Unmanned aerial vehicle adaptive control system for image cutting

Abstract

The invention discloses an unmanned aerial vehicle adaptive control system for image cutting, and the system comprises an unmanned aerial vehicle camera attitude control module which controls an unmanned aerial vehicle camera to be in a relative vertical state with the midpoint of a tested photovoltaic panel; the unmanned aerial vehicle camera attitude control module also controls the flight height and coordinates of the unmanned aerial vehicle according to the standard vertical relative height; the standard vertical relative height is a theoretical value and is determined according to the pixel size of the tested photovoltaic panel, known unmanned aerial vehicle camera parameters and actual parameters of the tested photovoltaic panel which are set in the system; the image cutting module is used for cutting the image according to the standard vertical relative height of the image shot by the unmanned aerial vehicle camera, the real-time vertical height of the shot image and the pixel size of the tested photovoltaic panel set in the system; the method has the beneficial effects that the difference of the obtained images of different photovoltaic panels is small, and cutting is more convenient.

Description

Technical Field

[0001] The present invention relates to the technical field of unmanned aerial vehicle control and image cropping, and in particular to an unmanned aerial vehicle adaptive control system for image cropping. Background Art

[0002] In the photovoltaic inspection by drone, due to the differences in the distribution elevation and inclination angle of photovoltaic panels in photovoltaic power stations, when the drone is cruising at a fixed altitude and taking fixed-focus high-definition orthophoto photos, the size of the photovoltaic panel strings in different photos taken will vary due to the change in the relative distance of the photos. In order to facilitate the use of AI recognition and positioning annotation in the later stage, it is necessary to unify the size of the photovoltaic panel strings in the photos to further improve the management level.

[0003] At present, most UAV image acquisition and analysis systems usually use neural networks to directly crop and splice the acquired images according to pixel points or residuals. In this regard, there are the following defects. Due to the shooting lens angle, the change in the flight altitude of the UAV and the differences in altitude and inclination of different photovoltaic panels, there will be certain image pixel differences when acquiring different target images, which will affect the cropping accuracy. In addition, due to the influence of factors such as the storage memory of the image and the estimated total number of images to be taken, while ensuring that the image shooting task is completed as much as possible, the image clarity should also be guaranteed to be optimal to facilitate subsequent cropping and marking. Summary of the invention

[0004] The present invention aims to solve at least one of the technical problems existing in the prior art; to this end, the present invention proposes a drone adaptive control system for image cropping, which is used to solve the technical problem that the prior art does not consider the influence of factors such as altitude on pixels, resulting in insufficient cropping accuracy.

[0005] To achieve the above object, the present invention provides a UAV adaptive control system for image cropping, comprising:

[0006] A drone camera attitude control module, wherein the drone camera attitude control module controls the drone camera and the midpoint of the photovoltaic panel under test to be in a relatively vertical state;

[0007] The drone camera attitude control module also controls the drone flight altitude and coordinates according to the standard vertical relative altitude;

[0008] The standard vertical relative height is a theoretical value, which is determined according to the pixel size of the photovoltaic panel under test set in the system, the known parameters of the drone camera and the actual parameters of the photovoltaic panel under test;

[0009] An image cropping module, which crops the image according to the standard vertical relative height of the image taken by the drone camera, the real-time vertical height of the image taken, and the pixel size of the photovoltaic panel under test set in the system;

[0010] To ensure that the acquired image is a standard image, the camera should be in a relatively vertical state with the midpoint of the photographed plane rather than a horizontal vertical state. The relative vertical state is based on the photographed plane as the horizontal plane. Secondly, it is necessary to ensure that the picture taken by the camera meets the requirements of the system, so the flight altitude, that is, the vertical distance between the camera and the photographed plane, needs to be adjusted.

