Method for collecting vertical face reference object visual positioning coordinates based on unmanned aerial vehicle side-mounted camera

Abstract

The invention relates to the technical field of unmanned aerial vehicle positioning, and discloses a method for collecting vertical face reference object visual positioning coordinates based on an unmanned aerial vehicle side-mounted camera, and the method comprises the steps: obtaining the coordinates of an unmanned aerial vehicle based on a satellite positioning technology, recording the coordinates as first fuzzy positioning coordinates, and carrying out the first precision processing of the first fuzzy positioning coordinates, obtaining a first accurate positioning coordinate; acquiring a starting vertical view and an ending vertical view through a camera acquisition device; respectively converting the starting elevation map and the ending elevation map into grey-scale maps and forming a key point; obtaining the calculated height of the unmanned aerial vehicle; obtaining a calculation longitude and a calculation latitude; obtaining a second accurate positioning coordinate again based on a satellite positioning technology; determining a height adjustment coefficient, a longitude adjustment coefficient and a latitude adjustment coefficient; the actual coordinate of the unmanned aerial vehicle is the product of the accurate positioning coordinate obtained in real time and the corresponding adjustment coefficient. According to the invention, the satellite positioning technology and the visual positioning technology are combined, and the high-precision positioning performance is maintained in various environments.

Description

Technical Field

[0001] The present invention relates to the technical field of unmanned aerial vehicle positioning, and in particular to a method for collecting visual positioning coordinates of a facade reference object based on a side-mounted camera of an unmanned aerial vehicle. Background Art

[0002] An efficient and accurate positioning coordinate method is an important factor affecting the performance of UAV missions. Currently, satellite navigation positioning technology, inertial navigation technology, or a combination of the two are commonly used as methods for UAV navigation and positioning. Among them, satellite navigation positioning technology can achieve global coverage, work around the clock, has no geographical restrictions and is not affected by weather; inertial navigation technology can achieve high-precision positioning in the short term.

[0003] However, from the perspective of positioning accuracy, the horizontal positioning accuracy of civilian GPS receivers in an open environment is usually around 5-10 meters, so high-precision operation of drones cannot be achieved based on satellite navigation and positioning technology; inertial navigation technology is a navigation technology that uses inertial sensors (such as accelerometers and gyroscopes) to measure the motion state of an object, thereby achieving continuous measurement and calculation of the object's position, speed and attitude. However, due to accumulated errors, drift, etc., drones cannot have high accuracy when operating for a long time.

[0004] Therefore, how to design a high-precision positioning coordinate method is particularly important. Summary of the invention

[0005] In view of this, the present invention proposes a method for collecting visual positioning coordinates of facade reference objects based on a side-mounted camera of a drone, aiming to solve the problem of poor positioning accuracy when drones operate for a long time in current technology.

[0006] The present invention proposes a method for collecting visual positioning coordinates of a facade reference object based on a side-mounted camera of an unmanned aerial vehicle, comprising:

[0007] Based on the satellite positioning technology, the coordinates of the UAV are obtained and recorded as first fuzzy positioning coordinates, and the first fuzzy positioning coordinates are subjected to first precision processing to obtain first precise positioning coordinates;

[0008] A camera collection device with a fixed angle is arranged on the side of the drone, and the drone moves at a constant speed, and obtains a starting elevation image and an ending elevation image through the camera collection device at the beginning and the end of the movement respectively;

[0009] The starting elevation image and the ending elevation image are respectively converted into grayscale images and a key point in the grayscale image is selected based on the SI FT algorithm; the calculated height of the drone is obtained according to the moving distance of the drone and the moving distance of the key points in the two grayscale images;

[0010] Obtain the calculated longitude and latitude based on the precise positioning coordinates and heading angle of the drone;

[0011] Based on the satellite positioning technology, the coordinates of the UAV are obtained again and recorded as second fuzzy positioning coordinates, and the second fuzzy positioning coordinates are subjected to second precision processing to obtain second precise positioning coordinates;

[0012] The altitude adjustment coefficient, longitude adjustment coefficient and latitude adjustment coefficient are determined according to the calculated altitude, calculated longitude, calculated latitude of the drone and the relationship between the altitude, longitude and latitude in the second precise positioning coordinates; the actual coordinates of the drone are the product of the precise positioning coordinates obtained in real time and the corresponding adjustment coefficient.

