A method for calibrating attitude angle error of a UAV

Abstract

The application provides a UAV attitude angle error calibration method, comprising: acquiring a current vertical azimuth angle of a UAV; using the UAV to perform tilt photography, establishing a first mapping relationship between a distance between a UAV photography center point and a target photography point and a vertical azimuth angle correction value of the UAV; measuring a distance real value between the UAV photography center point and a plurality of target photography points through map calibration; solving an optimal vertical azimuth angle correction value of the UAV according to an iterative regression method; and calculating a vertical azimuth angle error value of the UAV according to the optimal vertical azimuth angle correction value and the current vertical azimuth angle of the UAV. The application can correct the attitude angle of the UAV through the iterative regression method.

Description

Technical Field

[0001] The present invention relates to the field of unmanned aerial vehicle (UAV) photography, and more particularly to a method for calibrating an attitude angle error of a UAV. Background Art

[0002] With the development of drone technology, monitoring and regulatory services are placing higher demands on real-time drone monitoring. For example, combining drone video streams with spatiotemporal data to provide a more intuitive understanding of the actual situation in the target area is crucial. To meet this demand, the efficient and accurate conversion of real-time drone footage into geographic coordinates and integration with maps has become a pressing technical challenge.

[0003] The current technical solution is to collect historical data from the monitoring area and combine it with the results of drone flight to build a simulation model. Although this solution can achieve the effect of video fusion, it requires a lot of data collection and processing work in the early stage, and a lot of processing of drone flight results in the later stage. There are problems such as implementation difficulties and low real-time performance, making it difficult to achieve real-time fusion. Summary of the Invention

[0004] In view of the technical problems existing in the prior art, the present invention provides a method for calibrating the attitude angle error of a UAV, comprising:

[0005] Get the current vertical azimuth of the drone;

[0006] Using a drone to perform oblique photography, establishing a first mapping relationship between the distance between the drone photography center point and the target photography point and the drone vertical azimuth correction value;

[0007] Measure the true value of the distance between the center point of drone photography and multiple target photography points through map positioning;

[0008] Calculate the predicted distance between the drone photography center point and each target photography point based on the first mapping relationship according to the vertical azimuth angle correction value updated in each iteration;

[0009] Calculate the loss value of each iteration based on the distance prediction value and the true distance value;

[0010] Based on the loss value and learning rate of each iteration, the vertical azimuth correction value of the next iteration is updated, and the iteration is repeated until the loss value meets the preset conditions to obtain the optimal vertical azimuth correction value;

[0011] The vertical azimuth angle error value of the UAV is calculated according to the optimal vertical azimuth angle correction value and the current vertical azimuth angle of the UAV.

[0012] The present invention provides a method for calibrating the attitude angle error of a UAV. The method establishes mapping relationships between the vertical azimuth angle, horizontal azimuth angle, and viewing angle of the UAV, solves the error values ​​of the vertical azimuth angle, horizontal azimuth angle, and viewing angle of the UAV through an iterative regression method, and corrects the vertical azimuth angle, horizontal azimuth angle, and viewing angle of the UAV respectively according to the error values. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] Figure 1 A flow chart of a method for calibrating the vertical azimuth angle error in the attitude angle of a UAV provided by one embodiment of the present invention;

[0014] Figure 2 Schematic diagram of the relationship between the distance between the center point of drone photography and the target photography point and the vertical azimuth angle of the drone according to an embodiment of the present invention;

[0015] Figure 3 A schematic diagram of a method for calibrating the horizontal azimuth angle error in the attitude angle of a drone provided by an embodiment of the present invention;

[0016] Figure 4 Schematic diagram of the relationship between the geographical coordinates of the center point of drone photography and the horizontal azimuth angle of the drone according to an embodiment of the present invention;

[0017] Figure 5 A flow chart of a method for calibrating the viewing angle error in the attitude angle of a drone provided by one embodiment of the present invention;

[0018] Figure 6 Schematic diagram of the relationship between the drone's photography range boundary distance and the drone's viewing angle according to an embodiment of the present invention. DETAILED DESCRIPTION

[0019] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention. In addition, the technical features in the various embodiments or single embodiments provided by the present invention can be arbitrarily combined with each other to form a feasible technical solution. This combination is not restricted by the sequence of steps and / or structural composition mode, but must be based on the ability of ordinary technicians in this field to implement it. When the combination of technical solutions is contradictory or cannot be implemented, it should be deemed that this combination of technical solutions does not exist and is not within the scope of protection required by the present invention.

