Machine vision-based automatic identification and statistics system and method for houses around a fan

Abstract

The application discloses a kind of automatic identification statistics method and system of house around fan based on machine vision, comprising: the camera of unmanned aerial vehicle is calibrated, and the internal parameter matrix M1 of the camera of unmanned aerial vehicle and focal length F are obtained;The camera of unmanned aerial vehicle is photographed to fan, and the image of the top of fan nacelle is obtained, the image coordinates of the top feature point of fan nacelle are obtained based on the image of the top of fan nacelle, the actual coordinates of each top feature point are obtained, the camera external parameter matrix M2 is obtained based on the image coordinates of the top feature point of fan nacelle, the actual coordinates of each top feature point and internal parameter matrix;The distance H1 from camera coordinate system plane to the top of fan nacelle is extracted based on camera external parameter matrix M2.The application can realize the automatic identification statistics method and system of house around fan based on machine vision.

Description

Technical Field

[0001] This invention relates to the field of automatic building identification, and in particular to an automatic identification and statistical system and method for buildings around wind turbines based on machine vision. Background Technology

[0002] With the booming development of the wind power industry, wind power generation has become one of the main driving forces in the field of clean energy. However, accompanying its development are increasingly stringent regulations imposed by governments around the world regarding the potential impact of wind turbines on surrounding residents and the environment. To ensure the sustainable development of the wind power industry and protect public safety and environmental health, many countries have stipulated that no residential buildings may be built within the radius of wind turbines. Common regulations require that wind turbines be at least four times their total height away from residential areas to prevent noise pollution, obstructed views, and potential safety hazards.

[0003] However, when constructing wind farms in hilly areas, determining the radius of wind turbines and accurately counting the total number of houses becomes extremely difficult due to the scattered surrounding residents and dense vegetation. Traditional manual measurement methods are not only difficult and time-consuming, but satellite map measurements suffer from information lag and insufficient accuracy, as new houses cannot be reflected in a timely manner, and manual verification of the number of houses is still required. To solve this problem, this invention proposes an automatic identification and counting method for houses around wind turbines based on machine vision. By combining drone photography with machine vision technology, it achieves automatic identification and counting of houses around wind turbines, providing an efficient and accurate solution for wind farm construction. Summary of the Invention

[0004] The purpose of this invention is to provide an automatic identification and statistical system and method for buildings around wind turbines based on machine vision, aiming to solve the problem of automatic identification and statistical system and method for buildings around wind turbines based on machine vision.

[0005] This invention provides a machine vision-based method for automatic identification and statistical analysis of buildings surrounding wind turbines, comprising:

[0006] S1. Calibrate the drone camera to obtain the drone camera's intrinsic parameter matrix M1 and focal length F;

[0007] S2. The drone camera takes pictures of the wind turbine and obtains an image of the top of the wind turbine nacelle. Based on the image of the top of the wind turbine nacelle, the image coordinates of the feature points on the top of the wind turbine nacelle are obtained. The actual coordinates of each feature point on the top are obtained. Based on the image coordinates of the feature points on the top of the wind turbine nacelle, the actual coordinates of each feature point on the top, and the intrinsic parameter matrix, the extrinsic parameter matrix M2 of the camera is obtained.

[0008] S3. Extract the distance H1 from the camera coordinate system plane to the top of the wind turbine nacelle based on the camera extrinsic parameter matrix M2. Calculate the distance H between the camera coordinate system plane and the ground based on the distance H1 from the camera coordinate system plane to the top of the wind turbine nacelle and the height from the ground to the top of the nacelle. Calculate the pixel size r of the radius R of the area where no buildings are allowed based on H and the focal length F.

[0009] S4. Adjust the drone altitude to capture images within the complete fixed radius of the camera position;

[0010] S5. Obtain the image coordinates of the center point of the top of the tower, and draw a circle with a pixel size of r in the image with the center point of the top of the tower as the center.

[0011] S6. Detect the number of houses within a circle of size r in S5 based on a trained artificial intelligence model.