[0011] Further: the drone camera attitude control module includes the following steps:

[0012] S2-1: The drone acquires the image plane of the photovoltaic panel under test;

[0013] S2-2: extracting four vertices of the photovoltaic panel under test in the image plane according to the edge detection algorithm, wherein the four vertices are respectively defined as targets 1 to 4, and obtaining the corresponding Cartesian coordinates of targets 1 to 4 in the GPS module of the drone;

[0014] S2-3: According to the Cartesian coordinates (X N ,Y N ,Z N ) Obtain the midpoint coordinates and tilt angle of the photovoltaic panel under test in the image plane, N∈(1,4);

[0015] S2-4: Adjust the attitude of the drone according to the midpoint coordinates of the photovoltaic panel under test, the standard vertical relative height, the flight altitude of the drone and the tilt angle of the photovoltaic panel under test.

[0016] Further: The specific method of S2-2 is as follows:

[0017] Read the image and convert it into a grayscale image; use the Canny edge detection algorithm to detect the edge information of the image, that is, the approximate outline of the photovoltaic panel to be tested; further use OpenCV contour detection to find the rectangular outline of the photovoltaic panel to be tested in the approximate outline; select the pixel coordinates of the four vertices of the rectangular outline of the photovoltaic panel to be tested and record them as targets 1 to 4;

[0018] The pixel coordinates of targets 1 to 4 are converted into Cartesian coordinates (X N ,Y N ,Z N ), specifically, the pixel coordinates are obtained by inverting the focal length parameter f in the intrinsic parameter matrix of the drone camera;

[0019] The pixel coordinates can directly obtain the coordinate values ​​of the pixel points in the three-dimensional space. The pixel coordinates of the four vertices of the rectangular outline of the photovoltaic panel under test can be extracted by contour algorithm and then inverted to obtain the coordinate values ​​of the four vertices of the photovoltaic panel under test in the three-dimensional space.

[0020] Further: The specific method of S2-3 is as follows:

[0021] Determine the midpoint coordinates of the photovoltaic panel under test: According to the Cartesian coordinates (X N ,Y N ,Z N ) are determined by the average values ​​of the coordinate values ​​on the X-axis, Y-axis and Z-axis respectively;

[0022] Determine the tilt angle of the rectangular plane represented by the photovoltaic panel under test: randomly select the Cartesian coordinates (X N ,Y N ,Z N ) and any three coordinate values ​​of the midpoint coordinates of the photovoltaic panel under test are used to solve the normal vector of the rectangular plane, and then the inclination angle of the rectangular plane is calculated according to the angle between the normal vector and the X-axis-Y-axis horizontal plane of the Cartesian coordinate system of the GPS module of the drone, as follows:

[0023] Determine the normal vector of the rectangular plane from the Cartesian coordinates (X N ,Y N ,Z N ) and the three coordinate values ​​arbitrarily selected from the midpoint coordinates of the photovoltaic panel under test are recorded as P 1 , P 2 , P 3 ;

[0024] Create vector and The normal vector of the rectangular plane is the vector and The cross product of

[0025] Then, the tilt angle is calculated based on the normal vector of the rectangular plane and the normal vector of the X-axis-Y-axis horizontal plane of the Cartesian coordinate system of the GPS module of the drone, as shown in the following formula:

[0026]

[0027] Where n is the normal vector of the rectangular plane;

[0028] k is the X-axis-Y-axis horizontal plane normal vector of the Cartesian coordinate system of the drone’s built-in GPS module.

[0029] Further: The specific method of S2-4 is as follows:

[0030] Get the current flight altitude of the drone from the drone GPS module;

[0031] Adjust the flight altitude according to the standard vertical relative height and the tilt angle of the photovoltaic panel being tested, i.e. z f , specifically: 1.5Hsinθ+z M ≥z f ≥1.2Hsinθ+z M ; Where H is the standard vertical relative height; θ is the tilt angle of the photovoltaic panel under test; z M is the z coordinate of the midpoint coordinates of the photovoltaic panel under test;

[0032] After adjusting the flight altitude, z f The distance that the drone needs to translate is solved jointly with the midpoint coordinates of the photovoltaic panel under test in step 3, specifically:

[0033] Calculate the ratio of the x-coordinate, y-coordinate and z-coordinate of the midpoint of the photovoltaic panel under test in the Cartesian coordinate system;

[0034] According to the above ratio and the z coordinate of the drone in the Cartesian coordinate system f The coordinates directly obtain the estimated horizontal coordinates (x f ,y f ), which is the horizontal coordinate to which it should move at this height;

[0035] By comparing the current horizontal coordinate of the drone in the Cartesian coordinate system with the above estimated horizontal coordinate (x f ,y f ) position, the drone adjusts its own flight position;

[0036] The drone camera is further adjusted according to the tilt angle of the photovoltaic panel being tested, specifically: the drone camera adjusts its angle with the drone as the horizontal plane, ensuring that the angle between the drone camera and the horizontal plane represented by the drone is also the tilt angle of the photovoltaic panel being tested;

[0037] There are two purposes for obtaining the coordinates of the midpoint of the photovoltaic panel under test: 1. To determine the coordinates to which the drone should fly based on the coordinates of the midpoint of the photovoltaic panel under test, because the camera of the drone needs to be in a relatively vertical state with the photovoltaic panel under test; 2. To adjust the slope of the camera according to the slope of the photovoltaic panel to ensure that the shooting angle of the camera is relatively vertical to the plane of the photovoltaic panel under test;

[0038] The standard height is a theoretical value deduced from the final picture obtained under ideal conditions, the actual object being photographed and the various parameters of the camera. However, in actual shooting, the shooting is carried out according to the theoretical value, and the shooting angle of view is also the theoretical state, which only shoots the photovoltaic panel under test. If the drone has any deviation, it will not be able to capture the full picture of the photovoltaic panel under test. Therefore, it is necessary to perform an elevation operation under this ideal state to expand the camera's angle of view and shooting range, but it cannot be raised indefinitely, otherwise the pixels of the final cropped photovoltaic panel image under test will not meet the requirements. Therefore, the elevation height is in the range of 0.2 to 0.5 of the standard height.

[0039] Further: the steps for obtaining the standard vertical height are specifically as follows:

[0040] S6-1: Enter the pixel size I of the photovoltaic panel to be tested;

[0041] S6-2: Enter the drone camera parameters, including focal length F and camera sensor size D;

[0042] S6-3: Enter the actual size i of the photovoltaic panel under test;

[0043] S6-4: Directly calculate the standard vertical height: H = (F*I) / (D*i).

[0044] Furthermore: the specific steps of the image cropping module to crop an image are as follows:

[0045] S7-1: Use the direct distance measurement method to measure the relative position between the drone camera and the photovoltaic panel under test, which is the real-time vertical height of the captured image;

[0046] S7-2: cropping the image according to the ranging result, the standard vertical relative height and the real-time pixel size of the photovoltaic panel being measured. The cropping result is specifically: H / Hi×the real-time pixel size of the photovoltaic panel being measured, where H is the standard vertical height and Hi is the real-time vertical height of the captured image. The four sides of the captured image are evenly cropped accordingly.

[0047] Among them, the image pixels captured at the real-time vertical height should be cropped according to the pixel size I of the photovoltaic panel under test that needs to be obtained, that is, the image should be cropped according to the ratio of the standard vertical relative height and the actual shooting height to the vertical height ratio of the real-time shooting.

[0048] Furthermore: the direct ranging method includes two methods: one is to perform GPS ranging according to the coordinates of the drone after flight controlled by the drone camera attitude control module and the coordinates of the midpoint of the photovoltaic panel under test; the other is to perform image ranging according to the image of the photovoltaic panel under test taken by the drone camera and the drone camera parameters.