[0020] See also Figure 1, is a flow chart of a method for calibrating the attitude angle error of a UAV provided by the present invention, such as Figure 1 As shown, the method includes:

[0021] Step 11: Get the current vertical azimuth angle of the drone.

[0022] It is understandable that the attitude angle of a drone may deviate due to some objective factors, so it is necessary to correct the attitude angle of the drone. The attitude angle of a drone includes the vertical azimuth angle, the horizontal azimuth angle, and the viewing angle.

[0023] First, the vertical azimuth angle of the UAV is corrected, so the vertical azimuth angle error value of the UAV needs to be solved. The vertical azimuth angle error value of the UAV is calibrated below.

[0024] Step 12: Use a drone to perform oblique photography, and establish a first mapping relationship between the distance between the drone photography center point and the target photography point and the drone vertical azimuth correction value.

[0025] Understandably, see Figure 2 , a first mapping relationship between the distance between the drone photography center point and the target photography point and the drone vertical azimuth angle correction value can be established.

[0026] Among them, the first mapping relationship is:

[0027] ;

[0028] Among them, r represents the distance between the center point of the drone photography and the target photography point, h is the height of the drone photography, is the elevation correction value of the UAV position, is the elevation correction value of the target point, t is the vertical azimuth angle of the UAV, is the vertical azimuth error value of the UAV, It is the vertical azimuth correction value of the UAV.

[0029] Among them, the UAV position elevation correction value and the elevation correction value of the target point It can be calculated using a digital elevation model.

[0030] Specifically, based on the digital elevation model, a second mapping relationship between geographic coordinates and elevation correction values ​​is established:

[0031]

[0032] in, is the elevation correction value, is the elevation correction function of x and y coordinates, x and y represent geographic coordinates respectively, i is the column index of the digital elevation model converted according to the x coordinate, m(x) is the column index conversion function, j is the row index of the digital elevation model converted according to the y coordinate, n(y) is the row index conversion function, h is the ground elevation of the current geographic coordinate, and h(i,j) represents the elevation value of the i-th column and j-th row in the digital elevation model;

[0033] ;

[0034] ;

[0035] Among them, x and y are the geographic horizontal and vertical coordinates respectively. 、 are the starting point abscissa and ordinate of the digital elevation model respectively. 、 are the end point abscissa and ordinate of the digital elevation model, respectively; W is the width of the digital elevation model, and H is the height of the digital elevation model;

[0036] Calculate the elevation correction value of the drone position based on the geographic coordinates of the drone photography position and the second mapping relationship ;

[0037] According to the characteristic relationship of the center point of the photographic target, the corresponding point is found on the map, the geographic coordinates of the target point are obtained by measuring on the map, and the elevation correction value of the target point is calculated based on the geographic coordinates of the target point and the second mapping relationship. .

[0038] Step 13: Measure the true value of the distance between the drone photography center point and multiple target photography points through map positioning.

[0039] It is understandable that after establishing the first mapping relationship between the distance between the drone photography center point and the target photography point and the drone vertical azimuth correction value, multiple target points can be selected and the true value of the distance between the drone photography center point and multiple target photography points can be measured through map calibration.

[0040] Step 14: Calculate the predicted distance between the drone photography center point and each target photography point based on the first mapping relationship according to the vertical azimuth angle correction value updated in each iteration.

[0041] Step 15: Calculate the loss value of each iteration based on the distance prediction value and the true distance value.

[0042] Step 16, based on the loss value of each iteration and the learning rate, updating the vertical azimuth correction value of the next iteration, repeating the iteration until the loss value meets the preset condition, and obtaining the optimal vertical azimuth correction value.