[0013] Furthermore, the first refinement process includes:

[0014] Obtaining a plurality of the fuzzy positioning coordinates and an accuracy trust factor corresponding to each of the fuzzy positioning coordinates through at least two different satellite positioning methods, and obtaining the first precise positioning coordinates by processing the fuzzy positioning coordinates and the accuracy trust factor;

[0015] The precise positioning coordinates satisfy the following relationship:

[0016]

[0017] Among them, X, Y, and Z represent the longitude, latitude, and altitude of the precise positioning coordinates of the drone, respectively; xi, yi, and zi represent the longitude, latitude, and altitude given by the drone in the i-th satellite system, respectively.

[0018] Furthermore, the accuracy trust factor is obtained based on the signal-to-noise ratio, the estimated positioning error, and the number of satellites that can be received and used at the current position of the UAV;

[0019] The accuracy confidence factor is obtained through the following relationship:

[0020]

[0021] Among them, ωi represents the accuracy trust factor of the i-th satellite system, σi represents the estimated positioning error of the i-th satellite system; SNRi represents the signal-to-noise ratio of the i-th satellite system; SNRmax represents the highest signal-to-noise ratio of all satellite systems; Ni represents the number of visible satellites of the i-th satellite system; Nmax represents the maximum number of visible satellites of all satellite systems.

[0022] Further, when a key point in the grayscale image is selected based on the SI FT algorithm, it includes:

[0023] The contrast of the feature points in all the grayscale images is obtained, the contrast average value and the contrast mean square error value of the same feature point in different grayscale images are calculated respectively, and a key point is selected according to the contrast average value and the contrast mean square error value of each feature point.

[0024] Furthermore, when selecting a key point according to the contrast average value and the contrast mean square error value of each feature point, the feature point with the largest reliability evaluation value is selected as the key point, and the reliability evaluation value satisfies the following relationship:

[0025]

[0026] Among them, F is the reliability evaluation value, α and β are constants, is the average contrast, M 2 is the contrast mean square error value.

[0027] Furthermore, the calculated height of the drone is obtained according to the moving distance of the drone and the moving distance of the key points in the two grayscale images, including:

[0028] For the two grayscale images, a rectangular coordinate system is established with the same position as the origin and the same direction as the x-axis and y-axis, the coordinates of the key points in the two grayscale images are obtained respectively, and the distance between the key points in the two grayscale images is calculated according to the coordinates;

[0029] The movement time and heading angle of the UAV are obtained, and the displacement distance of the UAV is calculated. The movement distance of the UAV satisfies the following relationship:

[0030] x=v·t·cosθ;

[0031] Among them, x is the displacement distance of the UAV, which represents the displacement value in the direction parallel to the shooting surface; v is the movement speed of the UAV, and t is the movement time of the UAV.

[0032] The size of the scale is obtained according to the distance between the key points in the two grayscale images and the movement distance of the UAV; the distance between the key point in any grayscale image and the ground is obtained, and the calculated height of the UAV is calculated according to the scale.

[0033] Furthermore, the longitude and latitude are calculated based on the precise positioning coordinates and heading angle of the drone, including:

[0034] The calculation latitude satisfies the following relationship:

[0035]

[0036] Among them, Φ is the calculation latitude and R is the radius of the earth.

[0037] Furthermore, the longitude and latitude are calculated according to the precise positioning coordinates and heading angle of the drone, and the method further includes:

[0038] The calculation latitude satisfies the following relationship:

[0039]

[0040] Where λ is the calculated longitude.

[0041] Furthermore, the altitude adjustment coefficient is the ratio of the calculated altitude to the altitude in the second precise positioning coordinates; the longitude adjustment coefficient is the ratio of the calculated longitude to the longitude in the second precise positioning coordinates; and the latitude adjustment coefficient is the ratio of the calculated latitude to the latitude in the second precise positioning coordinates.

[0042] Furthermore, in the second precision processing, the precise positioning coordinates satisfy the following relationship:

[0043]

[0044] Among them, X, Y, and Z represent the longitude, latitude, and altitude of the precise positioning coordinates of the drone, respectively; xi, yi, and zi represent the longitude, latitude, and altitude given by the drone in the i-th satellite system, respectively.

