point, in the data of flying around the tower as the center, a point is selected as a typical angle every 60°, and each angle picture is obtained. The elements in the image are numbered according to the relative position of the tower element in the forward position and the current drone angle, and the corresponding pictures, element position pixel coordinates, element numbers, and pixel distances of the longest crossarm are saved in the storage module.
[0032] In particular, the calculation module also obtains the picture when the drone is in the P0 position and numbers the components in the image according to the relative positions of the tower components in the forward position, and the corresponding pictures, component position pixel coordinates, component numbers, and pixel distances of the longest crossarms are saved in the storage module.
[0033] In particular, the display module also displays data information of 7 typical angles, including 6 circumferential angles and the top center angle. The information includes at least the tower space model, images of each angle, the position of the element at each angle in the image, the element number at each angle and the operating status of the element, the image of the faulty element, and the faulty element information.
[0034] Furthermore, when the relevant information displayed by the display module is wrong, the system provides a modification interface.
[0035] Accordingly, by adopting the data acquisition method and the tower element coordinate calculation and numbering method provided by the present invention, an absolute matching relationship between the number of each tower and the geographical location can be established. At the same time, a tower element management system is established, which can dynamically display detailed images of the towers. Moreover, because a spatial model of the tower is established, when using it, the user can directly view images of the tower at various angles by dragging the tower model (it should be noted that the 7 typical angles mentioned in the present invention are images of only the typical 7 angles that can be detected during inspection, so as to achieve the purpose of rapid detection). The user can also click on a certain element in the model to view the details of the tower. The corresponding component can be viewed in the component, including the basic information of the component, such as location, current operation status, recently updated pictures, etc. The updated pictures can view the pictures collected from various angles. Therefore, when managing the pole tower, on the one hand, the pole tower situation can be viewed more intuitively, and on the other hand, the pole tower data can be analyzed differently by angle, by component type, by fault type, etc., to obtain more multidimensional feature data. Moreover, this intuitive pole tower situation, especially the faulty component viewing when there is a fault, can share the spatial viewing model. Therefore, it is more convenient to judge the component situation and determine the operation and maintenance plan during on-site operation and maintenance. Of course, the system of the present invention also provides a manual modification interface, and when the corresponding relationship of a certain component in the pole tower data storage is wrong, it can be modified manually.
[0036] Embodiment 2: When the operation and maintenance module performs actual inspection calculations, it will extract image information from seven typical angles to detect component faults. The detection results will calculate the positions of all components in the image and compare them with the component positions in the corresponding typical angles. Each component will be numbered and the faulty component number will be output. The corresponding detection images and fault detection results will be stored in the storage module, and the display content of the display module will be updated.
[0037] In particular, the operation and maintenance module also includes faulty component navigation. When maintenance personnel need to conduct inspections, the operation and maintenance module will obtain the faulty component number, call out the component's spatial position, the tower base position P0, and obtain the person's current position to calculate the precise route for the maintenance personnel to the faulty component.
[0038] Furthermore, the operation and maintenance module will retrieve the tower-related information and calculate the precise route that at least includes the route to the tower and the climbing route on the tower.
[0039] In particular, the detection system also dynamically marks the relative position in the tower space model of the display module according to the longitude and latitude of the operation and maintenance personnel and the longitude and latitude of the faulty component to help the operation and maintenance personnel locate the faulty component.
[0040] In particular, operation and maintenance personnel can plan drone inspection routes in advance based on the precise longitude and latitude positions of tower components to achieve automatic drone inspections.
[0041] During actual maintenance, especially in some mountains where the terrain is complex and the road conditions are bad, since the longitude and latitude coordinates of each pole tower and pole tower component are known, the technicians can preset the automatic inspection route of the drone in advance and set the attitude position of the drone lens to ensure that the lens automatically aims at the pole tower target during the drone inspection. Especially when re-inspecting the faulty component, since the longitude and latitude of the faulty component are known and the shooting angle when the faulty component is detected can be clear, the drone can find the position of the component more accurately and quickly adjust the lens attitude angle to shoot and inspect the faulty component. When the component clearly needs to be repaired or replaced, the system of the present invention can obtain the location of the maintenance personnel, the location of the pole tower and the location of the faulty component, so it can provide the route for the maintenance personnel to reach the pole tower and provide the crawling route on the pole tower in combination with the type of pole tower, to prevent the maintenance personnel from reaching the bottom of the pole tower, because the pole tower is too high, it is impossible to judge whether the component is on the left or right, and blindly climb up and find that the direction is wrong. When there is no obvious crawling guide on the tower, since the maintenance personnel's own longitude and latitude, the tower's longitude and latitude and the faulty component's longitude and latitude are known, the system can at least point to the faulty component's longitude and latitude after matching the personnel's longitude and latitude with the tower's longitude and latitude. At the same time, after the maintenance order is dispatched, the tower space model in the system display module can also render the location points of the maintenance personnel and the faulty component, helping the maintenance personnel to quickly locate and prevent the target from being lost.
[0042] The above shows and describes the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The above embodiments and descriptions are only for explaining the principles of the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, which fall within the scope of the present invention. The scope of protection claimed by the present invention.
Claims
1. A tower element numbering method based on drone, characterized in that: The numbering method obtains the tower crossarm elevation information Hhi, the longitude and latitude coordinates P1 (Lo1, La1) and P2 (Lo2, La2) of the points on both sides of the longest crossarm, where i starts from 1 and represents the i-th crossarm from bottom to top. At the same time, the unmanned aerial vehicle flies around the tower and adopts the target detection method to obtain the pixel positions of the tower crossarm and the tower elements on the crossarm, where the pixel length PW_0 of the longest crossarm and the pixel distance PW_ij from the center of the element to the end point P1 of the longest crossarm, where j starts from 1 and represents the different elements from left to right on the i-th crossarm, calculates the longitude and latitude coordinate positions of the tower elements and numbers them T_ij (x, y, z), where T_ij represents the j-th element of category T on the i-th crossarm, z=Hhi, represents the elevation of the crossarm where the element is located, and x and y are calculated by the equation group: ; in, , indicating the latitude and longitude coordinate length of the longest cross arm.