[0043] It can be understood that for the geographic coordinates of each target point, the elevation correction value of each target point can be calculated according to the second mapping relationship, and then the distance prediction value between the unmanned aerial vehicle photography center point and each target photography point can be calculated according to the first mapping relationship.

[0044] A plurality of target points constitute a sample set for iterative training, and the optimal vertical azimuth correction value of the unmanned aerial vehicle is solved through iterative training.

[0045] The process of iterative training is:

[0046] Initialize the vertical azimuth correction value , and set the initial learning rate .

[0047] Simplify the first mapping relationship as a target function:

[0048] ;

[0049] In the formula, , is the vertical azimuth correction value, which will be updated as a whole in the subsequent iteration process.

[0050] In the iterative training process, according to the vertical azimuth correction value updated for the nth time , the distance prediction value between the unmanned aerial vehicle photography center point and each target photography point is calculated based on the first mapping relationship .

[0051] According to the target point features and the unmanned aerial vehicle photography points, the distance true value r between the unmanned aerial vehicle photography center point and each target photography point is obtained by calibrating the measurement on the map n .

[0052] According to the distance prediction value and the distance true value, the loss value of each iteration is calculated, including:

[0053]

[0054] Wherein, represents the loss value of the nth iteration, is the distance true value corresponding to the ith target point, is the distance prediction value corresponding to the ith target point, and M is the number of target points.

[0055] ​The updating of the vertical azimuth correction value for the next iteration based on the loss value and learning rate of each iteration includes:

[0056]

[0057] = +

[0058] Where, is the derivative of the objective function for the nth iteration, is the vertical azimuth correction value of the nth iteration, is the vertical azimuth correction value of the n+1th iteration, is the learning rate of the nth iteration;

[0059] Among them, the learning rate of the nth iteration is The expression is:

[0060]

[0061] Where, is the learning rate for the nth iteration of training in a training cycle, n is the index of the number of iterations, and N is the total number of iterations in a training cycle.

[0062] Through the above-mentioned iterative training method, the optimal vertical azimuth correction value of the UAV can be iteratively solved.

[0063] Step 17: Calculate the vertical azimuth error value of the UAV based on the optimal vertical azimuth correction value and the current vertical azimuth of the UAV.

[0064] It can be understood that the vertical azimuth error value of the drone is calculated based on the optimal vertical azimuth correction value of the drone trained iteratively and the current vertical azimuth of the drone:

[0065]

[0066] Where, is the vertical azimuth error value, is the optimal vertical azimuth angle correction value obtained after iteration, and t is the current vertical azimuth angle of the UAV.

[0067] Through steps 1 to 7, the vertical azimuth error value of the UAV is obtained. The following introduces how to solve the horizontal azimuth error value and the viewing angle error value of the drone.

[0068] Among them, the solution process of the horizontal azimuth error value of the drone is first explained. The solution of the horizontal azimuth error value of the drone includes the following steps:

[0069] Step 21, based on the drone position coordinates, the distance between the drone photography center point and the target photography point, and the drone's current horizontal azimuth correction value, establish a third mapping relationship between the drone photography center point position coordinates and the drone's current horizontal azimuth correction value, wherein the distance between the drone photography center point and the target photography point is calculated based on the first mapping relationship.

[0070] It is understandable that the geographic coordinates of the drone photography center point are calculated based on the actual distance between the drone camera position and the target photography point and the current horizontal azimuth angle of the drone.

[0071] See also Figure 3 , the following third mapping relationship can be established:

[0072]

[0073]

[0074]

[0075] in, The mathematical position converted from the current horizontal azimuth correction value of the drone. is the current horizontal azimuth of the UAV, is the horizontal azimuth error value of the UAV, is the current horizontal azimuth correction value of the UAV, 、 They represent the horizontal and vertical coordinates of the drone's position respectively, r is the distance between the drone's photography center point and the target photography point, which can be solved according to the first mapping relationship mentioned above, and x and y are the coordinates of the drone's photography center point.