[0045] Compared with the prior art, the present invention has the following beneficial effects:

[0046] By combining satellite positioning and visual positioning technology, a high-precision UAV positioning method is provided. First, satellite positioning technology provides preliminary coordinate information for the UAV, but due to its limited accuracy, this solution obtains more accurate positioning data by refining these fuzzy coordinates. Then, the vertical image collected by the side-mounted camera of the UAV during the movement of the UAV is combined with the key points extracted by the S IFT algorithm to further improve the positioning accuracy. By fusing visual data with satellite data, visual positioning can still provide important auxiliary information even in an environment with weak satellite signals, ensuring the stability and reliability of the positioning results. The multi-dimensional data of the UAV, such as the moving height, longitude and latitude, can be accurately calculated, and the positioning error can be further corrected by adjusting the coefficient. In addition, the scheme also includes a reliability evaluation of key points to ensure that the selected feature points can be matched stably and reliably. Through dynamic adjustment and multiple precision processing, this method can provide the precise positioning coordinates of the UAV in real time and accurately, and is suitable for UAV tasks in long-term operations and various complex environments. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Various other advantages and benefits will become apparent to those of ordinary skill in the art by reading the detailed description of the preferred embodiments below. The accompanying drawings are only for the purpose of illustrating the preferred embodiments and are not to be considered as limiting the present invention. Moreover, the same reference symbols are used throughout the accompanying drawings to represent the same components. In the accompanying drawings:

[0048] Figure 1 The present invention is a flowchart of a method for collecting visual positioning coordinates of a vertical reference object based on a side-mounted camera of an unmanned aerial vehicle. DETAILED DESCRIPTION

[0049] Exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although exemplary embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments described herein. On the contrary, these embodiments are provided in order to enable a more thorough understanding of the present disclosure and to fully convey the scope of the present disclosure to those skilled in the art. It should be noted that, in the absence of conflict, the embodiments of the present invention and the features described in the embodiments can be combined with each other. The present invention will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0050] like Figure 1 As shown, an embodiment of the present invention provides a method for collecting visual positioning coordinates of a facade reference object based on a side-mounted camera of an unmanned aerial vehicle, which includes:

[0051] Based on the satellite positioning technology, the coordinates of the UAV are obtained and recorded as first fuzzy positioning coordinates, and the first fuzzy positioning coordinates are subjected to first precision processing to obtain first precise positioning coordinates;

[0052] A camera collection device with a fixed angle is arranged on the side of the drone, and the drone moves at a constant speed, and obtains a starting elevation image and an ending elevation image through the camera collection device at the beginning and the end of the movement respectively;

[0053] The starting elevation image and the ending elevation image are respectively converted into grayscale images and a key point in the grayscale image is selected based on the SI FT algorithm; the calculated height of the drone is obtained according to the moving distance of the drone and the moving distance of the key points in the two grayscale images;

[0054] Obtain the calculated longitude and latitude based on the precise positioning coordinates and heading angle of the drone;

[0055] Based on the satellite positioning technology, the coordinates of the UAV are obtained again and recorded as second fuzzy positioning coordinates, and the second fuzzy positioning coordinates are subjected to second precision processing to obtain second precise positioning coordinates;

[0056] The altitude adjustment coefficient, longitude adjustment coefficient and latitude adjustment coefficient are determined according to the calculated altitude, calculated longitude, calculated latitude of the drone and the relationship between the altitude, longitude and latitude in the second precise positioning coordinates; the actual coordinates of the drone are the product of the precise positioning coordinates obtained in real time and the corresponding adjustment coefficient.

[0057] By refining the initial fuzzy positioning coordinates, the positioning accuracy of the drone can be significantly improved. Satellite positioning technology provides rough coordinates as a basis. After precision processing, more accurate positioning information can be obtained to reduce positioning errors. The drone's side-mounted camera uses facade image acquisition and extracts key points through the SIFT algorithm during its uniform movement to effectively calculate the height of the drone. This visual positioning method supplements the shortcomings of satellite positioning, especially providing additional positioning information in complex environments. By determining the adjustment coefficients for the precise positioning coordinates of the drone and the calculated altitude, longitude, and latitude, dynamic correction of real-time positioning data is achieved. In this way, accurate positioning results can be continuously provided even during long-term operations. Combining satellite positioning and visual positioning technology, this method can maintain high-precision positioning performance in various environments. Even when the satellite signal is weak or blocked, the visual positioning data can still provide important supplements, improving the stability and reliability of the overall positioning system. Through precision processing and dynamic adjustment, this solution is particularly suitable for long-term, high-precision drone operations, reducing the accuracy drop caused by the accumulation of positioning errors, and ensuring the reliability and accuracy of drones when performing complex tasks. The altitude, longitude and latitude adjustment coefficients used are used to optimize the positioning data based on the actual measurement results, ensuring that the final actual coordinates are more consistent with the actual situation and improving the overall performance of the system.

