Photovoltaic panel positioning modeling method and system based on UAV image measurement

Abstract

The present invention relates to the technical field of photovoltaic panel positioning modeling, and is a method and system for photovoltaic panel positioning modeling based on drone image measurement, comprising: identifying regional features of a photovoltaic panel measurement area, obtaining adjustable parameters of the drone, fitting a positioning path, driving the drone to fly, and during the flight, capturing images in real time to obtain a detection image set, performing image preprocessing operations on the detection image set to obtain a clear detection image set, performing feature recognition on each clear detection image in the clear detection image set to obtain photovoltaic panel features, obtaining a target monitoring heading shooting angle and a target monitoring lateral shooting angle, obtaining a target recognition image set, obtaining a photovoltaic panel model, storing the photovoltaic panel model, obtaining a photovoltaic panel benchmark model, and completing photovoltaic panel positioning modeling based on drone image measurement based on the photovoltaic panel benchmark model. The present invention can improve the efficiency and accuracy of photovoltaic panel positioning measurement.

Description

Technical Field

[0001] The present invention relates to the technical field of photovoltaic panel positioning modeling, and in particular to a photovoltaic panel positioning modeling method and system based on drone image measurement. Background Art

[0002] Drone imaging is the process of using drones equipped with image acquisition equipment to acquire and analyze relevant image data while flying over a specific area. Photovoltaic panel positioning modeling uses algorithms and techniques to determine the exact position of photovoltaic panels in three-dimensional space based on data obtained from drone imaging, and then construct a three-dimensional model of the panels and the surrounding environment.

[0003] Traditional photovoltaic panel positioning and inspection relies on workers carrying measuring tools to measure the position and related parameters of each panel within the photovoltaic power plant. This method is not only inefficient but also labor-intensive and time-consuming. Furthermore, some photovoltaic power plants use fixed monitoring equipment with a limited monitoring range, failing to fully cover the entire plant. This makes it difficult to effectively monitor photovoltaic panels in remote or obscured areas. Therefore, improving the efficiency and accuracy of photovoltaic panel positioning and inspection is an urgent technical challenge. Summary of the Invention

[0004] The present invention provides a photovoltaic panel positioning modeling method based on drone image measurement and a computer-readable storage medium, the main purpose of which is to improve the efficiency and accuracy of photovoltaic panel positioning measurement.

[0005] To achieve the above objectives, the present invention provides a photovoltaic panel positioning modeling method based on drone image measurement, comprising:

[0006] Confirming a photovoltaic panel measurement area and identifying regional features of the photovoltaic panel measurement area, wherein the regional features include: height and slope, and the photovoltaic panel measurement area includes multiple photovoltaic panels;

[0007] Obtaining the adjustable parameters of the drone, where the adjustable parameters of the drone include: flight altitude, heading shooting angle, and sideways shooting angle;

[0008] Fitting a positioning path according to the photovoltaic panel measurement area, regional characteristics, and adjustable parameters of the drone, using the positioning path to drive a pre-built drone to fly, and during the flight, using preset shooting parameters to capture images in real time to obtain a detection image set;

[0009] Performing image preprocessing on each detection image in the detection image set to obtain a clear detection image set, and performing feature recognition on each clear detection image in the clear detection image set to obtain photovoltaic panel features;

[0010] Obtain target monitoring heading shooting angle and target monitoring side shooting angle according to the characteristics of the photovoltaic panel, and obtain a target recognition image set based on the target monitoring heading shooting angle and target monitoring side shooting angle;

[0011] Obtaining a photovoltaic panel model based on a target recognition image set;

[0012] The photovoltaic panel model is stored to obtain a photovoltaic panel benchmark model, and the photovoltaic panel positioning modeling based on drone image measurement is completed based on the photovoltaic panel benchmark model.

[0013] Optionally, fitting a positioning path according to the photovoltaic panel measurement area, regional characteristics, and adjustable parameters of the drone includes:

[0014] Gridding the photovoltaic panel measurement area to obtain a grid measurement area set, wherein the grid measurement area set includes a plurality of grid measurement areas, and each grid measurement area has the same area;

[0015] Obtain the flight start and end points of the drone based on the grid measurement area set, and generate multiple initial positioning paths based on the pre-built route planning algorithm, the flight start and end points;

[0016] Extract initial positioning paths from multiple initial positioning paths in sequence, and perform the following operations on each of the extracted initial positioning paths:

[0017] Calculate the actual energy consumption of the drone based on the initial positioning path and regional characteristics, and compare the actual energy consumption with the preset energy consumption threshold;

[0018] If it is confirmed that the actual energy consumption is greater than the preset energy consumption threshold, the initial positioning path is eliminated, and the retained initial positioning paths are summarized to obtain multiple retained positioning paths. The multiple retained positioning paths are used as multiple initial positioning paths, and the process returns to the step of sequentially extracting initial positioning paths from the multiple initial positioning paths until the actual energy consumption is less than or equal to the preset energy consumption threshold, thereby obtaining an optimized positioning path.

[0019] The optimized positioning paths are summarized to obtain an optimized positioning path set, and the optimized positioning path corresponding to the minimum actual energy consumption is extracted from the optimized positioning path set to obtain a positioning path.

[0020] Optionally, calculating the actual energy consumption of the UAV based on the initial positioning path and regional characteristics includes:

[0021] Obtaining air density based on the flight altitude and regional characteristics, obtaining regional wind direction and regional wind speed in the photovoltaic panel measurement area, and obtaining the flight direction angle of the UAV based on the regional wind direction and regional wind speed;

[0022] Obtain the flight speed of the drone based on the flight direction angle, regional wind direction, and regional wind speed, and calculate the air resistance based on the flight speed;

[0023] Calculate the energy consumption to overcome the resistance based on the air resistance and air density, obtain the energy consumption to overcome the gravity of the UAV, obtain the path length of the initial positioning path, and obtain the heading shooting angle change and the side shooting angle change based on the heading shooting angle and the side shooting angle;

[0024] The energy consumption of adjusting the shooting angle is calculated according to the path length, the change in the heading shooting angle, and the change in the side shooting angle. The energy consumption calculation formula of adjusting the shooting angle is as follows:

[0025] W z =k(|Δα|+|Δβ|)×s

[0026] Among them, W z represents the energy consumption of camera angle adjustment, k represents the preset attitude adjustment force coefficient, Δα represents the change of heading camera angle, Δβ represents the change of side camera angle, and s represents the path length;

[0027] The actual energy consumption of the drone is calculated based on the energy consumption for overcoming resistance, the energy consumption for overcoming gravity, the energy consumption for adjusting the shooting angle, and the preset power efficiency of the drone.

[0028] Optionally, the calculating the actual energy consumption of the drone based on the energy consumption for overcoming resistance, the energy consumption for overcoming gravity, the energy consumption for adjusting the shooting angle, and a preset power efficiency of the drone includes:

[0029] The actual energy consumption of the drone is calculated based on the energy consumption for overcoming resistance, the energy consumption for overcoming gravity, the energy consumption for adjusting the camera angle, the preset drone power efficiency, and the pre-built total energy consumption formula. The total energy consumption formula is as follows:

[0030]

[0031] Among them, E represents actual energy consumption, W d Indicates the energy consumption to overcome resistance, W g represents the energy consumption to overcome gravity, η represents the preset UAV power efficiency, ρ0 represents the air density, e represents the natural constant, H h Indicates the flight altitude, H0 indicates the preset reference altitude, H indicates the preset atmospheric altitude, v indicates the regional wind direction, v w represents the regional wind speed, cosθ represents the flight direction angle, C d represents the preset air resistance coefficient, and A represents the preset projected area.

[0032] Optionally, performing an image preprocessing operation on each detection image in the detection image set to obtain a clear detection image set includes:

[0033] Extract detection images from the detection image set in sequence, and perform the following operations on the extracted detection images:

[0034] Cropping the detection image to obtain a cropped image, and acquiring flight attitude data according to the shooting parameters;

[0035] Obtain the distortion model and projection model, use the distortion model and camera parameters to perform distortion correction on the cropped image to obtain an initial corrected image;

[0036] Performing geometric correction on the initial correction image using the projection model and flight attitude data to obtain a correction image;

[0037] A histogram equalization operation is performed on the corrected image to obtain a clear detection image, and the clear detection images are summarized to obtain a clear detection image set corresponding to the detection image set.

[0038] Optionally, acquiring a photovoltaic panel model based on the target recognition image set includes:

[0039] Generate a three-dimensional point cloud image using the target recognition image set, fit the photovoltaic panel geometric representation plane using the three-dimensional point cloud image, identify the photovoltaic panel geometric representation plane, and obtain the photovoltaic panel recognition plane;

[0040] Determine whether the photovoltaic panel identification plane meets the photovoltaic panel characteristics;

[0041] If the photovoltaic panel identification plane meets the photovoltaic panel characteristics, the photovoltaic panel identification plane is analyzed to obtain a photovoltaic panel model;

[0042] If the photovoltaic panel recognition plane does not meet the photovoltaic panel characteristics, then return to the step of using the positioning path to drive the pre-built drone to fly until the photovoltaic panel recognition plane meets the photovoltaic panel characteristics, thereby obtaining a photovoltaic panel model.