[0076] Step 22: Based on the characteristics of the center point of drone photography, the real coordinates of the center points of multiple drone photography are collected on the map.

[0077] It is understandable that step 21 establishes a third mapping relationship, which can correct the current horizontal azimuth angle of the drone As a whole, based on the characteristics of the center point of drone photography, the real coordinates of multiple drone photography center points are collected on the map, that is, multiple sets of (x, y) real values ​​are collected.

[0078] Step 23: Calculate the predicted coordinates of the center point of the drone photography based on the third mapping relationship according to the horizontal azimuth correction value updated in each iteration.

[0079] In step 24, the loss value of each iteration is calculated based on the predicted coordinates and the actual coordinates of the center point of the drone photography.

[0080] Step 25: Based on the loss value and learning rate of each iteration, the horizontal azimuth angle correction value of the next iteration is updated, and the iteration is repeated until the loss value meets the preset conditions, thereby obtaining the optimal horizontal azimuth angle correction value.

[0081] It is understandable that according to the third mapping relationship, the horizontal azimuth correction value is updated by continuous iteration. ,According to the third mapping relationship, the predicted coordinates of the center point position of each drone photography,are calculated in each iteration.

[0082] Based on the predicted and true coordinates of the center points of multiple drone photography images, the loss value for each iteration is calculated. Based on the loss value and learning rate of the current iteration, the horizontal azimuth correction value for the next iteration is updated. This iteration is repeated until the loss value meets the preset conditions, resulting in the current optimal horizontal azimuth correction value. The iterative process for finding the current optimal horizontal azimuth correction value is similar to the iterative process for finding the current optimal vertical azimuth correction value and is not described in detail here.

[0083] Step 26: Calculate the horizontal azimuth error value of the UAV based on the current optimal horizontal azimuth correction value and the current horizontal azimuth of the UAV.

[0084] It can be understood that the horizontal azimuth error value of the UAV is calculated based on the iterative current optimal horizontal azimuth correction value and the current horizontal azimuth angle of the UAV.

[0085] The following is an explanation of the process of solving the UAV's viewing angle error value. The solution of the UAV's viewing angle error value mainly includes the following steps:

[0086] Step 31, obtaining the current horizontal azimuth angle p, current vertical azimuth angle t, and current viewing angle z of the UAV;

[0087] Step 32: according to the current horizontal azimuth angle p of the UAV and the horizontal azimuth angle error value of the UAV , calculate the current horizontal azimuth correction value of the drone; and according to the current vertical azimuth t of the drone and the vertical azimuth error value of the drone , calculate the current vertical azimuth correction value of the drone.

[0088] Step 33: Determine whether the current horizontal azimuth correction value of the drone is , and whether the current horizontal azimuth correction value of the drone is 0.

[0089] It should be noted that when calibrating the UAV’s viewing angle error value, it is relatively easy to use the UAV for orthographic projection. Therefore, firstly, according to the current horizontal azimuth angle p of the UAV and the horizontal azimuth angle error value of the UAV, , calculate the current horizontal azimuth correction value of the drone, and the vertical azimuth error value of the drone according to the current vertical azimuth t of the drone and the vertical azimuth error value of the drone , calculate the current vertical azimuth correction value of the drone.

[0090] Determine the current horizontal azimuth correction value of the drone and the current vertical azimuth correction value of the drone. Determine whether the current horizontal azimuth correction value of the drone is , and whether the current horizontal azimuth correction value of the drone is 0. If so, it means that the drone is in orthographic projection, then step 34 is executed. If not, the position of the orthographic projection of the drone is found.

[0091] Step 34: If yes, a fourth mapping relationship between the current viewing angle correction value of the drone and the boundary distance of the drone's photography range is established based on the drone's elevation value.

[0092] It is understandable that when the drone is in the orthographic projection position, the drone viewing angle error value can be calibrated.