[0049] The beneficial effects of the present invention are: ensuring that the shooting angle of the drone camera is relatively perpendicular to the plane of the photovoltaic panel being measured, and controlling the drone shooting height to remain within a relative range, so that the difference in the images of different photovoltaic panels obtained is small, and finally the method of cropping the image according to the theoretical optimal pixel acquisition height and the actual shooting height is more convenient. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] In order to more clearly illustrate the embodiments of the present invention and the technical solutions in the prior art, the drawings required for use in the embodiments and the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some implementation methods recorded in the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0051] Figure 1 It is a system schematic diagram of the present invention;

[0052] Figure 2 It is a schematic diagram of shooting angle and cropping of the present invention. DETAILED DESCRIPTION

[0053] In order to make the purpose, technical scheme and advantages of the present invention more clearly understood, the present invention is described and illustrated below in conjunction with the accompanying drawings and embodiments. 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. Based on the embodiments provided by the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work belong to the scope of protection of the present invention.

[0054] Obviously, the drawings described below are only some examples or embodiments of the present invention. For ordinary technicians in this field, the present invention can be applied to other similar scenarios based on these drawings without any creative work.

[0055] In addition, it can be understood that although the efforts made in this development process may be complex and lengthy, for ordinary technicians in the field related to the contents disclosed in the present invention, some design, manufacturing or production changes based on the technical contents disclosed in the present invention are just conventional technical means and should not be understood as the contents disclosed in the present invention are insufficient.

[0056] If not otherwise specified, all embodiments and optional embodiments of the present invention can be combined with each other to form a new technical solution.

[0057] See also Figure 1 , the system flow chart of the present invention comprises:

[0058] The drone camera attitude control module controls the drone camera and the midpoint of the photovoltaic panel under test to be in a relatively vertical state;

[0059] The drone camera attitude control module also controls the drone flight altitude according to the standard vertical relative altitude;

[0060] Among them, the standard vertical relative height is a theoretical value, which is determined according to the pixel size of the photovoltaic panel under test set in the system, the known parameters of the drone camera and the actual parameters of the photovoltaic panel under test;

[0061] The image cropping module crops the image according to the standard vertical relative height of the image taken by the drone camera, the real-time vertical height of the image taken, and the pixel size of the photovoltaic panel under test set in the system.

[0062] Specifically:

[0063] To ensure that the shooting angle of the drone camera is relatively vertical to the plane of the photovoltaic panel being measured, the drone camera attitude control module processes the drone's flight attitude and camera in the following steps:

[0064] The flight attitude part is mainly adjusted according to the coordinate value of the midpoint coordinate of the photovoltaic panel under test in the GPS three-dimensional space. The first step is to directly obtain the image plane of the photovoltaic panel under test; read the image and convert it into a grayscale image; use the Canny edge detection algorithm to detect the edge information of the image, that is, the approximate outline of the photovoltaic panel under test; further use OpenCV contour detection to detect the rectangular outline of the photovoltaic panel under test in the approximate outline; screen the pixel coordinates of the four vertices of the rectangular outline of the photovoltaic panel under test, where the pixel coordinates of the four vertices are respectively defined as targets 1 to 4, and the pixel coordinates can be inverted according to the focal length parameter f in the intrinsic parameter matrix of the drone camera to obtain the corresponding Cartesian coordinates of targets 1 to 4 in the drone GPS module, that is, the spatial coordinates of the four vertex coordinates of the photovoltaic panel under test in the GPS space coordinate system with the earth as the earth, thereby solving the position of the center point of the photovoltaic panel under test in the GPS geodetic coordinate system, for which the drone only needs to adjust its position in the GPS geodetic coordinate system accordingly according to the GPS; due to the shooting angle of the drone, such as Figure 2 The plane of the photovoltaic panel being photographed should be relatively perpendicular to the plane of the photovoltaic panel being photographed. At this time, the plane of the photovoltaic panel being measured should be the horizontal plane. Therefore, in order to achieve that the shooting angle of the drone is relatively perpendicular to the plane of the photovoltaic panel being measured, in the GPS geodetic coordinate system, the position of the drone should be on the straight line from the origin of the geodetic coordinate system to the midpoint coordinate of the photovoltaic panel being measured; and if the drone camera wants to shoot the midpoint of the photovoltaic panel being measured, the angle of the camera also needs to be adjusted accordingly;

[0065] First point: adjust the angle of the drone camera relative to the horizontal plane of the drone to θ, where θ is the angle corresponding to the tilt angle of the photovoltaic panel under test relative to the horizontal plane of the ground. This ensures that when the photovoltaic panel under test is tilted, the observation angle of the camera is perpendicular to the horizontal plane of the photovoltaic panel under test under ideal conditions.