[0043] Optionally, fitting a photovoltaic panel geometric representation plane using a three-dimensional point cloud image includes:

[0044] Acquiring three-dimensional point cloud data according to the three-dimensional point cloud image, filtering the three-dimensional point cloud data to obtain filtered point cloud data, and obtaining a filtered point cloud coordinate set according to the filtered point cloud data, wherein the three-dimensional point cloud data includes a plurality of three-dimensional point cloud points;

[0045] Determine a plane fitting algorithm and obtain an initial plane equation according to the plane fitting algorithm, where the initial plane equation is expressed as:

[0046] z=ax+by+c

[0047] Among them, a, b, and c are the initial parameters in the initial plane equation, x represents the x-axis filtered point cloud coordinate, y represents the y-axis filtered point cloud coordinate, and z represents the z-axis filtered point cloud coordinate;

[0048] The matrix equation is constructed based on the filtered point cloud coordinate set and the initial plane equation, where the matrix equation is expressed as:

[0049]

[0050] Among them, z1 represents the first z-axis filtered point cloud coordinate, z2 represents the second z-axis filtered point cloud coordinate, and z n Indicates the nth z-axis filtered point cloud coordinate, x1 indicates the first x-axis filtered point cloud coordinate, x2 indicates the second x-axis filtered point cloud coordinate, x n Indicates the nth x-axis filtered point cloud coordinate, y1 indicates the first y-axis filtered point cloud coordinate, y2 indicates the second y-axis filtered point cloud coordinate, y n Indicates the nth y-axis filtered point cloud coordinate;

[0051] Solve the matrix equation to obtain the plane parameters, determine the fitting plane based on the plane parameters, calculate the distance between each filtered point cloud coordinate in the filtered point cloud coordinate set and the fitting plane, and obtain the point cloud distance set;

[0052] Calculate the average error of the point cloud distance set and determine whether the average error is within the preset error range;

[0053] If it is confirmed that the average error is within the preset error range, the fitting plane is confirmed as the photovoltaic panel geometric representation plane.

[0054] Optionally, filtering the three-dimensional point cloud data to obtain filtered point cloud data includes:

[0055] Extracting a plurality of sample point cloud data from the three-dimensional point cloud data, wherein the sample point cloud data includes a plurality of sample point cloud points;

[0056] Extract a sample point cloud data from multiple sample point cloud data, aggregate the retained multiple sample point cloud data to obtain multiple remaining point cloud data, and perform the following operations on the extracted sample point cloud data:

[0057] Extract sample point cloud points from the sample point cloud data in sequence, and perform the following operations on the extracted sample point cloud points:

[0058] Obtain the neighborhood point set of the sample point cloud point, calculate the distance between the sample point cloud point and each neighborhood point in the neighborhood point set, and obtain the neighborhood distance set;

[0059] Determine whether there is a neighborhood distance in the neighborhood distance set that is not within the preset standard distance interval;

[0060] If there is a neighborhood distance in the neighborhood distance set that is not within the preset standard distance interval, the sample point cloud point corresponding to the neighborhood distance is confirmed as an abnormal point cloud point;

[0061] If there is no neighborhood distance in the neighborhood distance set that is not within the preset standard distance interval, returning to the step of sequentially extracting sample point cloud points from the sample point cloud data until the sample point cloud data is an empty set;

[0062] Summarize abnormal point cloud points to obtain an abnormal point cloud point set, and build an abnormal point cloud model based on the abnormal point cloud point set;

[0063] The abnormal point cloud model is used to filter multiple remaining point cloud data to obtain filtered point cloud data.

[0064] Optionally, calculating the average error of the point cloud distance set includes:

[0065] Get the median, maximum point cloud distance and minimum point cloud distance in the point cloud distance set;

[0066] The average error is calculated using the median, maximum point cloud distance, minimum point cloud distance, and point cloud distance set. The average error calculation formula is as follows:

[0067]

[0068] in, represents the average error, n represents the number of point cloud distance sets, d i Represents the i-th point cloud distance, i represents the index of the point cloud distance set, m represents the median, d max Indicates the maximum point cloud distance, d min Represents the minimum point cloud distance, γ1 represents the weight coefficient of the arithmetic mean of the point cloud distance, γ2 represents the weight coefficient of the median, and γ3 represents the weight coefficient of the mean of the maximum point cloud distance and the minimum point cloud distance.

[0069] To achieve the above objectives, the present invention further provides a photovoltaic panel positioning modeling system based on drone image measurement, comprising:

[0070] A path fitting module is used to confirm the photovoltaic panel measurement area, identify the regional characteristics of the photovoltaic panel measurement area, where the regional characteristics include: height and slope, the photovoltaic panel measurement area includes multiple photovoltaic panels, and obtain the drone's adjustable parameters, where the drone's adjustable parameters include: flight altitude, heading shooting angle, and sideways shooting angle;

[0071] The image acquisition module is used to drive the pre-built UAV to fly using the positioning path, and during the flight, it uses the preset shooting parameters to capture images in real time to obtain a detection image set;

[0072] an image analysis module, configured to perform image preprocessing on each detection image in the detection image set to obtain a clear detection image set, perform feature recognition on each clear detection image in the clear detection image set to obtain photovoltaic panel features, obtain a target monitoring heading shooting angle and a target monitoring side shooting angle based on the photovoltaic panel features, and obtain a target recognition image set based on the target monitoring heading shooting angle and the target monitoring side shooting angle;

[0073] The modeling completion module is used to obtain a photovoltaic panel model based on a target recognition image set, store the photovoltaic panel model, obtain a photovoltaic panel benchmark model, and complete the photovoltaic panel positioning modeling based on drone image measurement based on the photovoltaic panel benchmark model.

[0074] In order to solve the above problem, the present invention further provides an electronic device, comprising:

[0075] a memory storing at least one instruction;

[0076] The processor executes the instructions stored in the memory to implement the above-mentioned photovoltaic panel positioning modeling method based on drone image measurement.

[0077] In order to solve the above problems, the present invention also provides a computer-readable storage medium, which stores at least one instruction, and the at least one instruction is executed by a processor in an electronic device to implement the above-mentioned photovoltaic panel positioning modeling method based on drone image measurement.

[0078] The present invention is to solve the problems described in the background technology. The present invention confirms the photovoltaic panel measurement area and identifies the regional characteristics of the photovoltaic panel measurement area, wherein the regional characteristics include: height and slope. The photovoltaic panel measurement area includes multiple photovoltaic panels. The present invention clarifies the photovoltaic panel measurement area, which helps to limit the operating range of the drone to the target area, avoid unnecessary flight and data collection, improve work efficiency, reduce resource waste, and obtain drone adjustable parameters, wherein the drone adjustable parameters include: flight altitude, heading shooting angle and lateral shooting angle. Different photovoltaic panel measurement areas and regional characteristics of the present invention may require different shooting angles and altitudes. By obtaining adjustable parameters such as flight altitude, heading shooting angle and lateral shooting angle, it can be adjusted according to actual conditions. The invention can flexibly adjust the flight and shooting mode of the drone to obtain the best image data and meet the needs of photovoltaic panel information collection in different scenarios. According to the photovoltaic panel measurement area, regional characteristics and the drone's adjustable parameters, the positioning path is fitted, and the pre-built drone is driven to fly using the positioning path. During the flight, images are taken in real time using the preset shooting parameters to obtain a detection image set. The invention combines the measurement area, regional characteristics and the drone's adjustable parameters to fit the positioning path, and can plan the optimal flight route to ensure that the drone covers the entire photovoltaic panel measurement area while reducing the energy consumption required for the drone to locate the photovoltaic panel. The image preprocessing operation is performed on each detection image in the detection image set to obtain a clear detection image set. Each clear detection image in the clear detection image set is subjected to feature recognition to obtain photovoltaic panel features. The image preprocessing operation of the present invention can remove interference factors such as noise and distortion in the image, improve the clarity and quality of the image, make the features of the photovoltaic panel more obvious, and facilitate subsequent feature recognition work. The target monitoring heading shooting angle and the target monitoring lateral shooting angle are obtained according to the photovoltaic panel features, and the target recognition image set is obtained based on the target monitoring heading shooting angle and the target monitoring lateral shooting angle. The present invention determines the target monitoring heading shooting angle and the target monitoring lateral shooting angle according to the photovoltaic panel features, and can obtain clearer and more accurate photovoltaic panel image data in a targeted manner, highlight the key features of the photovoltaic panel, reduce the interference of irrelevant information, and improve subsequent The accuracy of model construction is achieved by obtaining a photovoltaic panel model based on a target recognition image set. The present invention utilizes the rich information in the target recognition image set to construct an intuitive and accurate three-dimensional model of the photovoltaic panel. The model can truly reflect the actual shape, position and layout of the photovoltaic panel, and provide a visual decision-making basis for the planning, design, operation and maintenance of photovoltaic power stations. The photovoltaic panel model is stored to obtain a photovoltaic panel benchmark model. Based on the photovoltaic panel benchmark model, photovoltaic panel positioning modeling based on drone image measurement is completed. The present invention stores the photovoltaic panel model as a photovoltaic panel benchmark model to facilitate subsequent query, comparison and analysis. At the same time, the benchmark model can be used as a reference standard and reused in the subsequent construction, transformation and operation and maintenance of the photovoltaic power station to improve work efficiency and accuracy.Therefore, the present invention can improve the efficiency and accuracy of photovoltaic panel positioning measurement. BRIEF DESCRIPTION OF THE DRAWINGS

[0079] Figure 1 A schematic diagram of a flow chart of a photovoltaic panel positioning modeling method based on drone image measurement provided by one embodiment of the present invention;

[0080] Figure 2 This is a functional module diagram of a photovoltaic panel positioning modeling system based on drone image measurement provided by one embodiment of the present invention;

[0081] Figure 3 A schematic structural diagram of an electronic device for implementing the photovoltaic panel positioning modeling method based on drone image measurement provided by one embodiment of the present invention.