2. A method for numbering pole tower elements based on drones according to claim 1, characterized in that: The specific steps of obtaining Hhi, P1, and P2 by the numbering method are as follows: S1. The drone takes off vertically from the ground. Before taking off, the drone altitude information H0 is recorded, and target detection is started. When the cross arm is detected to appear at the center height of the image field of view, the drone altitude information Hhi is recorded; S2. The drone flies to the top of the tower, with the camera pointing vertically downward, and flies horizontally along the direction of the longest crossarm. Target detection is started. When one side of the longest crossarm is detected in the center of the image field of view, the current drone position P1 (Lo1, La1) is recorded. When the concrete pole of the tower is detected in the center of the image field of view, the current drone position P0 (Lo0, La0) is recorded. When the other side of the longest crossarm is detected in the center of the image field of view, the current drone position P2 (Lo2, La2) is recorded.
3. A method for numbering pole tower elements based on drones according to claim 2, characterized in that: When there are multiple crossarms of the same length as the longest crossarm, the lowest crossarm is selected as the longest crossarm for reference.
4. A method for numbering pole tower elements based on drones according to claim 3, characterized in that: The drone-based tower element numbering method is for the case where multiple crossarms are on a plane. When there are multiple crossarms and they are not on a line in the top view, the method is applied separately to the crossarms in different directions to establish multiple crossarm equations and perform calculations.
5. A method for numbering pole tower elements based on drones according to claim 4, characterized in that: The drone-based tower element numbering method is also used to establish a tower space model in combination with tower types.
6. A tower element detection system based on drone, the system using a tower element numbering method based on drone according to any one of claims 1 to 5, characterized in that: The system includes a calculation module, a storage module, a display module and an operation and maintenance module. The calculation module uses a pole tower component numbering method based on a drone to complete the calculation of the longitude and latitude of the pole tower and the longitude and latitude height of the pole tower components and number them. The data is stored in the storage module. The display module displays the pole tower position information and component numbering information, and displays the inspection results of each component according to the operation and maintenance results. The operation and maintenance module is used for on-site inspections, on-site data collection and pole tower component fault detection, and the detection results are matched with the pole tower component numbers and stored.
7. The tower element detection system based on drone according to claim 6 is characterized in that: The calculation module also selects the data of flying around the tower at the same height with the tower element as the center of the field of view, obtains the position of the drone when the longest crossarm has the largest pixel size in the image, and records it as the positive position. Taking the positive position as the starting point, in the data of flying around the tower as the center, selects a point as a typical angle every 60°, obtains each angle picture, and numbers the elements in the image according to the relative position of the tower element in the positive position and the current drone angle. The corresponding pictures, pixel coordinates of the element position, element number, and pixel distance of the longest crossarm are saved in the storage module.
8. The tower element detection system based on drone according to claim 7 is characterized in that: The calculation module also obtains the image when the drone is in position P0 and numbers the components in the image according to the relative positions of the tower components in the forward position. The corresponding image, component position pixel coordinates, component number, and pixel distance of the longest crossarm are saved in the storage module.
9. The tower element detection system based on drone according to claim 8, characterized in that: The display module also displays data information of 7 typical angles, including 6 circumferential angles and the top center angle. The information includes at least the tower space model, images of each angle, the position of the element at each angle in the image, the element number at each angle and the operating status of the element, the image of the faulty element, and the faulty element information.
10. The tower element detection system based on drone according to claim 9, characterized in that: When the relevant information displayed by the display module is wrong, the system provides a modification interface.
11. The tower element detection system based on drone according to claim 10, characterized in that: When the operation and maintenance module performs actual inspection calculations, it will extract image information from seven typical angles to detect component faults. The detection results will calculate the positions of all components in the image and compare them with the component positions in the corresponding typical angles. Each component will be numbered and the faulty component number will be output. The corresponding detection images and fault detection results will be stored in the storage module, and the display content of the display module will be updated.
12. The tower element detection system based on drone according to claim 11, characterized in that: The operation and maintenance module also includes faulty component navigation. When maintenance personnel need to conduct inspections, the module will obtain the faulty component number, call out the component's spatial position, the tower base position P0, and obtain the person's current position to calculate the precise route for the maintenance personnel to the faulty component.
13. The tower element detection system based on drone according to claim 12, characterized in that: The operation and maintenance module will retrieve the tower-related information and calculate the precise route that includes at least the route to the tower and the crawling route on the tower.
14. The tower element detection system based on drone according to claim 10, characterized in that: The detection system also dynamically marks the relative position in the tower space model of the display module according to the longitude and latitude of the operation and maintenance personnel and the longitude and latitude of the faulty component to help the operation and maintenance personnel locate the faulty component.
15. The tower element detection system based on drone according to claim 10, characterized in that: Operation and maintenance personnel can plan drone inspection routes in advance based on the precise longitude and latitude positions of tower components to achieve automatic drone inspections.
Citation Information
Patent Citations
Method and system for establishing three-dimensional dynamic model of power transmission line tower subjected to external force deformation
CN104091029A
Method for automatically identifying defects of electric power tower part based on aerial photography of unmanned aerial vehicle
CN115471464A
Power inspection method and system based on unmanned aerial vehicle
CN119562036A
Transformer substation intelligent inspection system and method based on image recognition
WO2024040566A1