[0066] The method for obtaining this angle θ is as follows: randomly select the Cartesian coordinates (X N ,Y N ,Z N ) and any three coordinate values ​​of the midpoint coordinates of the photovoltaic panel under test are used to solve the normal vector of the rectangular plane, and then the inclination angle of the rectangular plane is calculated according to the angle between the normal vector and the X-axis-Y-axis horizontal plane of the Cartesian coordinate system of the GPS module of the drone, as follows:

[0067] Determine the normal vector of the rectangular plane from the Cartesian coordinates (X N ,Y N ,Z N ) and the three coordinate values ​​arbitrarily selected from the midpoint coordinates of the photovoltaic panel under test are recorded as P 1 , P 2 , P 3 ;

[0068] Create vector and The normal vector of the rectangular plane is the vector and The cross product of

[0069] Then, the tilt angle is calculated based on the normal vector of the rectangular plane and the normal vector of the X-axis-Y-axis horizontal plane of the Cartesian coordinate system of the GPS module of the drone, as shown in the following formula:

[0070]

[0071] Where n is the normal vector of the rectangular plane;

[0072] k is the X-axis-Y-axis horizontal plane normal vector of the Cartesian coordinate system of the GPS module of the drone;

[0073] The second point: ensure that the drone can move to a position where the camera can correctly observe the photovoltaic panel under test, that is, the position of the drone should also be on the straight line from the origin of the GPS geodetic coordinate system to the midpoint coordinates of the photovoltaic panel under test. However, due to the requirements for image pixels, the flight altitude of the drone (i.e., shooting distance) needs to be limited to a certain extent. It is determined according to the derived theoretical value standard vertical height H. Its flight altitude should meet the shooting distance requirements and the horizontal coordinates adjusted according to the flight altitude should meet the shooting distance requirements. In order to ensure that the full view of the photovoltaic panel under test is captured, it is necessary to perform an elevation operation under this ideal state to expand the camera's viewing angle and shooting range, but it cannot be raised infinitely, otherwise the pixels of the final cropped photovoltaic panel image under test will not meet the requirements. For this reason, the elevation height is in the range of 0.2 to 0.5 of the standard height, that is, the flight altitude should be 1.5Hsinθ+z M ≥z f ≥1.2Hsinθ+z M , where θ is the tilt angle of the photovoltaic panel under test, because the midpoint coordinates of the drone and the photovoltaic panel under test should be on the same oblique line in the GPS geodetic coordinate system; z M is the z coordinate of the midpoint of the photovoltaic panel under test in the GPS geodetic coordinate system, f It should be a height higher than the altitude of the midpoint of the photovoltaic panel being tested. f Yes (x f ,y f ) is solved (the solution here can directly refer to the ratio between the coordinates of the midpoint of the photovoltaic panel under test), that is, the position where the drone should move to in the GPS geodetic coordinate system.

[0074] Further:

[0075] The specific steps for obtaining the standard vertical height H are as follows: input the pixel size I of the photovoltaic panel to be tested; input the parameters of the drone camera, including focal length F and camera sensor size D; input the actual size i of the photovoltaic panel to be tested; directly obtain the standard vertical height as: H = (F*I) / (D*i), where the pixel size of the photovoltaic panel to be tested is set according to the drone memory and mission requirements (number of shots), which is set by the user.