[0082] Description of reference numerals:

[0083] 1. Electronic device; 10. Processor; 11. Memory; 12. Bus.

[0084] The purpose, features and advantages of the present invention will be further described with reference to the accompanying drawings and in conjunction with the embodiments. DETAILED DESCRIPTION

[0085] It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0086] The embodiment of the present application provides a photovoltaic panel positioning modeling method based on drone image measurement. The execution subject of the photovoltaic panel positioning modeling method based on drone image measurement includes but is not limited to at least one of the electronic devices such as a server, a terminal, etc. that can be configured to execute the method provided by the embodiment of the present application. In other words, the photovoltaic panel positioning modeling method based on drone image measurement can be executed by software or hardware installed on a terminal device or a server device, and the software can be a blockchain platform. The server includes but is not limited to: a single server, a server cluster, a cloud server or a cloud server cluster, etc.

[0087] Reference Figure 1 FIG. 1 is a flow chart of a photovoltaic panel positioning modeling method based on drone image measurement provided by an embodiment of the present invention. In this embodiment, the photovoltaic panel positioning modeling method based on drone image measurement includes:

[0088] S1. Confirm the photovoltaic panel measurement area and identify the regional features of the photovoltaic panel measurement area.

[0089] Specifically, the regional characteristics include: height and slope, and the photovoltaic panel measurement area includes multiple photovoltaic panels.

[0090] It should be clarified that the PV panel measurement area refers to the geographic area containing the PV panels to be measured, analyzed, and modeled. For example, a large centralized PV power plant covers the area occupied by numerous PV arrays on a large piece of land, or a small distributed PV power plant, such as the rooftop area where PV panels are installed on a residential roof.

[0091] It should be explained that identifying the regional characteristics of the photovoltaic panel measurement area refers to identifying the photovoltaic panel measurement area using a geographic information system to obtain regional characteristics. The use of a geographic information system to identify the photovoltaic panel measurement area in the embodiment of the present invention is a prior art and will not be repeated here.

[0092] S2. Obtaining adjustable parameters of the drone, wherein the adjustable parameters of the drone include: flight altitude, heading shooting angle, and sideways shooting angle.

[0093] It should be explained that obtaining the drone's adjustable parameters refers to obtaining them from the drone's technical manual or related equipment documentation. The drone's adjustable parameters consist of flight altitude, heading angle, and sideways angle. Flight altitude refers to the drone's vertical distance from the ground. The heading angle refers to the angle between the drone's camera's optical axis in the flight direction and the vertical. The sideways angle refers to the angle between the drone's camera's optical axis, perpendicular to the flight direction, and the vertical.

[0094] S3. Fitting a positioning path according to the photovoltaic panel measurement area, regional characteristics, and adjustable parameters of the drone.

[0095] In detail, the positioning path is fitted according to the photovoltaic panel measurement area, regional characteristics and adjustable parameters of the drone, including:

[0096] Gridding the photovoltaic panel measurement area to obtain a grid measurement area set, wherein the grid measurement area set includes a plurality of grid measurement areas, and each grid measurement area has the same area;

[0097] Obtain the flight start and end points of the drone based on the grid measurement area set, and generate multiple initial positioning paths based on the pre-built route planning algorithm, the flight start and end points;

[0098] Extract initial positioning paths from multiple initial positioning paths in sequence, and perform the following operations on each of the extracted initial positioning paths:

[0099] Calculate the actual energy consumption of the drone based on the initial positioning path and regional characteristics, and compare the actual energy consumption with the preset energy consumption threshold;

[0100] If it is confirmed that the actual energy consumption is greater than the preset energy consumption threshold, the initial positioning path is eliminated, and the retained initial positioning paths are summarized to obtain multiple retained positioning paths. The multiple retained positioning paths are used as multiple initial positioning paths, and the process returns to the step of sequentially extracting initial positioning paths from the multiple initial positioning paths until the actual energy consumption is less than or equal to the preset energy consumption threshold, thereby obtaining an optimized positioning path.

[0101] The optimized positioning paths are summarized to obtain an optimized positioning path set, and the optimized positioning path corresponding to the minimum actual energy consumption is extracted from the optimized positioning path set to obtain a positioning path.

[0102] It should be explained that the gridded photovoltaic panel measurement area refers to the operation of gridding the photovoltaic panel measurement area using a division tool and a preset grid size. For example, the division tool can be ArcGIS, QGIS, and GiobalMapper. The grid size is a pre-set size. The grid measurement area set refers to the set of all grid measurement areas obtained after gridding the photovoltaic panel measurement area. The acquisition of the flight starting point and flight end point of the drone based on the grid measurement area set refers to selecting a vertex on the boundary of the photovoltaic panel measurement area close to the drone's take-off point and entering the grid measurement area as the flight starting point, and selecting a vertex in the grid measurement area on the boundary of the photovoltaic panel measurement area that is relative to the flight starting point as the flight end point.

[0103] For example, the grid measurement area is in a rectangular coordinate system. When the drone takes off from the south side of the grid measurement area, the southwest corner of the grid with coordinates (50, 10) located on the southern boundary of the area is selected as the flight start point, and the northeast corner of the grid with coordinates (50, 200) located on the northern boundary of the area is selected as the flight end point.

[0104] Importantly, a route planning algorithm refers to an algorithm used to generate a UAV flight path. Examples include genetic algorithms, ant colony algorithms, and Dijkstra's algorithm. The initial positioning path is a set of multiple feasible UAV flight paths generated based on the route planning algorithm, the flight start point, and the flight end point. The energy consumption threshold is a pre-set upper limit to ensure that the UAV will not be unable to complete its mission or return safely due to excessive energy consumption during flight. Retained positioning paths are those paths that remain after energy consumption screening of the initial positioning paths. Optimized positioning paths are those paths that are continuously compared with a preset energy consumption threshold, eliminating paths with actual energy consumption exceeding the threshold until all retained positioning paths have actual energy consumption less than or equal to the threshold. The optimized positioning path set is a set of optimized positioning paths. A positioning path is the path with the lowest actual energy consumption extracted from the optimized positioning path set. The purpose of the positioning path is to enable the UAV to capture all images used to characterize the photovoltaic panel measurement area during flight while minimizing energy consumption during the photovoltaic panel positioning process.

[0105] Specifically, the actual energy consumption of the UAV is calculated based on the initial positioning path and regional characteristics, including:

[0106] Obtaining air density based on the flight altitude and regional characteristics, obtaining regional wind direction and regional wind speed in the photovoltaic panel measurement area, and obtaining the flight direction angle of the UAV based on the regional wind direction and regional wind speed;

[0107] Obtain the flight speed of the drone based on the flight direction angle, regional wind direction, and regional wind speed, and calculate the air resistance based on the flight speed;

[0108] Calculate the energy consumption to overcome the resistance based on the air resistance and air density, obtain the energy consumption to overcome the gravity of the UAV, obtain the path length of the initial positioning path, and obtain the heading shooting angle change and the side shooting angle change based on the heading shooting angle and the side shooting angle;

[0109] The energy consumption of adjusting the shooting angle is calculated according to the path length, the change in the heading shooting angle, and the change in the side shooting angle. The energy consumption calculation formula of adjusting the shooting angle is as follows:

[0110] W z =k(|Δα|+|Δβ|)×s

[0111] Among them, W z represents the energy consumption of camera angle adjustment, k represents the preset attitude adjustment force coefficient, Δα represents the change of heading camera angle, Δβ represents the change of side camera angle, and s represents the path length;

[0112] The actual energy consumption of the drone is calculated based on the energy consumption for overcoming resistance, the energy consumption for overcoming gravity, the energy consumption for adjusting the shooting angle, and the preset power efficiency of the drone.

[0113] It should be explained that the air density obtained according to the flight altitude and regional characteristics means that the air density will decrease as the flight altitude and regional characteristics increase. The regional wind direction and regional wind speed obtained in the photovoltaic panel measurement area refer to the regional wind direction and regional wind speed obtained by equipment such as meteorological stations, meteorological satellites, and ground wind speed and direction meters in the photovoltaic panel measurement area. The flight direction angle refers to the angle between the flight direction of the drone and the regional wind direction. The flight speed refers to the speed at which the drone flies. When the drone flies against the wind, the flight speed will decrease, and when the drone flies with the wind, the flight speed will increase. The formula for calculating the air resistance in the step of calculating the air resistance according to the flight speed is as follows:

[0114]

[0115] Among them, F d represents air resistance, v f Indicates flight speed.