[0012] The present invention also provides an automatic identification and statistical system for buildings around wind turbines based on machine vision, comprising:

[0013] Calibration module: Used to calibrate the drone camera and obtain the drone camera's intrinsic parameter matrix M1 and focal length F;

[0014] The extrinsic parameter matrix calculation module is used by the drone camera to capture images of the wind turbine, obtain images of the top of the wind turbine nacelle, obtain image coordinates of feature points on the top of the wind turbine nacelle based on the images, obtain the actual coordinates of each feature point on the top, and obtain the camera extrinsic parameter matrix M2 based on the image coordinates of the feature points on the top of the wind turbine nacelle, the actual coordinates of each feature point on the top, and the intrinsic parameter matrix.

[0015] Pixel size calculation module: used to extract the distance H1 from the camera coordinate system plane to the top of the wind turbine nacelle based on the camera extrinsic parameter matrix M2, calculate the distance H between the camera coordinate system plane and the ground based on the distance H1 from the camera coordinate system plane to the top of the wind turbine nacelle and the height from the ground to the top of the nacelle, and calculate the pixel size r of the radius R of the area where no houses are allowed based on H and the focal length F.

[0016] Shooting module: Used to adjust the drone's altitude and capture images within a fixed radius of the entire camera position;

[0017] Circle drawing module: Used to obtain the image coordinates of the center point of the top of the tower, and draw a circle with a pixel size of r in the image with the center point of the top of the tower as the center;

[0018] Detection module: Used to detect the number of houses within a circle of size r based on a trained artificial intelligence model.

[0019] Using the embodiments of the present invention, automatic identification and statistics of houses around wind turbines can be achieved through machine vision.

[0020] The above description is merely an overview of the technical solution of the present invention. In order to better understand the technical means of the present invention and to implement it in accordance with the contents of the specification, and in order to make the above and other objects, features and advantages of the present invention more apparent and understandable, specific embodiments of the present invention are described below. Attached Figure Description

[0021] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0022] Figure 1 This is a flowchart of a machine vision-based automatic identification and statistical method for buildings around a wind turbine, according to an embodiment of the present invention.

[0023] Figure 2 This is a schematic diagram of a machine vision-based automatic identification and statistics system for buildings around a wind turbine, according to an embodiment of the present invention.

[0024] Figure 3 This is a schematic diagram of a machine vision-based automatic identification and statistics device for buildings around a wind turbine, according to an embodiment of the present invention. Detailed Implementation

[0025] The technical solution of the present invention will be clearly and completely described below with reference to the embodiments. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0026] Method Implementation Examples

[0027] According to embodiments of the present invention, a machine vision-based method for automatic identification and statistical analysis of buildings surrounding wind turbines is provided. Figure 1 This is a flowchart of a machine vision-based automatic identification and statistical method for buildings around wind turbines, as described in an embodiment of the present invention. Figure 1 As shown, it specifically includes:

[0028] S1. Calibrate the drone camera to obtain the drone camera's intrinsic parameter matrix M1 and focal length F;

[0029] S1 specifically includes: the camera calibration adopts the Zhang Zhengyou calibration method, which involves acquiring multiple sets of checkerboard images from different directions to calibrate the camera and obtain the camera's intrinsic parameter matrix M1 and focal length F.

[0030] S2. The drone camera takes pictures of the wind turbine and obtains an image of the top of the wind turbine nacelle. Based on the image of the top of the wind turbine nacelle, the image coordinates of the feature points on the top of the wind turbine nacelle are obtained. The actual coordinates of each feature point on the top are obtained. Based on the image coordinates of the feature points on the top of the wind turbine nacelle, the actual coordinates of each feature point on the top, and the intrinsic parameter matrix, the extrinsic parameter matrix M2 of the camera is obtained.

[0031] S2 specifically includes: operating the drone to take off to a specific altitude, acquiring images of the top of the wind turbine nacelle, obtaining the image coordinates of the feature points on the top of the nacelle, establishing a coordinate system on the top of the nacelle according to the nacelle design drawings and obtaining the actual coordinates of each feature point on the top, and obtaining the camera extrinsic parameter matrix M2 based on the intrinsic parameter matrix by mapping the relationship between the image coordinates and the actual coordinates of the feature points on the top of the nacelle. M2 represents the pose relationship between the camera and the top of the nacelle.