[0058] In some embodiments of the present application, the first refinement process includes:

[0059] Obtaining a plurality of the fuzzy positioning coordinates and an accuracy trust factor corresponding to each of the fuzzy positioning coordinates through at least two different satellite positioning methods, and obtaining the first precise positioning coordinates by processing the fuzzy positioning coordinates and the accuracy trust factor;

[0060] The precise positioning coordinates satisfy the following relationship:

[0061]

[0062] Among them, X, Y, and Z represent the longitude, latitude, and altitude of the precise positioning coordinates of the drone, respectively; xi, yi, and zi represent the longitude, latitude, and altitude given by the drone in the i-th satellite system, respectively.

[0063] It is understandable that by integrating positioning data from at least two different satellite systems, the accuracy of positioning can be significantly improved. The fuzzy positioning coordinates provided by different satellite systems each have their own characteristics and limitations. Combining these data can reduce the impact of single system errors and obtain more accurate coordinates. By assigning an accuracy trust factor to each fuzzy positioning coordinate, a weighted average can be performed according to the reliability weight of each coordinate. This method helps to filter out data with poor accuracy and improve the reliability and accuracy of the final precise positioning coordinates. Using the accuracy trust factor and the fuzzy positioning coordinates for comprehensive processing, the precise positioning coordinates can be optimized according to the weights of different data. This method makes the final precise positioning coordinates more in line with the actual situation and reduces errors.

[0064] In some embodiments of the present application, the accuracy trust factor is obtained based on the signal-to-noise ratio, the estimated positioning error, and the number of satellites that can be received and used at the current position of the drone;

[0065] The accuracy confidence factor is obtained through the following relationship:

[0066]

[0067] Among them, ωi represents the accuracy trust factor of the i-th satellite system, σi represents the estimated positioning error of the i-th satellite system; SNRi represents the signal-to-noise ratio of the i-th satellite system; SNRmax represents the highest signal-to-noise ratio of all satellite systems; Ni represents the number of visible satellites of the i-th satellite system; Nmax represents the maximum number of visible satellites of all satellite systems.

[0068] It is understandable that by combining the signal-to-noise ratio, the number of visible satellites, and the estimated positioning error, multiple aspects of the positioning data can be comprehensively considered to provide a more accurate quality assessment. The calculation of the accuracy trust factor can effectively distinguish the data quality of different satellite systems, giving higher weight to high-quality data, thereby improving the accuracy of the final positioning result. In complex environments (such as high-rise buildings in cities or inclement weather), the signal-to-noise ratio may be low, the positioning error may be large, or the number of visible satellites may be small. The calculation of the accuracy trust factor can dynamically adjust the reliability of the positioning result.

[0069] In some embodiments of the present application, when a key point in the grayscale image is selected based on the SI FT algorithm, it includes:

[0070] The contrast of the feature points in all the grayscale images is obtained, the contrast average value and the contrast mean square error value of the same feature point in different grayscale images are calculated respectively, and a key point is selected according to the contrast average value and the contrast mean square error value of each feature point.

[0071] It can be understood that by calculating the contrast mean square error value, feature points with stable contrast changes can be effectively identified, reducing mismatches caused by noise or other factors. Feature points with high contrast usually have more obvious image features, and selecting these feature points as key points can improve the accuracy and reliability of matching. Combining the contrast average and mean square error values ​​to select feature points makes the algorithm more adaptable to image changes (such as changes in illumination and viewing angle), thereby enhancing the robustness of the overall system.

[0072] In some embodiments of the present application, when selecting a key point according to the contrast average value and the contrast mean square error value of each feature point, the feature point with the largest reliability evaluation value is selected as the key point, and the reliability evaluation value satisfies the following relationship:

[0073]

[0074] Among them, F is the reliability evaluation value, α and β are constants, is the average contrast, M 2 is the contrast mean square error value.