[0116] It is understood that the energy consumption for overcoming resistance refers to the energy consumption calculated by air resistance and path length. The energy consumption for overcoming gravity refers to the energy consumption obtained by multiplying the mass of the drone, the flight altitude, and the acceleration due to gravity. The path length is the length of the initial positioning path. The change in the heading shooting angle refers to the change in the heading shooting angle of the drone on the initial positioning path. The change in the lateral shooting angle refers to the change in the lateral shooting angle of the drone on the initial positioning path. The energy consumption for adjusting the shooting angle is the energy consumption calculated by the attitude adjustment force coefficient, the heading shooting angle change, the lateral shooting angle change, and the path length. The attitude adjustment force coefficient refers to a preset coefficient that reflects the proportional relationship between the energy required by the drone to adjust the heading shooting angle and the lateral shooting angle, the angle change, and the path length.

[0117] Specifically, the actual energy consumption of the drone is calculated based on the energy consumption for overcoming resistance, the energy consumption for overcoming gravity, the energy consumption for adjusting the shooting angle, and the preset power efficiency of the drone, including:

[0118] The actual energy consumption of the drone is calculated based on the energy consumption for overcoming resistance, the energy consumption for overcoming gravity, the energy consumption for adjusting the camera angle, the preset drone power efficiency, and the pre-built total energy consumption formula. The total energy consumption formula is as follows:

[0119]

[0120] Among them, E represents actual energy consumption, W d Indicates the energy consumption to overcome resistance, W grepresents the energy consumption to overcome gravity, η represents the preset UAV power efficiency, ρ0 represents the air density, e represents the natural constant, H h Indicates the flight altitude, H0 indicates the preset reference altitude, H indicates the preset atmospheric altitude, v indicates the regional wind direction, v w represents the regional wind speed, cosθ represents the flight direction angle, C d represents the preset air resistance coefficient, and A represents the preset projected area.

[0121] It should be explained that actual energy consumption refers to the energy actually consumed by a drone during flight to overcome air resistance, gravity, and camera angle adjustments. The base altitude refers to a pre-set reference altitude value used to describe the relationship between air density and flight altitude. The drag coefficient reflects the relationship between the amount of air resistance encountered by a drone in the air and factors such as the drone's shape and surface smoothness. For example, a streamlined drone has a low drag coefficient, which can reduce energy consumption and improve flight efficiency. Projected area refers to the projected area of ​​the drone in the direction of flight. The projected area described in this embodiment of the present invention is derived from the drone's design specifications. Atmospheric scale altitude is a physical quantity related to the change in air density with altitude. It is the characteristic length at which air density decays exponentially with altitude. Drone power efficiency refers to a pre-set measure of the efficiency of the drone's power system in converting input energy into the energy required for efficient flight.

[0122] S4. Use the positioning path to drive the pre-built UAV to fly, and during the flight, use the preset shooting parameters to shoot images in real time to obtain a detection image set.

[0123] It should be explained that a drone is an unmanned aerial vehicle controlled by a self-contained programmable control device. Shooting parameters are pre-set parameters used to obtain test images during the drone's flight. Examples include aperture, shutter speed, and focal length. A test image set is a collection of images captured using pre-set shooting parameters while the drone follows a positioning path.

[0124] S5. Perform image preprocessing on each detection image in the detection image set to obtain a clear detection image set.

[0125] Specifically, performing image preprocessing on each detection image in the detection image set to obtain a clear detection image set includes:

[0126] Extract detection images from the detection image set in sequence, and perform the following operations on the extracted detection images:

[0127] Cropping the detection image to obtain a cropped image, and acquiring flight attitude data according to the shooting parameters;

[0128] Obtain the distortion model and projection model, use the distortion model and camera parameters to perform distortion correction on the cropped image to obtain an initial corrected image;

[0129] Performing geometric correction on the initial correction image using the projection model and flight attitude data to obtain a correction image;

[0130] A histogram equalization operation is performed on the corrected image to obtain a clear detection image, and the clear detection images are summarized to obtain a clear detection image set corresponding to the detection image set.

[0131] It should be explained that the cropping of the detection image refers to the operation of cropping the detection image according to a preset image size. Flight attitude data refers to the spatial attitude data of the UAV during flight, including: the heading and roll of the UAV. Heading refers to the flight direction of the UAV. Roll refers to the rotational movement of the UAV around the longitudinal axis of the fuselage. The longitudinal axis of the fuselage is a straight line running from the nose to the tail of the UAV. The distortion model is a mathematical model used to describe the image distortion law produced by the camera lens of the UAV. The distortion model described in the embodiment of the present invention can perform distortion correction on the cropped image and restore the distorted image to an image that is closer to the real scene. The projection model is a mathematical model that describes the projection relationship from the photovoltaic panel in three-dimensional space to the two-dimensional image plane. The projection model described in the embodiment of the present invention, combined with the flight attitude data, can be used to perform geometric correction on the initial corrected image after distortion correction, eliminate the image deformation caused by the shooting attitude and projection relationship, and restore the photovoltaic panel in the image to the correct geometric shape and position relationship.

[0132] Importantly, geometric correction refers to the operation of processing the initial corrected image using a projection model and flight attitude data to eliminate the geometric deformation of the initial corrected image caused by factors such as the UAV's flight attitude and shooting angle. The corrected image refers to the image obtained after geometric correction of the initial corrected image. Histogram equalization is an image processing technology used to enhance image contrast. The histogram equalization operation performed on the corrected image in the embodiment of the present invention refers to the operation of adjusting the grayscale histogram of the corrected image and remapping the grayscale value distribution of the corrected image so that the grayscale value of the corrected image is more evenly distributed throughout the grayscale range, thereby enhancing the overall contrast of the corrected image and making the details in the corrected image more clearly visible. Histogram equalization is a prior art and will not be described in detail here.

[0133] S6. Perform feature recognition on each clear detection image in the clear detection image set to obtain photovoltaic panel features.

[0134] It should be noted that the clear detection image set is a collection of all clear detection images. Determining photovoltaic panel features refers to extracting features from the clear detection image set using a CNN model to obtain photovoltaic panel features. The CNN model is a historically trained model used to identify photovoltaic panel features. Photovoltaic panel features include panel shape and texture.

[0135] S7. Obtain a target monitoring heading shooting angle and a target monitoring side shooting angle according to the characteristics of the photovoltaic panel, and obtain a target recognition image set based on the target monitoring heading shooting angle and the target monitoring side shooting angle.

[0136] It should be explained that the target monitoring heading shooting angle refers to the shooting angle of the drone along the flight direction relative to a specific reference direction (such as the front, geographic north, etc.) when monitoring photovoltaic panels. The target monitoring side shooting angle refers to the side shooting angle of the drone relative to the flight direction during flight when monitoring photovoltaic panels, that is, the shooting angle in a direction perpendicular to the target monitoring heading shooting angle. The target recognition image set refers to the collection of images captured by the drone based on the target monitoring heading shooting angle and the target monitoring side shooting angle.

[0137] S8. Obtain a photovoltaic panel model based on the target recognition image set.

[0138] In detail, the method of acquiring a photovoltaic panel model based on a target recognition image set includes:

[0139] Generate a three-dimensional point cloud image using the target recognition image set, fit the photovoltaic panel geometric representation plane using the three-dimensional point cloud image, identify the photovoltaic panel geometric representation plane, and obtain the photovoltaic panel recognition plane;

[0140] Determine whether the photovoltaic panel identification plane meets the photovoltaic panel characteristics;

[0141] If the photovoltaic panel identification plane meets the photovoltaic panel characteristics, the photovoltaic panel identification plane is analyzed to obtain a photovoltaic panel model;

[0142] If the photovoltaic panel recognition plane does not meet the photovoltaic panel characteristics, then return to the step of using the positioning path to drive the pre-built drone to fly until the photovoltaic panel recognition plane meets the photovoltaic panel characteristics, thereby obtaining a photovoltaic panel model.

[0143] It should be explained that the step of generating a three-dimensional point cloud map using the target recognition image set is a prior art and will not be repeated here. The photovoltaic panel recognition plane refers to a plane obtained after further judgment and confirmation on the basis of the identified geometric characterization plane of the photovoltaic panel. The step of parsing the photovoltaic panel recognition plane to obtain a photovoltaic panel model is as follows: extracting the geometric parameters of the photovoltaic panel recognition plane, including the position of the plane (coordinates in three-dimensional space), size (length, width), tilt angle, etc., combining the image information in the target recognition image set, extracting the texture and material information of the photovoltaic panel recognition plane, integrating the extracted geometric parameters, texture and material information, and constructing a photovoltaic panel model. The photovoltaic panel model refers to a three-dimensional virtual model that integrates multiple feature information such as the geometric shape, size, position, texture, and material of the photovoltaic panel.