[0076] The image cropping module can crop images in a simple manner according to the influence of the size of the field of view of the photo (i.e., the ratio of the flight altitude) on the pixels. Specifically, the relative position of the drone camera and the photovoltaic panel under test is measured by the direct ranging method, which is the real-time vertical height of the captured image; the image is cropped according to the ranging result, the standard vertical relative height and the real-time pixel size of the photovoltaic panel under test. The cropping result is specifically: H / Hi×the real-time pixel size of the photovoltaic panel under test, where H is the standard vertical height and Hi is the real-time vertical height of the captured image. The four sides of the captured image are evenly cropped accordingly; the direct ranging method includes two methods: one is to perform GPS ranging according to the coordinates of the drone after the drone is controlled by the drone camera attitude control module and the coordinates of the midpoint of the photovoltaic panel under test; the other is to perform image ranging according to the image of the photovoltaic panel under test taken by the drone camera and the drone camera parameters.

[0077] Therefore, the present invention performs the following operations to ensure accurate cropping of the image:

[0078] The camera posture is adjusted according to the posture of the photovoltaic panel under test, and the flight altitude of the drone is adjusted according to the shooting pixel requirements, and then the position of the drone observing the photovoltaic panel is adjusted to ensure that the shooting angle of the drone is perpendicular to the photovoltaic panel under test.

[0079] The above is a very detailed description of one or more embodiments of the present invention, but the content is only a specific example of the present invention and cannot be considered to limit the scope of application of the present invention. Any other methods and changes proposed based on the content of the present invention should fall within the scope of patent protection of the present invention.

Claims

1. An adaptive control system for unmanned aerial vehicles for image cropping, characterized in that: include: A drone camera attitude control module, wherein the drone camera attitude control module controls the drone camera and the midpoint of the photovoltaic panel under test to be in a relatively vertical state; The drone camera attitude control module also controls the drone flight altitude and coordinates according to the standard vertical relative altitude; The standard vertical relative height is a theoretical value, which is determined according to the pixel size of the photovoltaic panel under test set in the system, the known parameters of the drone camera and the actual parameters of the photovoltaic panel under test; The image cropping module crops the image according to the standard vertical relative height of the image taken by the drone camera, the real-time vertical height of the image taken, and the pixel size of the photovoltaic panel under test set in the system.

2. The UAV adaptive control system for image cropping according to claim 1, characterized in that: The drone camera attitude control module includes the following steps: S2-1: The drone acquires the image plane of the photovoltaic panel under test; S2-2: extracting four vertices of the photovoltaic panel under test in the image plane according to the edge detection algorithm, wherein the four vertices are respectively defined as targets 1 to 4, and obtaining the corresponding Cartesian coordinates of targets 1 to 4 in the GPS module of the drone; S2-3: According to the Cartesian coordinates (X N ,Y N ,Z N ) Obtain the midpoint coordinates and tilt angle of the photovoltaic panel under test in the image plane, N∈(1,4); S2-4: Adjust the attitude of the drone according to the midpoint coordinates of the photovoltaic panel under test, the standard vertical relative height, the flight altitude of the drone and the tilt angle of the photovoltaic panel under test.

3. The UAV adaptive control system for image cropping according to claim 2, characterized in that: The specific method of S2-2 is as follows: Read the image and convert it into a grayscale image; use the Canny edge detection algorithm to detect the edge information of the image, that is, the approximate outline of the photovoltaic panel to be tested; further use OpenCV contour detection to find the rectangular outline of the photovoltaic panel to be tested in the approximate outline; select the pixel coordinates of the four vertices of the rectangular outline of the photovoltaic panel to be tested and record them as targets 1 to 4; The pixel coordinates of targets 1 to 4 are converted into Cartesian coordinates (X N ,Y N ,Z N ), specifically, the pixel coordinates are inverted by the focal length parameter f in the intrinsic parameter matrix of the drone camera.