[0075] It should be noted that by comprehensively considering the contrast average and mean square error value, it is possible to better select feature points that are stable and obvious in different images, thereby improving the accuracy of matching. Selecting feature points with high reliability evaluation values ​​means that these feature points are more consistent in different images, which can reduce mismatches caused by image noise or lighting changes. By adjusting the weight coefficients α and β, the feature point selection algorithm can be optimized for different application scenarios and its adaptability to various environmental conditions can be enhanced.

[0076] In some embodiments of the present application, obtaining the calculated height of the drone according to the moving distance of the drone and the moving distances of key points in two grayscale images includes:

[0077] For the two grayscale images, a rectangular coordinate system is established with the same position as the origin and the same direction as the x-axis and y-axis, the coordinates of the key points in the two grayscale images are obtained respectively, and the distance between the key points in the two grayscale images is calculated according to the coordinates;

[0078] The movement time and heading angle of the UAV are obtained, and the displacement distance of the UAV is calculated. The movement distance of the UAV satisfies the following relationship:

[0079] x=v·t·cosθ;

[0080] Among them, x is the displacement distance of the UAV, which represents the displacement value in the direction parallel to the shooting surface; v is the movement speed of the UAV, and t is the movement time of the UAV.

[0081] The size of the scale is obtained according to the distance between the key points in the two grayscale images and the movement distance of the UAV; the distance between the key point in any grayscale image and the ground is obtained, and the calculated height of the UAV is calculated according to the scale.

[0082] It should be noted that by combining the moving distance of the key points in the image with the actual displacement distance of the drone, the drone's altitude can be accurately calculated. This method is applicable to different environments and shooting conditions, and the altitude can be calculated by simply using the key points in the grayscale image and the drone's motion data. By combining image processing technology and motion data, the positioning accuracy and reliability of drones in practical applications can be significantly improved.

[0083] In some embodiments of the present application, the longitude and latitude are calculated according to the precise positioning coordinates and heading angle of the drone, including:

[0084] The calculation latitude satisfies the following relationship:

[0085]

[0086] Among them, Φ is the calculation latitude and R is the radius of the earth.

[0087] In some embodiments of the present application, obtaining the calculated longitude and calculated latitude according to the precise positioning coordinates and heading angle of the drone also includes:

[0088] The calculation latitude satisfies the following relationship:

[0089]

[0090] Where λ is the calculated longitude.

[0091] It should be noted that the calculation of longitude and latitude is corrected by taking into account the curvature of the earth and the heading angle of the drone, thereby improving the accuracy of positioning. This is crucial for application scenarios that require high-precision geographic positioning, such as map making, drone monitoring and inspection, etc. These calculation methods can adapt to complex geographical environments and changes, such as terrain undulations and movement directions, thereby ensuring positioning accuracy in different environments. Through accurate longitude and latitude calculations, the stability and reliability of the navigation system can be improved, problems caused by positioning errors can be reduced, and the efficiency and effectiveness of drone mission execution can be improved.

[0092] In some embodiments of the present application, the altitude adjustment coefficient is the ratio of the calculated altitude to the altitude in the second precise positioning coordinates; the longitude adjustment coefficient is the ratio of the calculated longitude to the longitude in the second precise positioning coordinates; and the latitude adjustment coefficient is the ratio of the calculated latitude to the latitude in the second precise positioning coordinates.

[0093] In some embodiments of the present application, in the second refinement process, the precise positioning coordinates satisfy the following relationship:

[0094]

[0095] Among them, X, Y, and Z represent the longitude, latitude, and altitude of the precise positioning coordinates of the drone, respectively; xi, yi, and zi represent the longitude, latitude, and altitude given by the drone in the i-th satellite system, respectively.

[0096] It should be noted that the altitude, longitude and latitude adjustment coefficients are used to correct the positioning data of the drone to make it closer to the actual position, thereby improving the overall positioning accuracy. By adjusting the ratio of the calculated altitude, longitude and latitude to the second precise positioning coordinates, the deviation caused by various errors can be effectively corrected. The second precision processing provides a weighted average precise positioning coordinate by combining the data of multiple satellite systems and their accuracy trust factors. This method effectively improves the consistency and accuracy of the data. By adjusting the coefficients, the stability and reliability of the drone positioning system can be maintained in complex environments (such as high buildings, extreme weather, etc.), thereby reducing positioning errors and improving the accuracy of drone mission execution. The precision processing combines multiple information from different satellites and weightedly considers the signal-to-noise ratio and error of each satellite system, which helps to optimize the performance of the algorithm and ensure accurate positioning in practical applications. This adjustment mechanism supports drones to provide higher positioning accuracy and reliability when performing delicate tasks (such as terrain mapping, building inspection, etc.) to meet higher operational requirements.