[0144] In detail, the method of fitting the photovoltaic panel geometric representation plane using the three-dimensional point cloud image includes:

[0145] Acquiring three-dimensional point cloud data according to the three-dimensional point cloud image, filtering the three-dimensional point cloud data to obtain filtered point cloud data, and obtaining a filtered point cloud coordinate set according to the filtered point cloud data, wherein the three-dimensional point cloud data includes a plurality of three-dimensional point cloud points;

[0146] Determine a plane fitting algorithm and obtain an initial plane equation according to the plane fitting algorithm, where the initial plane equation is expressed as:

[0147] z=ax+by+c

[0148] Among them, a, b, and c are the initial parameters in the initial plane equation, x represents the x-axis filtered point cloud coordinate, y represents the y-axis filtered point cloud coordinate, and z represents the z-axis filtered point cloud coordinate;

[0149] The matrix equation is constructed based on the filtered point cloud coordinate set and the initial plane equation, where the matrix equation is expressed as:

[0150]

[0151] Among them, z1 represents the first z-axis filtered point cloud coordinate, z2 represents the second z-axis filtered point cloud coordinate, and z n Indicates the nth z-axis filtered point cloud coordinate, x1 indicates the first x-axis filtered point cloud coordinate, x2 indicates the second x-axis filtered point cloud coordinate, x n Indicates the nth x-axis filtered point cloud coordinate, y1 indicates the first y-axis filtered point cloud coordinate, y2 indicates the second y-axis filtered point cloud coordinate, y n Indicates the nth y-axis filtered point cloud coordinate;

[0152] Solve the matrix equation to obtain the plane parameters, determine the fitting plane based on the plane parameters, calculate the distance between each filtered point cloud coordinate in the filtered point cloud coordinate set and the fitting plane, and obtain the point cloud distance set;

[0153] Calculate the average error of the point cloud distance set and determine whether the average error is within the preset error range;

[0154] If it is confirmed that the average error is within the preset error range, the fitting plane is confirmed as the photovoltaic panel geometric representation plane.

[0155] It should be explained that the three-dimensional point cloud image refers to a graphic representation of the digital representation of the photovoltaic panel in three-dimensional space. Three-dimensional point cloud data refers to data containing multiple three-dimensional point cloud points. Filtered point cloud data refers to data obtained after filtering the three-dimensional point cloud data. The purpose of filtering the three-dimensional point cloud data is to remove abnormal data such as noise points and outliers in the three-dimensional point cloud data, so as to make the data smoother and more accurate, thereby improving the accuracy and reliability of subsequent plane fitting. The acquisition of the filtered point cloud coordinate set based on the filtered point cloud data means that each point in the filtered point cloud data has a corresponding three-dimensional coordinate (x, y, z), and the x-coordinate, y-coordinate and z-coordinate of all points in the filtered point cloud data are extracted respectively, and arranged in sequence according to the order of the points to obtain the filtered point cloud coordinate set. The initial parameter refers to an unknown parameter. The filtered point cloud coordinate set refers to a set consisting of all filtered point cloud coordinates.

[0156] Importantly, the plane fitting algorithm is an algorithm for finding a plane in three-dimensional point cloud data so that the plane can best approximate the distribution of these three-dimensional point cloud data. The plane fitting algorithm described in the embodiment of the present invention refers to the least squares method. The initial plane equation is a mathematical equation used to represent the fitting plane. The solution of the matrix equation is a prior art and will not be described in detail here. The plane parameters refer to the parameters obtained by solving the matrix equation. The fitting plane refers to the plane determined according to the plane parameters obtained by the solution. The point cloud distance set refers to a set consisting of the distance between each filtered point cloud coordinate in the filtered point cloud coordinate set and the fitting plane. The calculation formula in the step of calculating the distance between each filtered point cloud coordinate in the filtered point cloud coordinate set and the fitting plane is the point-to-plane distance formula. The error interval is a pre-set interval used to determine whether the fitting plane meets the requirements.

[0157] In detail, filtering the three-dimensional point cloud data to obtain filtered point cloud data includes:

[0158] Extracting a plurality of sample point cloud data from the three-dimensional point cloud data, wherein the sample point cloud data includes a plurality of sample point cloud points;

[0159] Extract a sample point cloud data from multiple sample point cloud data, aggregate the retained multiple sample point cloud data to obtain multiple remaining point cloud data, and perform the following operations on the extracted sample point cloud data:

[0160] Extract sample point cloud points from the sample point cloud data in sequence, and perform the following operations on the extracted sample point cloud points:

[0161] Obtain the neighborhood point set of the sample point cloud point, calculate the distance between the sample point cloud point and each neighborhood point in the neighborhood point set, and obtain the neighborhood distance set;

[0162] Determine whether there is a neighborhood distance in the neighborhood distance set that is not within the preset standard distance interval;

[0163] If there is a neighborhood distance in the neighborhood distance set that is not within the preset standard distance interval, the sample point cloud point corresponding to the neighborhood distance is confirmed as an abnormal point cloud point;

[0164] If there is no neighborhood distance in the neighborhood distance set that is not within the preset standard distance interval, returning to the step of sequentially extracting sample point cloud points from the sample point cloud data until the sample point cloud data is an empty set;

[0165] Summarize abnormal point cloud points to obtain an abnormal point cloud point set, and build an abnormal point cloud model based on the abnormal point cloud point set;

[0166] The abnormal point cloud model is used to filter multiple remaining point cloud data to obtain filtered point cloud data.

[0167] It should be explained that the sample point cloud data refers to a portion of point cloud data extracted from the three-dimensional point cloud data. The remaining point cloud data refers to the remaining sample point cloud data after extracting a sample point cloud data from multiple sample point cloud data. The step of obtaining the neighborhood point set of the sample point cloud point is: setting a distance threshold, selecting a sample point cloud point from the sample point cloud data, calculating the distance from the sample point cloud point to all other sample point cloud points, and collecting the sample point cloud points whose distance is less than or equal to the distance threshold to obtain a neighborhood point set. The calculation formula in the step of calculating the distance between the sample point cloud point and each neighborhood point in the neighborhood point set to obtain the neighborhood distance set is as follows:

[0168]

[0169] Among them, o represents the neighborhood distance, (t0, p0, q0) represents the sample point cloud point, and (t1, p1, q1) represents the neighborhood point.

[0170] Importantly, the neighborhood distance set refers to a set consisting of the distances between a sample point cloud point and each of its neighborhood points in the neighborhood point set. An abnormal point cloud point refers to a sample point cloud point in the sample point cloud data, for which the neighborhood distance set has a neighborhood distance that is not within a preset standard distance interval. An abnormal point cloud point set refers to a set consisting of all abnormal point cloud points. Constructing an abnormal point cloud model based on the abnormal point cloud point set means using the abnormal point cloud point set as training data, and using a machine learning algorithm (such as a clustering algorithm, a classification algorithm, etc.) for training to obtain a model that can identify abnormal points. Using the abnormal point cloud model to filter multiple remaining point cloud data to obtain filtered point cloud data means identifying and removing abnormal points that match the characteristics of the abnormal point cloud model from the remaining point cloud data, thereby obtaining filtered point cloud data. The standard distance interval refers to a pre-set distance range.

[0171] In detail, the calculation of the average error of the point cloud distance set includes:

[0172] Get the median, maximum point cloud distance and minimum point cloud distance in the point cloud distance set;

[0173] The average error is calculated using the median, maximum point cloud distance, minimum point cloud distance, and point cloud distance set. The average error calculation formula is as follows:

[0174]

[0175] in, represents the average error, n represents the number of point cloud distance sets, d i Represents the i-th point cloud distance, i represents the index of the point cloud distance set, m represents the median, d max Indicates the maximum point cloud distance, d min Represents the minimum point cloud distance, γ1 represents the weight coefficient of the arithmetic mean of the point cloud distance, γ2 represents the weight coefficient of the median, and γ3 represents the weight coefficient of the mean of the maximum point cloud distance and the minimum point cloud distance.

[0176] It should be explained that the average error refers to the average value of the point cloud distance set. The weight coefficient of the arithmetic mean of the point cloud distance is the coefficient that determines the weight of the arithmetic mean of the point cloud distance in the final average error calculation. When the weight coefficient of the arithmetic mean of the point cloud distance is large, it means that the arithmetic mean of the point cloud distance has a greater impact on the average error. The weight coefficient of the median determines the coefficient of the weight of the median of the point cloud distance in the final average error calculation. When the weight coefficient of the median is large, it means that the weight coefficient of the median has a greater impact on the average error. The weight coefficient of the maximum point cloud distance and the minimum point cloud distance average determines the coefficient of the weight of the maximum point cloud distance and the minimum point cloud distance average in the average error calculation. When the weight coefficient of the median is large, it means that the maximum point cloud distance and the minimum point cloud distance average have a greater impact on the average error. In this embodiment of the present invention, γ1+γ2+γ3=1.

[0177] S9. Store the photovoltaic panel model, obtain a photovoltaic panel benchmark model, and complete photovoltaic panel positioning modeling based on drone image measurement based on the photovoltaic panel benchmark model.