[0032] S3. Extract the distance H1 from the camera coordinate system plane to the top of the wind turbine nacelle based on the camera extrinsic parameter matrix M2. Calculate the distance H between the camera coordinate system plane and the ground based on the distance H1 from the camera coordinate system plane to the top of the wind turbine nacelle and the height from the ground to the top of the nacelle. Calculate the pixel size r of the radius R of the area where no buildings are allowed based on H and the focal length F.

[0033] S3 specifically includes: extracting the distance H1 from the camera coordinate system plane to the nacelle surface from the pose relationship matrix between the camera and the top of the nacelle; calculating the distance H between the camera coordinate system plane and the ground based on the distance H1 from the camera coordinate system plane to the top of the wind turbine nacelle and the height from the ground to the top of the nacelle; and calculating the pixel size r of the radius R of the area where no buildings are allowed based on H and the focal length F. The calculation formula is as follows:

[0034] r = H × F / R.

[0035] S4. Adjust the drone altitude to capture images within the complete fixed radius of the camera position;

[0036] S5. Obtain the image coordinates of the center point of the top of the tower, and draw a circle with a pixel size of r in the image with the center point of the top of the tower as the center.

[0037] S6. Detect the number of houses within a circle of size r in S5 based on a trained artificial intelligence model.

[0038] S6 specifically includes: a pre-trained YOLO model, which is used to detect and count the number of houses within the circle drawn in S5.

[0039] Step 1: Perform drone camera calibration to obtain the drone camera's intrinsic parameter matrix M1 and focal length F.

[0040] The camera calibration adopts the Zhang Zhengyou calibration method, which involves acquiring multiple sets of chessboard-like images from different directions to calibrate the camera and obtain the camera's intrinsic parameter matrix M1 and focal length F.

[0041] The calibration algorithm is as follows:

[0042] To simplify the calculation process, the plane containing the two-dimensional target is set as the XOY plane of the world coordinate system. As can be seen from equation (2-17), the mapping relationship from a point on the target plane to the camera image can be described by a homography matrix H.

[0043] H=[h1 h2 h3]=sM1[r1 r2 t] (2-18)

[0044] Where h1, h2, and h3 represent the first, second, and third columns of the homography matrix H, respectively; M1 represents the camera intrinsic parameters; and r1, r2, and t represent the first two columns of the rotation matrix and the translation vector in the camera extrinsic parameters, respectively.

[0045] Decomposing equation (2-18), we get:

[0046]

[0047] Where λ = 1 / s.

[0048] Since the rotation matrices are all orthonormal matrices, r1⊥r2, and ||r1||=||r2||. Combining equation (2-19), we can obtain:

[0049]

[0050]

[0051] make Expand

[0052]

[0053] In fact, matrix B has a closed-form solution:

[0054]

[0055] The left side of both equations (2-20) and (2-21) can be written as Given the form, it is easy to see that B is a symmetric matrix. Rearranging the elements of matrix B yields a new vector b. Both of the above equations can be written as:

[0056]

[0057] From the two constraints of equations (2-20) and (2-21), we can obtain the following equation:

[0058]

[0059] If we have K target images, we can obtain K equations as shown in equation (2-25). Stacking these equations, we have:

[0060] Vb = 0 (2-26)

[0061] Where V is a 2K×6 matrix, and the equation has a solution when K≥2.

[0062] b = [B 11 B 12 B 22 B 13 B 23 B 33 ] T

[0063] Combining equation (2-23), the camera's intrinsic parameters can be obtained:

[0064]

[0065] in,

[0066] After obtaining the camera's intrinsic parameters, substitute them into equation (2-18) to calculate the camera's extrinsic parameters:

[0067]

[0068] The proportionality coefficient λ here is determined by the orthogonality condition.

[0069] Step Two: Operate the drone to a specific altitude to photograph the wind turbine. During this process, images of the top of the wind turbine nacelle are acquired, obtaining the image coordinates of feature points on the nacelle top. A coordinate system is established on the top of the nacelle according to the nacelle design drawings, and the actual coordinates of each feature point on the top are obtained. By mapping the relationship between the image coordinates and actual coordinates of the feature points on the top of the nacelle, the camera extrinsic parameter matrix M2 is obtained. This matrix represents the pose relationship between the camera and the top of the nacelle. By adjusting the pose of the drone camera, the camera lens is ensured to be parallel to the surface of the top of the nacelle, ensuring the accuracy of the measured radius range of the wind turbine.