4. The UAV adaptive control system for image cropping according to claim 2, characterized in that: The specific method of S2-3 is as follows: Determine the midpoint coordinates of the photovoltaic panel under test: According to the Cartesian coordinates (X N ,Y N ,Z N ) are determined by the average values ​​of the coordinate values ​​on the X-axis, Y-axis and Z-axis respectively; Determine the tilt angle of the rectangular plane represented by the photovoltaic panel under test: randomly select the Cartesian coordinates (X N ,Y N ,Z N ) and any three coordinate values ​​of the midpoint coordinates of the photovoltaic panel under test are used to solve the normal vector of the rectangular plane, and then the inclination angle of the rectangular plane is calculated according to the angle between the normal vector and the X-axis-Y-axis horizontal plane of the Cartesian coordinate system of the GPS module of the drone, as follows: Determine the normal vector of the rectangular plane from the Cartesian coordinates (X N ,Y N ,Z N ) and the coordinates of the midpoint of the photovoltaic panel under test are arbitrarily selected as P1, P2, and P3 respectively; Create vector and The normal vector of the rectangular plane is the vector and The cross product of Then, the tilt angle is calculated based on the normal vector of the rectangular plane and the normal vector of the X-axis-Y-axis horizontal plane of the Cartesian coordinate system of the GPS module of the drone, as shown in the following formula: Where n is the normal vector of the rectangular plane; k is the X-axis-Y-axis horizontal plane normal vector of the Cartesian coordinate system of the drone’s built-in GPS module.

5. The UAV adaptive control system for image cropping according to claim 2, characterized in that: The specific method of S2-4 is as follows: Get the current flight altitude of the drone from the drone GPS module; Adjust the flight altitude according to the standard vertical relative height and the tilt angle of the photovoltaic panel being tested, i.e. Z f , specifically: 1.5Hsinθ+z M ≥z f ≥1.2Hsinθ+z M ; Where H is the standard vertical relative height; θ is the tilt angle of the photovoltaic panel under test; Z M is the z coordinate of the midpoint coordinates of the photovoltaic panel under test; After adjusting the flight altitude, Z f The distance that the drone needs to translate is solved jointly with the midpoint coordinates of the photovoltaic panel under test in step 3, specifically: Calculate the ratio of the x-coordinate, y-coordinate and z-coordinate of the midpoint of the photovoltaic panel under test in the Cartesian coordinate system; According to the above ratio and the Z of the drone in the Cartesian coordinate system f The coordinates directly obtain the estimated horizontal coordinates (x f ,y f ), which is the horizontal coordinate to which it should move at this height; By comparing the current horizontal coordinate of the drone in the Cartesian coordinate system with the above estimated horizontal coordinate (x f ,y f ) position, the drone adjusts its own flight position; The drone camera is further adjusted according to the tilt angle of the photovoltaic panel being measured. Specifically, the drone camera adjusts its angle with the drone as the horizontal plane to ensure that the angle between the drone camera and the horizontal plane represented by the drone is also the tilt angle of the photovoltaic panel being measured.

6. The UAV adaptive control system for image cropping according to claim 1, characterized in that: The steps for obtaining the standard vertical height are as follows: S6-1: Enter the pixel size I of the photovoltaic panel to be tested; S6-2: Enter the drone camera parameters, including focal length F and camera sensor size D; S6-3: Enter the actual size i of the photovoltaic panel under test; S6-4: Directly calculate the standard vertical height: H = (F*I) / (D*i).

7. The UAV adaptive control system for image cropping according to claim 1, characterized in that: The specific steps of the image cropping module to crop an image are as follows: S7-1: Use the direct distance measurement method to measure the relative position between the drone camera and the photovoltaic panel under test, which is the real-time vertical height of the captured image; S7-2: Crop the image according to the ranging result, the standard vertical relative height and the real-time pixel size of the photovoltaic panel being measured. The cropping result is: H / Hi×the real-time pixel size of the photovoltaic panel being measured, where H is the standard vertical height and Hi is the real-time vertical height of the captured image. The four sides of the captured image are evenly cropped accordingly.

8. The UAV adaptive control system for image cropping according to claim 7, characterized in that: The direct distance measurement method includes two methods: one is to perform GPS distance measurement based on the coordinates of the drone after flight and the coordinates of the midpoint of the photovoltaic panel under test controlled by the drone camera attitude control module; the other is to perform image distance measurement based on the image of the photovoltaic panel under test taken by the drone camera and the drone camera parameters.

Images