[0178] It should be noted that storing the photovoltaic panel model refers to storing the photovoltaic panel model file on a local computer's hard drive or an external storage device. For example, the external storage device may be a USB flash drive or portable hard drive. The photovoltaic panel benchmark model refers to the model obtained after this storage. This photovoltaic panel benchmark model can be used as a reference standard and reused during the subsequent construction, renovation, and operation and maintenance of photovoltaic power plants, improving efficiency and accuracy.

[0179] The present invention is to solve the problems described in the background technology. The present invention confirms the photovoltaic panel measurement area and identifies the regional characteristics of the photovoltaic panel measurement area, wherein the regional characteristics include: height and slope. The photovoltaic panel measurement area includes multiple photovoltaic panels. The present invention clarifies the photovoltaic panel measurement area, which helps to limit the operating range of the drone to the target area, avoid unnecessary flight and data collection, improve work efficiency, reduce resource waste, and obtain drone adjustable parameters, wherein the drone adjustable parameters include: flight altitude, heading shooting angle and lateral shooting angle. Different photovoltaic panel measurement areas and regional characteristics of the present invention may require different shooting angles and altitudes. By obtaining adjustable parameters such as flight altitude, heading shooting angle and lateral shooting angle, it can be adjusted according to actual conditions. The invention can flexibly adjust the flight and shooting mode of the drone to obtain the best image data and meet the needs of photovoltaic panel information collection in different scenarios. According to the photovoltaic panel measurement area, regional characteristics and the drone's adjustable parameters, the positioning path is fitted, and the pre-built drone is driven to fly using the positioning path. During the flight, images are taken in real time using the preset shooting parameters to obtain a detection image set. The invention combines the measurement area, regional characteristics and the drone's adjustable parameters to fit the positioning path, and can plan the optimal flight route to ensure that the drone covers the entire photovoltaic panel measurement area while reducing the energy consumption required for the drone to locate the photovoltaic panel. The image preprocessing operation is performed on each detection image in the detection image set to obtain a clear detection image set. Each clear detection image in the clear detection image set is subjected to feature recognition to obtain photovoltaic panel features. The image preprocessing operation of the present invention can remove interference factors such as noise and distortion in the image, improve the clarity and quality of the image, make the features of the photovoltaic panel more obvious, and facilitate subsequent feature recognition work. The target monitoring heading shooting angle and the target monitoring lateral shooting angle are obtained according to the photovoltaic panel features, and the target recognition image set is obtained based on the target monitoring heading shooting angle and the target monitoring lateral shooting angle. The present invention determines the target monitoring heading shooting angle and the target monitoring lateral shooting angle according to the photovoltaic panel features, and can obtain clearer and more accurate photovoltaic panel image data in a targeted manner, highlight the key features of the photovoltaic panel, reduce the interference of irrelevant information, and improve subsequent The accuracy of model construction is achieved by obtaining a photovoltaic panel model based on a target recognition image set. The present invention utilizes the rich information in the target recognition image set to construct an intuitive and accurate three-dimensional model of the photovoltaic panel. The model can truly reflect the actual shape, position and layout of the photovoltaic panel, and provide a visual decision-making basis for the planning, design, operation and maintenance of photovoltaic power stations. The photovoltaic panel model is stored to obtain a photovoltaic panel benchmark model. Based on the photovoltaic panel benchmark model, photovoltaic panel positioning modeling based on drone image measurement is completed. The present invention stores the photovoltaic panel model as a photovoltaic panel benchmark model to facilitate subsequent query, comparison and analysis. At the same time, the benchmark model can be used as a reference standard and reused in the subsequent construction, transformation and operation and maintenance of the photovoltaic power station to improve work efficiency and accuracy.Therefore, the present invention can improve the efficiency and accuracy of photovoltaic panel positioning measurement.

[0180] like Figure 2 , which is a functional module diagram of a photovoltaic panel positioning modeling system based on drone image measurement provided by one embodiment of the present invention.

[0181] The photovoltaic panel positioning modeling system 100 based on drone imagery described in the present invention can be installed in an electronic device. Depending on the functionality implemented, the photovoltaic panel positioning modeling system 100 based on drone imagery can include a path fitting module 101, an image acquisition module 102, an image analysis module 103, and a modeling completion module 104. The modules described in the present invention, also referred to as units, refer to a series of computer program segments that can be executed by an electronic device processor and can perform fixed functions, and are stored in the memory of the electronic device.

[0182] The path fitting module 101 is used to determine the photovoltaic panel measurement area, identify regional characteristics of the photovoltaic panel measurement area, wherein the regional characteristics include: height and slope, the photovoltaic panel measurement area includes multiple photovoltaic panels, and obtain drone adjustable parameters, wherein the drone adjustable parameters include: flight altitude, heading shooting angle, and sideways shooting angle;

[0183] The image acquisition module 102 is used to drive the pre-built UAV to fly using the positioning path, and to capture images in real time using preset shooting parameters during the flight to obtain a detection image set;

[0184] The image analysis module 103 is configured to perform an image preprocessing operation on each detection image in the detection image set to obtain a clear detection image set, perform feature recognition on each clear detection image in the clear detection image set to obtain photovoltaic panel features, obtain a target monitoring heading shooting angle and a target monitoring side shooting angle based on the photovoltaic panel features, and obtain a target recognition image set based on the target monitoring heading shooting angle and the target monitoring side shooting angle;

[0185] The modeling completion module 104 is used to obtain a photovoltaic panel model based on the target recognition image set, store the photovoltaic panel model, obtain a photovoltaic panel benchmark model, and complete the photovoltaic panel positioning modeling based on the drone image measurement based on the photovoltaic panel benchmark model.

[0186] In detail, each module in the photovoltaic panel positioning modeling system 100 based on drone image measurement in the embodiment of the present invention adopts the same method as above when in use. Figure 1 The photovoltaic panel positioning modeling method based on drone image measurement described in the previous section uses the same technical means and can produce the same technical effects, so I will not go into details here.

[0187] like Figure 3 , which is a structural diagram of an electronic device for implementing a photovoltaic panel positioning modeling method based on drone image measurement provided by an embodiment of the present invention.

[0188] The electronic device 1 may include a processor 10, a memory 11 and a bus 12, and may also include a computer program stored in the memory 11 and executable on the processor 10, such as a photovoltaic panel positioning modeling method program based on drone image measurement.

[0189] Among them, the memory 11 includes at least one type of readable storage medium, and the readable storage medium includes a flash memory, a mobile hard disk, a multimedia card, a card-type memory (for example, SD or DX memory, etc.), a magnetic memory, a disk, an optical disk, etc. In some embodiments, the memory 11 can be an internal storage unit of the electronic device 1, such as a mobile hard disk of the electronic device 1. In other embodiments, the memory 11 can also be an external storage device of the electronic device 1, such as a plug-in mobile hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (Flash Card), etc. equipped on the electronic device 1. Furthermore, the memory 11 also includes an internal storage unit of the electronic device 1 and an external storage device. The memory 11 can not only be used to store application software and various types of data installed on the electronic device 1, such as the code of the photovoltaic panel positioning modeling method program based on drone image measurement, but can also be used to temporarily store data that has been output or is to be output.

[0190] In some embodiments, the processor 10 may be composed of an integrated circuit, such as a single packaged integrated circuit, or a plurality of packaged integrated circuits with the same or different functions, including one or more central processing units (CPUs), microprocessors, digital processing chips, graphics processors, and a combination of various control chips. The processor 10 is the control core (Control Unit) of the electronic device, connecting the various components of the entire electronic device using various interfaces and lines, and executing programs or modules stored in the memory 11 (such as a photovoltaic panel positioning modeling method program based on drone image measurement, etc.), as well as calling data stored in the memory 11, to perform various functions of the electronic device 1 and process data.

[0191] The bus 12 may be a peripheral component interconnect (PCI) bus or an extended industry standard architecture (EISA) bus, etc. The bus 12 may be divided into an address bus, a data bus, a control bus, etc. The bus 12 is configured to enable communication between the memory 11 and at least one processor 10, etc.

[0192] Figure 3 Only the electronic device with components is shown, and it can be understood by those skilled in the art that Figure 3 The structure shown does not constitute a limitation on the electronic device 1 , and may include fewer or more components than shown in the figure, or combine certain components, or arrange the components differently.

[0193] For example, although not shown, the electronic device 1 may further include a power source (such as a battery) for powering the various components. Preferably, the power source may be logically connected to the at least one processor 10 via a power management device, thereby implementing functions such as charging management, discharging management, and power consumption management through the power management device. The power source may further include any components such as one or more DC or AC power sources, a recharging device, a power failure detection circuit, a power converter or inverter, a power status indicator, etc. The electronic device 1 may further include various sensors, Bluetooth modules, Wi-Fi modules, etc., which will not be described in detail here.

[0194] Furthermore, the electronic device 1 may also include a network interface. Optionally, the network interface may include a wired interface and / or a wireless interface (such as a WI-FI interface, a Bluetooth interface, etc.), which is generally used to establish a communication connection between the electronic device 1 and other electronic devices.