[0070] The criterion for determining whether the camera lens is parallel to the top surface of the cabin is the camera extrinsic parameter matrix. If the rotational part of the camera extrinsic parameter matrix M2... Then r 11 and r 22 (The rotated portions of the X and Y axes) are 0, while r 21 and r 12 The intersection of the X and Y axes is 1 or -1.

[0071] The conversion relationship between M1 and M2:

[0072] The process of determining the camera's intrinsic and extrinsic parameters by mapping the two-dimensional image coordinates of the target to the three-dimensional world coordinates is called camera calibration. This involves using the camera to map a point in space (X...)... W ,Y W Z W Projecting onto the image plane (u,v) requires transformations across four coordinate systems: world coordinate system, camera coordinate system, image physical coordinate system, and image pixel coordinate system. By establishing the transformation relationships between these coordinate systems, the image coordinate system UOV and the world coordinate system O are determined. w X w Y w Z w The mapping relationship between them

[0073]

[0074] M1 is the camera intrinsic parameter matrix, where f x f y S represents the number of pixels per unit millimeter along the x-axis and y-axis in the image pixel coordinate system, respectively. x S y These represent the x and y coordinates of the origin of the image's physical coordinate system in the image's pixel coordinate system, respectively. The camera intrinsic parameter matrix is ​​an inherent property of the camera and does not change with the camera's position. M2 is the camera extrinsic parameter matrix, describing the relative pose between the world coordinate system and the camera coordinate system; obviously, it changes with the camera's position.

[0075] Step 3: Extract the distance H1 from the camera coordinate system plane to the nacelle surface from the pose relationship matrix between the camera and the top of the nacelle obtained in Step 2. Calculate the distance H between the camera coordinate system plane and the ground, taking into account the dimensions of the wind turbine. Then, calculate the pixel size r of the radius R using the formula r = H × F / R. To ensure the sustainable development of the wind power industry and protect public safety and environmental health, many countries have stipulated that residential buildings cannot be built within the radius of wind turbines. Common regulations require that wind turbines be at least four times their total height away from residential areas to prevent noise pollution, view obstruction, and potential safety hazards.

[0076] R represents the area where housing is not permitted.

[0077] Step 4: Adjust the drone's altitude and capture images within the entire fixed radius of the camera position. The definition of a fixed radius varies depending on the laws of different countries, but is the same within the same country.

[0078] Step 5: Combining the wind turbine design drawings and the images obtained from the drone camera, find the image coordinates of the center point of the top of the tower. Draw a circle with a pixel size of r in the image, with the center point of the top of the tower as the center.

[0079] Step 6: Use a pre-trained YOLO model to detect and count the number of houses within the circle drawn in Step 5.

[0080] The beneficial effects are as follows: This invention proposes an automatic identification and statistical method for buildings around wind turbines based on machine vision. By combining drone photography with machine vision technology, it realizes the automatic identification and statistical analysis of buildings around wind turbines, providing an efficient and accurate solution for wind farm construction.

[0081] System Implementation Examples

[0082] According to embodiments of the present invention, an automatic identification and statistical system for buildings around wind turbines based on machine vision is provided. Figure 2 This is a schematic diagram of an automatic identification and statistics system for buildings around a wind turbine based on machine vision, as described in an embodiment of the present invention. Figure 2 As shown, it specifically includes:

[0083] Calibration module: Used to calibrate the drone camera and obtain the drone camera's intrinsic parameter matrix M1 and focal length F;

[0084] The extrinsic parameter matrix calculation module is used by the drone camera to capture images of the wind turbine, obtain images of the top of the wind turbine nacelle, obtain image coordinates of feature points on the top of the wind turbine nacelle based on the images, obtain the actual coordinates of each feature point on the top, and obtain the camera extrinsic parameter matrix M2 based on the image coordinates of the feature points on the top of the wind turbine nacelle, the actual coordinates of each feature point on the top, and the intrinsic parameter matrix.

[0085] Pixel size calculation module: used to extract the distance H1 from the camera coordinate system plane to the top of the wind turbine nacelle based on the camera extrinsic parameter matrix M2, calculate the distance H between the camera coordinate system plane and the ground based on the distance H1 from the camera coordinate system plane to the top of the wind turbine nacelle and the height from the ground to the top of the nacelle, and calculate the pixel size r of the radius R of the area where no houses are allowed based on H and the focal length F.