[0097] In summary, this solution combines satellite positioning technology, visual image processing and motion data to significantly improve the positioning accuracy of drones. By accurately processing the fuzzy positioning coordinates and analyzing the image data during the motion process, the errors existing in traditional satellite positioning technology can be corrected to provide more accurate positioning results. By calculating the adjustment coefficients of altitude, longitude and latitude and applying them to the actual coordinates, this solution can correct the positioning information of the drone in real time. This dynamic adjustment mechanism can effectively handle the deviations caused by environmental changes or equipment errors, ensuring that the altitude and position data of the drone are more accurate when performing tasks. By integrating satellite data and visual data, the reliability of the positioning system is enhanced. Especially in complex or extreme environments, traditional satellite positioning technology may be interfered. This solution ensures that the system can maintain high stability and reliability in various environments through the application of visual assistance and adjustment coefficients. The precise positioning capability of the solution enables drones to provide accurate data support when performing high-precision operations (such as terrain mapping, building inspection, environmental monitoring, etc.). This is particularly important for application scenarios that require detailed geographic information, and improves the operation capabilities of drones in professional fields. The use of precision trust factors effectively reduces the impact of satellite positioning errors on drone positioning. By weighted processing of data from multiple satellite systems, the impact of single system errors is reduced and the overall positioning accuracy is improved. By acquiring and processing positioning data in real time, it is able to respond quickly to changes in the drone's position and provide the latest positioning information. This real-time update capability is essential for drones to perform tasks in dynamic environments (such as rescue operations, emergency response, etc.). By selecting the best key points through efficient image processing algorithms (such as S IFT) and optimizing the reliability of feature points based on contrast and mean square error values, the application effect of image data in positioning is further improved, and the accuracy and robustness of the algorithm are increased.

[0098] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, ordinary technicians in the relevant field should understand that the specific implementation methods of the present invention can still be modified or replaced by equivalents. Any modification or equivalent replacement that does not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A method for collecting visual positioning coordinates of facade reference objects based on a side-mounted camera of an unmanned aerial vehicle, characterized in that: include: Based on the satellite positioning technology, the coordinates of the UAV are obtained and recorded as first fuzzy positioning coordinates, and the first fuzzy positioning coordinates are subjected to first precision processing to obtain first precise positioning coordinates; A camera collection device with a fixed angle is arranged on the side of the drone, and the drone moves at a constant speed, and obtains a starting elevation image and an ending elevation image through the camera collection device at the beginning and end of the movement respectively; The starting elevation image and the ending elevation image are respectively converted into grayscale images and a key point in the grayscale image is selected based on the SIFT algorithm; the calculated height of the drone is obtained according to the moving distance of the drone and the moving distance of the key points in the two grayscale images; Obtain the calculated longitude and latitude based on the precise positioning coordinates and heading angle of the drone; Based on the satellite positioning technology, the coordinates of the UAV are obtained again and recorded as second fuzzy positioning coordinates, and the second fuzzy positioning coordinates are subjected to second precision processing to obtain second precise positioning coordinates; The altitude adjustment coefficient, longitude adjustment coefficient and latitude adjustment coefficient are determined according to the calculated altitude, calculated longitude, calculated latitude of the drone and the relationship between the altitude, longitude and latitude in the second precise positioning coordinates; the actual coordinates of the drone are the product of the precise positioning coordinates obtained in real time and the corresponding adjustment coefficient.

2. According to claim 1, a method for collecting visual positioning coordinates of facade reference objects based on a side-mounted camera of an unmanned aerial vehicle is characterized in that: The first refinement process comprises: Obtaining a plurality of the fuzzy positioning coordinates and an accuracy trust factor corresponding to each of the fuzzy positioning coordinates through at least two different satellite positioning methods, and obtaining the first precise positioning coordinates by processing the fuzzy positioning coordinates and the accuracy trust factor; The precise positioning coordinates satisfy the following relationship: Among them, X, Y, and Z represent the longitude, latitude, and altitude of the precise positioning coordinates of the drone, respectively; xi, yi, and zi represent the longitude, latitude, and altitude given by the drone in the i-th satellite system, respectively.