[0195] Optionally, the electronic device 1 may further include a user interface, which may be a display or an input unit (such as a keyboard). Optionally, the user interface may also be a standard wired interface or a wireless interface. Optionally, in some embodiments, the display may be an LED display, a liquid crystal display, a touch-sensitive liquid crystal display, or an OLED (Organic Light-Emitting Diode) touch device. The display may also be appropriately referred to as a display screen or a display unit, which is used to display information processed in the electronic device 1 and to display a visual user interface.

[0196] The photovoltaic panel positioning modeling method program based on drone image measurement stored in the memory 11 of the electronic device 1 is a combination of multiple instructions. When running in the processor 10, it can achieve the following:

[0197] Confirming a photovoltaic panel measurement area and identifying regional features of the photovoltaic panel measurement area, wherein the regional features include: height and slope, and the photovoltaic panel measurement area includes multiple photovoltaic panels;

[0198] Obtaining the adjustable parameters of the drone, where the adjustable parameters of the drone include: flight altitude, heading shooting angle, and sideways shooting angle;

[0199] Fitting a positioning path according to the photovoltaic panel measurement area, regional characteristics, and adjustable parameters of the drone, using the positioning path to drive a pre-built drone to fly, and during the flight, using preset shooting parameters to capture images in real time to obtain a detection image set;

[0200] Performing image preprocessing on each detection image in the detection image set to obtain a clear detection image set, and performing feature recognition on each clear detection image in the clear detection image set to obtain photovoltaic panel features;

[0201] Obtain target monitoring heading shooting angle and target monitoring side shooting angle according to the characteristics of the photovoltaic panel, and obtain a target recognition image set based on the target monitoring heading shooting angle and target monitoring side shooting angle;

[0202] Obtaining a photovoltaic panel model based on a target recognition image set;

[0203] The photovoltaic panel model is stored to obtain a photovoltaic panel benchmark model, and the photovoltaic panel positioning modeling based on drone image measurement is completed based on the photovoltaic panel benchmark model.

[0204] Specifically, the specific implementation method of the processor 10 for the above instructions can refer to Figures 1 to 3 The description of the relevant steps in the corresponding embodiments will not be repeated here.

[0205] Furthermore, if the modules / units integrated into the electronic device 1 are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. The computer-readable storage medium can be volatile or non-volatile. For example, the computer-readable medium can include: any entity or device capable of carrying the computer program code, a recording medium, a USB flash drive, a mobile hard drive, a magnetic disk, an optical disk, a computer memory, or a read-only memory (ROM).

[0206] The present invention further provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program. When the computer program is executed by a processor of an electronic device, the computer program can implement:

[0207] Confirming a photovoltaic panel measurement area and identifying regional features of the photovoltaic panel measurement area, wherein the regional features include: height and slope, and the photovoltaic panel measurement area includes multiple photovoltaic panels;

[0208] Obtaining the adjustable parameters of the drone, where the adjustable parameters of the drone include: flight altitude, heading shooting angle, and sideways shooting angle;

[0209] Fitting a positioning path according to the photovoltaic panel measurement area, regional characteristics, and adjustable parameters of the drone, using the positioning path to drive a pre-built drone to fly, and during the flight, using preset shooting parameters to capture images in real time to obtain a detection image set;

[0210] Performing image preprocessing on each detection image in the detection image set to obtain a clear detection image set, and performing feature recognition on each clear detection image in the clear detection image set to obtain photovoltaic panel features;

[0211] Obtain target monitoring heading shooting angle and target monitoring side shooting angle according to the characteristics of the photovoltaic panel, and obtain a target recognition image set based on the target monitoring heading shooting angle and target monitoring side shooting angle;

[0212] Obtaining a photovoltaic panel model based on a target recognition image set;

[0213] The photovoltaic panel model is stored to obtain a photovoltaic panel benchmark model, and the photovoltaic panel positioning modeling based on drone image measurement is completed based on the photovoltaic panel benchmark model.

[0214] In the several embodiments provided by the present invention, it should be understood that the disclosed devices, systems and methods can be implemented in other ways. For example, the system embodiments described above are only exemplary, and actual implementations may have other division methods.

[0215] The modules described as separate components may or may not be physically separate, and the components shown as modules may or may not be physical units, that is, they may be located in one place or distributed across multiple network elements. Some or all of the modules may be selected to achieve the purpose of the solution of this embodiment according to actual needs.

[0216] In addition, the functional modules in various embodiments of the present invention may be integrated into a single processing unit, each unit may exist physically separately, or two or more units may be integrated into a single unit. The aforementioned integrated units may be implemented in the form of hardware or hardware plus software functional modules.

[0217] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0218] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not limiting. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A photovoltaic panel positioning modeling method based on drone image measurement, characterized in that: The method comprises: Confirming a photovoltaic panel measurement area and identifying regional features of the photovoltaic panel measurement area, wherein the regional features include: height and slope, and the photovoltaic panel measurement area includes multiple photovoltaic panels; Obtaining the adjustable parameters of the drone, where the adjustable parameters of the drone include: flight altitude, heading shooting angle, and sideways shooting angle; Fitting a positioning path according to the photovoltaic panel measurement area, regional characteristics, and adjustable parameters of the drone, driving a pre-built drone to fly using the positioning path, and capturing images in real time using preset shooting parameters during the flight to obtain a detection image set, wherein fitting a positioning path according to the photovoltaic panel measurement area, regional characteristics, and adjustable parameters of the drone includes: Gridding the photovoltaic panel measurement area to obtain a grid measurement area set, wherein the grid measurement area set includes a plurality of grid measurement areas, and each grid measurement area has the same area; Obtain the flight start and end points of the drone based on the grid measurement area set, and generate multiple initial positioning paths based on the pre-built route planning algorithm, the flight start and end points; Extract initial positioning paths from multiple initial positioning paths in sequence, and perform the following operations on each of the extracted initial positioning paths: Calculate the actual energy consumption of the drone based on the initial positioning path and regional characteristics, and compare the actual energy consumption with the preset energy consumption threshold; If it is confirmed that the actual energy consumption is greater than the preset energy consumption threshold, the initial positioning path is eliminated, and the retained initial positioning paths are summarized to obtain multiple retained positioning paths. The multiple retained positioning paths are used as multiple initial positioning paths, and the process returns to the step of sequentially extracting initial positioning paths from the multiple initial positioning paths until the actual energy consumption is less than or equal to the preset energy consumption threshold, thereby obtaining an optimized positioning path. Summarize the optimized positioning paths to obtain an optimized positioning path set, extract the optimized positioning path corresponding to the minimum actual energy consumption from the optimized positioning path set, and obtain the positioning path; Performing image preprocessing on each detection image in the detection image set to obtain a clear detection image set, and performing feature recognition on each clear detection image in the clear detection image set to obtain photovoltaic panel features; Obtain target monitoring heading shooting angle and target monitoring side shooting angle according to the characteristics of the photovoltaic panel, and obtain a target recognition image set based on the target monitoring heading shooting angle and target monitoring side shooting angle; Obtaining a photovoltaic panel model based on a target recognition image set; The photovoltaic panel model is stored to obtain a photovoltaic panel benchmark model, and the photovoltaic panel positioning modeling based on drone image measurement is completed based on the photovoltaic panel benchmark model.

2. The photovoltaic panel positioning modeling method based on drone image measurement according to claim 1, characterized in that: The actual energy consumption of the UAV is calculated based on the initial positioning path and regional characteristics, including: Obtaining air density based on the flight altitude and regional characteristics, obtaining regional wind direction and regional wind speed in the photovoltaic panel measurement area, and obtaining the flight direction angle of the UAV based on the regional wind direction and regional wind speed; Obtain the flight speed of the drone based on the flight direction angle, regional wind direction, and regional wind speed, and calculate the air resistance based on the flight speed; Calculate the energy consumption to overcome the resistance based on the air resistance and air density, obtain the energy consumption to overcome the gravity of the UAV, obtain the path length of the initial positioning path, and obtain the heading shooting angle change and the side shooting angle change based on the heading shooting angle and the side shooting angle; The energy consumption of adjusting the shooting angle is calculated according to the path length, the change in the heading shooting angle, and the change in the side shooting angle. The energy consumption calculation formula of adjusting the shooting angle is as follows: W z =k(|Δα|+|Δβ|)×s Among them, W z represents the energy consumption of camera angle adjustment, k represents the preset attitude adjustment force coefficient, Δα represents the change of heading camera angle, Δβ represents the change of side camera angle, and s represents the path length; The actual energy consumption of the drone is calculated based on the energy consumption for overcoming resistance, the energy consumption for overcoming gravity, the energy consumption for adjusting the shooting angle, and the preset power efficiency of the drone.

3. The photovoltaic panel positioning modeling method based on drone image measurement according to claim 2, characterized in that: The actual energy consumption of the drone is calculated based on the energy consumption for overcoming resistance, the energy consumption for overcoming gravity, the energy consumption for adjusting the shooting angle, and the preset power efficiency of the drone, including: The actual energy consumption of the drone is calculated based on the energy consumption for overcoming resistance, the energy consumption for overcoming gravity, the energy consumption for adjusting the camera angle, the preset drone power efficiency, and the pre-built total energy consumption formula. The total energy consumption formula is as follows: Among them, E represents actual energy consumption, W d Indicates the energy consumption to overcome resistance, W g represents the energy consumption to overcome gravity, η represents the preset UAV power efficiency, ρ0 represents the air density, e represents the natural constant, H h Indicates the flight altitude, H0 indicates the preset reference altitude, H indicates the preset atmospheric altitude, v indicates the regional wind direction, v w represents the regional wind speed, cosθ represents the flight direction angle, C d represents the preset air resistance coefficient, and A represents the preset projected area.