[0086] Shooting module: Used to adjust the drone's altitude and capture images within a fixed radius of the entire camera position;

[0087] Circle drawing module: Used to obtain the image coordinates of the center point of the top of the tower, and draw a circle with a pixel size of r in the image with the center point of the top of the tower as the center;

[0088] Detection module: Used to detect the number of houses within a circle of size r based on a trained artificial intelligence model.

[0089] The calibration module is specifically used for: camera calibration adopts Zhang Zhengyou calibration method, which involves acquiring multiple sets of checkerboard images from different directions by the camera to calibrate the camera and obtain the camera's intrinsic parameter matrix M1 and focal length F.

[0090] The extrinsic parameter matrix calculation module is specifically used for: operating the UAV to take off to a specific altitude, acquiring images of the top of the wind turbine nacelle, obtaining the image coordinates of the feature points on the top of the nacelle, establishing a coordinate system on the top of the nacelle according to the nacelle design drawings and obtaining the actual coordinates of each feature point on the top, and obtaining the camera extrinsic parameter matrix M2 based on the intrinsic parameter matrix by mapping the relationship between the image coordinates and the actual coordinates of the feature points on the top of the nacelle. M2 represents the pose relationship between the camera and the top of the nacelle.

[0091] The pixel size calculation module is specifically used to: extract the distance H1 from the camera coordinate system plane to the nacelle surface from the pose relationship matrix between the camera and the top of the nacelle; calculate the distance H between the camera coordinate system plane and the ground based on the distance H1 from the camera coordinate system plane to the top of the wind turbine nacelle and the height from the ground to the top of the nacelle; and calculate the pixel size r of the radius R of the area where no buildings are allowed based on H and the focal length F. The calculation formula is as follows:

[0092] r = H × F / R.

[0093] The detection module is specifically used to: pre-train a YOLO model, and detect and count the number of houses within a circle of size r based on the trained YOLO model.

[0094] The embodiments of the present invention are system embodiments corresponding to the above method embodiments. The specific operation of each module can be understood by referring to the description of the method embodiments, and will not be repeated here.

[0095] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features; and these modifications or substitutions to the technical solutions of the embodiments of the present invention do not cause the essence of the corresponding technical solutions to deviate from the scope of the present solution.

Claims

1. A machine vision-based method for automatic identification and statistical analysis of buildings surrounding wind turbines, characterized in that, include: S1. Calibrate the drone camera to obtain the drone camera's intrinsic parameter matrix M1 and focal length F; S2. The drone camera takes pictures of the wind turbine and obtains an image of the top of the wind turbine nacelle. Based on the image of the top of the wind turbine nacelle, the image coordinates of the feature points on the top of the wind turbine nacelle are obtained. The actual coordinates of each feature point on the top are obtained. Based on the image coordinates of the feature points on the top of the wind turbine nacelle, the actual coordinates of each feature point on the top, and the intrinsic parameter matrix, the extrinsic parameter matrix M2 of the camera is obtained. S3. Extract the distance H1 from the camera coordinate system plane to the top of the wind turbine nacelle based on the camera extrinsic parameter matrix M2. Calculate the distance H between the camera coordinate system plane and the ground based on the distance H1 from the camera coordinate system plane to the top of the wind turbine nacelle and the height from the ground to the top of the nacelle. Calculate the pixel size r of the radius R of the area where no buildings are allowed based on H and the focal length F. S4. Adjust the drone altitude to capture images within the complete fixed radius of the camera position; S5. Obtain the image coordinates of the center point of the top of the tower, and draw a circle with a pixel size of r in the image with the center point of the top of the tower as the center. S6. Detect the number of houses within a circle of size r in S5 based on a trained artificial intelligence model.

2. The method according to claim 1, characterized in that, S1 specifically includes: the camera calibration adopts the Zhang Zhengyou calibration method, which involves the camera acquiring multiple sets of chessboard images from different directions to calibrate the camera and obtain the camera's intrinsic parameter matrix M1 and focal length F.