3. The method for collecting visual positioning coordinates of facade reference objects based on a side-mounted camera of an unmanned aerial vehicle according to claim 1 is characterized in that: The accuracy trust factor is obtained based on the signal-to-noise ratio, the estimated positioning error, and the number of satellites that can be received and used at the current position of the drone; The accuracy confidence factor is obtained by the following relationship: Among them, ωi represents the accuracy trust factor of the i-th satellite system, σi represents the estimated positioning error of the i-th satellite system; SNRi represents the signal-to-noise ratio of the i-th satellite system; SNRmax represents the highest signal-to-noise ratio of all satellite systems; Ni represents the number of visible satellites of the i-th satellite system; Nmax represents the maximum number of visible satellites of all satellite systems.

4. The method for collecting visual positioning coordinates of facade reference objects based on a side-mounted camera of an unmanned aerial vehicle according to claim 3 is characterized in that: When a key point in the grayscale image is selected based on the SIFT algorithm, it includes: The contrast of the feature points in all the grayscale images is obtained, the contrast average value and the contrast mean square error value of the same feature point in different grayscale images are calculated respectively, and a key point is selected according to the contrast average value and the contrast mean square error value of each feature point.

5. The method for collecting visual positioning coordinates of facade reference objects based on a side-mounted camera of an unmanned aerial vehicle according to claim 4 is characterized in that: When selecting a key point based on the contrast average value and the contrast mean square error value of each feature point, the feature point with the largest reliability evaluation value is selected as the key point, and the reliability evaluation value satisfies the following relationship: Among them, F is the reliability evaluation value, α and β are constants, is the average contrast, M 2 is the contrast mean square error value.

6. The method for collecting visual positioning coordinates of facade reference objects based on a side-mounted camera of an unmanned aerial vehicle according to claim 5 is characterized in that: The calculated height of the drone is obtained based on the moving distance of the drone and the moving distance of the key points in the two grayscale images, including: For the two grayscale images, a rectangular coordinate system is established with the same position as the origin and the same direction as the x-axis and y-axis, the coordinates of the key points in the two grayscale images are obtained respectively, and the distance between the key points in the two grayscale images is calculated according to the coordinates; The movement time and heading angle of the UAV are obtained, and the displacement distance of the UAV is calculated. The movement distance of the UAV satisfies the following relationship: x=v·t·cosθ; Among them, x is the displacement distance of the UAV, which represents the displacement value in the direction parallel to the shooting surface; v is the movement speed of the UAV, and t is the movement time of the UAV. The size of the scale is obtained according to the distance between the key points in the two grayscale images and the movement distance of the UAV; the distance between the key point in any grayscale image and the ground is obtained, and the calculated height of the UAV is calculated according to the scale.

7. The method for collecting visual positioning coordinates of facade reference objects based on a side-mounted camera of an unmanned aerial vehicle according to claim 6 is characterized in that: The longitude and latitude are calculated based on the precise positioning coordinates and heading angle of the drone, including: The calculation latitude satisfies the following relationship: Among them, Φ is the calculation latitude and R is the radius of the earth.

8. The method for collecting visual positioning coordinates of facade reference objects based on a side-mounted camera of an unmanned aerial vehicle according to claim 7 is characterized in that: The longitude and latitude are calculated based on the precise positioning coordinates and heading angle of the drone, and also include: The calculation latitude satisfies the following relationship: Where λ is the calculated longitude.

9. The method for collecting visual positioning coordinates of facade reference objects based on a side-mounted camera of an unmanned aerial vehicle according to claim 8, characterized in that: The altitude adjustment coefficient is the ratio of the calculated altitude to the altitude in the second precise positioning coordinates; the longitude adjustment coefficient is the ratio of the calculated longitude to the longitude in the second precise positioning coordinates; the latitude adjustment coefficient is the ratio of the calculated latitude to the latitude in the second precise positioning coordinates.

10. The method for collecting visual positioning coordinates of facade reference objects based on a side-mounted camera of an unmanned aerial vehicle according to claim 9, characterized in that: In the second precision processing, the precise positioning coordinates satisfy the following relationship: Among them, X, Y, and Z represent the longitude, latitude, and altitude of the precise positioning coordinates of the drone, respectively; xi, yi, and zi represent the longitude, latitude, and altitude given by the drone in the i-th satellite system, respectively.

Images