4. The photovoltaic panel positioning modeling method based on drone image measurement according to claim 3 is characterized in that: The image preprocessing operation is performed on each detection image in the detection image set to obtain a clear detection image set, including: Extract detection images from the detection image set in sequence, and perform the following operations on the extracted detection images: Cropping the detection image to obtain a cropped image, and acquiring flight attitude data according to the shooting parameters; Obtain the distortion model and projection model, use the distortion model and camera parameters to perform distortion correction on the cropped image to obtain an initial corrected image; Performing geometric correction on the initial correction image using the projection model and flight attitude data to obtain a correction image; A histogram equalization operation is performed on the corrected image to obtain a clear detection image, and the clear detection images are summarized to obtain a clear detection image set corresponding to the detection image set.

5. The photovoltaic panel positioning modeling method based on drone image measurement according to claim 4, characterized in that: The method of obtaining a photovoltaic panel model based on a target recognition image set includes: Generate a three-dimensional point cloud image using the target recognition image set, fit the photovoltaic panel geometric representation plane using the three-dimensional point cloud image, identify the photovoltaic panel geometric representation plane, and obtain the photovoltaic panel recognition plane; Determine whether the photovoltaic panel identification plane meets the photovoltaic panel characteristics; If the photovoltaic panel identification plane meets the photovoltaic panel characteristics, the photovoltaic panel identification plane is analyzed to obtain a photovoltaic panel model; If the photovoltaic panel recognition plane does not meet the photovoltaic panel characteristics, then return to the step of using the positioning path to drive the pre-built drone to fly until the photovoltaic panel recognition plane meets the photovoltaic panel characteristics, thereby obtaining a photovoltaic panel model.

6. The photovoltaic panel positioning modeling method based on drone image measurement according to claim 5, characterized in that: The method of fitting the photovoltaic panel geometric representation plane using the three-dimensional point cloud image includes: Acquiring three-dimensional point cloud data according to the three-dimensional point cloud image, filtering the three-dimensional point cloud data to obtain filtered point cloud data, and obtaining a filtered point cloud coordinate set according to the filtered point cloud data, wherein the three-dimensional point cloud data includes a plurality of three-dimensional point cloud points; Determine a plane fitting algorithm and obtain an initial plane equation according to the plane fitting algorithm, where the initial plane equation is expressed as: z=ax+by+c Among them, a, b, and c are the initial parameters in the initial plane equation, x represents the x-axis filtered point cloud coordinate, y represents the y-axis filtered point cloud coordinate, and z represents the z-axis filtered point cloud coordinate; The matrix equation is constructed based on the filtered point cloud coordinate set and the initial plane equation, where the matrix equation is expressed as: Among them, z1 represents the first z-axis filtered point cloud coordinate, z2 represents the second z-axis filtered point cloud coordinate, and z n Indicates the nth z-axis filtered point cloud coordinate, x1 indicates the first x-axis filtered point cloud coordinate, x2 indicates the second x-axis filtered point cloud coordinate, x n Indicates the nth x-axis filtered point cloud coordinate, y1 indicates the first y-axis filtered point cloud coordinate, y2 indicates the second y-axis filtered point cloud coordinate, y n Indicates the nth y-axis filtered point cloud coordinate; Solve the matrix equation to obtain the plane parameters, determine the fitting plane based on the plane parameters, calculate the distance between each filtered point cloud coordinate in the filtered point cloud coordinate set and the fitting plane, and obtain the point cloud distance set; Calculate the average error of the point cloud distance set and determine whether the average error is within the preset error range; If it is confirmed that the average error is within the preset error range, the fitting plane is confirmed as the photovoltaic panel geometric representation plane.

7. The photovoltaic panel positioning modeling method based on drone image measurement according to claim 6, characterized in that: The filtering of the three-dimensional point cloud data to obtain filtered point cloud data includes: Extracting a plurality of sample point cloud data from the three-dimensional point cloud data, wherein the sample point cloud data includes a plurality of sample point cloud points; Extract a sample point cloud data from multiple sample point cloud data, aggregate the retained multiple sample point cloud data to obtain multiple remaining point cloud data, and perform the following operations on the extracted sample point cloud data: Extract sample point cloud points from the sample point cloud data in sequence, and perform the following operations on the extracted sample point cloud points: Obtain the neighborhood point set of the sample point cloud point, calculate the distance between the sample point cloud point and each neighborhood point in the neighborhood point set, and obtain the neighborhood distance set; Determine whether there is a neighborhood distance in the neighborhood distance set that is not within the preset standard distance interval; If there is a neighborhood distance in the neighborhood distance set that is not within the preset standard distance interval, the sample point cloud point corresponding to the neighborhood distance is confirmed as an abnormal point cloud point; If there is no neighborhood distance in the neighborhood distance set that is not within the preset standard distance interval, returning to the step of sequentially extracting sample point cloud points from the sample point cloud data until the sample point cloud data is an empty set; Summarize abnormal point cloud points to obtain an abnormal point cloud point set, and build an abnormal point cloud model based on the abnormal point cloud point set; The abnormal point cloud model is used to filter multiple remaining point cloud data to obtain filtered point cloud data.

8. The photovoltaic panel positioning modeling method based on drone image measurement according to claim 7, characterized in that: The calculation of the average error of the point cloud distance set includes: Get the median, maximum point cloud distance and minimum point cloud distance in the point cloud distance set; The average error is calculated using the median, maximum point cloud distance, minimum point cloud distance, and point cloud distance set. The average error calculation formula is as follows: in, represents the average error, n represents the number of point cloud distance sets, d i Represents the i-th point cloud distance, i represents the index of the point cloud distance set, m represents the median, d max Indicates the maximum point cloud distance, d min Represents the minimum point cloud distance, γ1 represents the weight coefficient of the arithmetic mean of the point cloud distance, γ2 represents the weight coefficient of the median, and γ3 represents the weight coefficient of the mean of the maximum point cloud distance and the minimum point cloud distance.

9. A photovoltaic panel positioning modeling system based on drone image measurement, characterized in that: The system comprises: A path fitting module is used to confirm the photovoltaic panel measurement area, identify the regional characteristics of the photovoltaic panel measurement area, where the regional characteristics include: height and slope, the photovoltaic panel measurement area includes multiple photovoltaic panels, and obtain the drone's adjustable parameters, where the drone's adjustable parameters include: flight altitude, heading shooting angle, and sideways shooting angle; An image acquisition module is configured to fit a positioning path based on the photovoltaic panel measurement area, regional characteristics, and adjustable parameters of the drone, drive a pre-built drone to fly using the positioning path, and capture images in real time during the flight using preset shooting parameters to obtain a detection image set, wherein fitting a positioning path based on the photovoltaic panel measurement area, regional characteristics, and adjustable parameters of the drone includes: Gridding the photovoltaic panel measurement area to obtain a grid measurement area set, wherein the grid measurement area set includes a plurality of grid measurement areas, and each grid measurement area has the same area; Obtain the flight start and end points of the drone based on the grid measurement area set, and generate multiple initial positioning paths based on the pre-built route planning algorithm, the flight start and end points; Extract initial positioning paths from multiple initial positioning paths in sequence, and perform the following operations on each of the extracted initial positioning paths: Calculate the actual energy consumption of the drone based on the initial positioning path and regional characteristics, and compare the actual energy consumption with the preset energy consumption threshold; If it is confirmed that the actual energy consumption is greater than the preset energy consumption threshold, the initial positioning path is eliminated, and the retained initial positioning paths are summarized to obtain multiple retained positioning paths. The multiple retained positioning paths are used as multiple initial positioning paths, and the process returns to the step of sequentially extracting initial positioning paths from the multiple initial positioning paths until the actual energy consumption is less than or equal to the preset energy consumption threshold, thereby obtaining an optimized positioning path. Summarize the optimized positioning paths to obtain an optimized positioning path set, extract the optimized positioning path corresponding to the minimum actual energy consumption from the optimized positioning path set, and obtain the positioning path; an image analysis module, configured to perform image preprocessing on each detection image in the detection image set to obtain a clear detection image set, perform feature recognition on each clear detection image in the clear detection image set to obtain photovoltaic panel features, obtain a target monitoring heading shooting angle and a target monitoring side shooting angle based on the photovoltaic panel features, and obtain a target recognition image set based on the target monitoring heading shooting angle and the target monitoring side shooting angle; The modeling completion module is used to obtain a photovoltaic panel model based on a target recognition image set, store the photovoltaic panel model, obtain a photovoltaic panel benchmark model, and complete the photovoltaic panel positioning modeling based on drone image measurement based on the photovoltaic panel benchmark model.

Images