3. The method according to claim 2, characterized in that, S2 specifically includes: operating the UAV to take off to a specific altitude, acquiring images of the top of the wind turbine nacelle, obtaining the image coordinates of feature points on the top of the nacelle, establishing a coordinate system on the top of the nacelle according to the nacelle design drawings and obtaining the actual coordinates of each feature point on the top, and obtaining the camera extrinsic parameter matrix M2 based on the intrinsic parameter matrix by mapping the relationship between the image coordinates and actual coordinates of the feature points on the top of the nacelle. M2 represents the pose relationship between the camera and the top of the nacelle.

4. The method according to claim 3, characterized in that, S3 specifically includes: extracting the distance H1 from the camera coordinate system plane to the nacelle surface from the pose relationship matrix between the camera and the top of the nacelle; calculating the distance H between the camera coordinate system plane and the ground based on the distance H1 from the camera coordinate system plane to the top of the wind turbine nacelle and the height from the ground to the top of the nacelle; and calculating the pixel size r of the radius R of the area where no buildings are allowed based on H and the focal length F. The calculation formula is as follows: r = H × F / R.

5. The method according to claim 4, characterized in that, S6 specifically includes: a pre-trained YOLO model, which is used to detect and count the number of houses within the circle drawn in S5.

6. A machine vision-based automatic identification and statistical system for buildings around a wind turbine, characterized in that, include: Calibration module: Used to calibrate the drone camera and obtain the drone camera's intrinsic parameter matrix M1 and focal length F; The extrinsic parameter matrix calculation module is used by the drone camera to capture images of the wind turbine, obtain images of the top of the wind turbine nacelle, obtain image coordinates of feature points on the top of the wind turbine nacelle based on the images, obtain the actual coordinates of each feature point on the top, and obtain the camera extrinsic parameter matrix M2 based on the image coordinates of the feature points on the top of the wind turbine nacelle, the actual coordinates of each feature point on the top, and the intrinsic parameter matrix. Pixel size calculation module: used to extract the distance H1 from the camera coordinate system plane to the top of the wind turbine nacelle based on the camera extrinsic parameter matrix M2, calculate the distance H between the camera coordinate system plane and the ground based on the distance H1 from the camera coordinate system plane to the top of the wind turbine nacelle and the height from the ground to the top of the nacelle, and calculate the pixel size r of the radius R of the area where no houses are allowed based on H and the focal length F. Shooting module: Used to adjust the drone's altitude and capture images within a fixed radius of the entire camera position; Circle drawing module: Used to obtain the image coordinates of the center point of the top of the tower, and draw a circle with a pixel size of r in the image with the center point of the top of the tower as the center; Detection module: Used to detect the number of houses within a circle of size r based on a trained artificial intelligence model.

7. The system according to claim 6, characterized in that, The calibration module is specifically used for: camera calibration using the Zhang Zhengyou calibration method, where the camera acquires multiple sets of checkerboard images from different directions to calibrate the camera and obtain the camera's intrinsic parameter matrix M1 and focal length F.

8. The system according to claim 7, characterized in that, The extrinsic parameter matrix calculation module is specifically used for: operating the UAV to take off to a specific altitude, acquiring images of the top of the wind turbine nacelle, obtaining the image coordinates of the feature points on the top of the nacelle, establishing a coordinate system on the top of the nacelle according to the nacelle design drawings and obtaining the actual coordinates of each feature point on the top, and obtaining the camera extrinsic parameter matrix M2 based on the intrinsic parameter matrix by mapping the relationship between the image coordinates and the actual coordinates of the feature points on the top of the nacelle. M2 represents the pose relationship between the camera and the top of the nacelle.

9. The system according to claim 8, characterized in that, The pixel size calculation module is specifically used to: extract the distance H1 from the camera coordinate system plane to the nacelle surface from the pose relationship matrix between the camera and the top of the nacelle; calculate the distance H between the camera coordinate system plane and the ground based on the distance H1 from the camera coordinate system plane to the top of the wind turbine nacelle and the height from the ground to the top of the nacelle; and calculate the pixel size r of the radius R of the area where no buildings are allowed based on H and the focal length F. The calculation formula is as follows: r = H × F / R.

10. The system according to claim 9, characterized in that, The detection module is specifically used to: pre-train a YOLO model, and detect and count the number of houses within a circle of size r based on the trained YOLO model.

Images