[0093] See also Figure 4 , the following fourth mapping relationship can be established:

[0094]

[0095]

[0096] in, is the half angle of the horizontal component of the drone’s viewing angle, is the error value of the lateral component, is the half angle of the longitudinal component of the drone’s viewing angle, is the error value of the longitudinal component, h is the altitude value of the UAV, 、 It is the horizontal and vertical distance of the boundary point of the drone photography range.

[0097] Step 35, based on the characteristics of the position of the boundary points of the drone photography range, the positions of multiple drone photography range boundary points are measured on the map, and the distances between multiple drone photography range boundaries are calculated;

[0098] Step 36: Substitute the boundary distances of the multiple drone photography ranges into the fourth mapping relationship, construct a regression equation, and solve the regression equation to obtain the current optimal viewing angle correction value of the drone.

[0099] According to the fourth mapping relationship established in step 34, multiple and multiple , multiple and multiple Substituting the fourth mapping relationship into the equation, a regression or iterative method is used to calculate the half-angle correction value for the lateral component of the drone's current optimal viewing angle and the half-angle correction value for the longitudinal component of the drone's current optimal viewing angle. The iterative process is the same as the iterative process for calculating the drone's current vertical azimuth correction value and the horizontal azimuth correction value described above, and is not further explained.

[0100] Step 37 : Calculate the viewing angle error value of the drone based on the current optimal viewing angle correction value of the drone and the current viewing angle of the drone.

[0101] It can be understood that the perspective error value of the drone is calculated based on the current optimal perspective correction value of the drone solved by iterative regression and the current perspective of the drone.

[0102] Specifically, the error value of the lateral component of the drone's current optimal viewing angle is calculated based on the half-angle correction value of the lateral component of the drone's viewing angle and the half-angle value of the lateral component of the drone's viewing angle. ; and calculate the error value of the longitudinal component of the drone's perspective based on the half-angle correction value of the drone's current optimal perspective longitudinal component and the half-angle value of the drone's perspective longitudinal component .

[0103] According to the above steps 11 to 17, the calibration of the vertical azimuth error value of the drone is completed, the calibration of the horizontal azimuth error value of the drone is completed in steps 21 to 26, and the calibration of the viewing angle error value of the drone is completed in steps 31 to 37.

[0104] An embodiment of the present invention provides a method for calibrating an attitude angle error value, which establishes a mapping relationship between the vertical azimuth angle, horizontal azimuth angle and viewing angle of a drone, and calibrates the vertical azimuth angle error, horizontal azimuth angle error and viewing angle error of the drone through an iterative regression method.

[0105] It should be noted that, in the above embodiments, the description of each embodiment has its own focus. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0106] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Thus, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0107] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowcharts and / or block diagrams, as well as combinations of processes and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded computer, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowcharts and / or block diagrams. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0108] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0109] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0110] Although the preferred embodiments of the present invention have been described, those skilled in the art may make additional changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present invention.

[0111] Obviously, those skilled in the art may make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if such changes and modifications fall within the scope of the claims and their equivalents, the present invention is intended to include such changes and modifications.

Claims

1. A method for calibrating the attitude angle error of a UAV, characterized in that: include: Get the current vertical azimuth of the drone; Using a drone to perform oblique photography, establishing a first mapping relationship between the distance between the drone photography center point and the target photography point and the drone vertical azimuth correction value; Measure the true value of the distance between the center point of drone photography and multiple target photography points through map positioning; Calculate the predicted distance between the drone photography center point and each target photography point based on the first mapping relationship according to the vertical azimuth angle correction value updated in each iteration; Calculate the loss value of each iteration based on the distance prediction value and the true distance value; Based on the loss value and learning rate of each iteration, the vertical azimuth correction value of the next iteration is updated, and the iteration is repeated until the loss value meets the preset conditions to obtain the optimal vertical azimuth correction value; Calculating a vertical azimuth error value of the UAV based on the optimal vertical azimuth correction value and the current vertical azimuth of the UAV; The first mapping relationship is: ； Among them, r represents the distance between the center point of the drone photography and the target photography point, h is the height of the drone photography, is the elevation correction value of the UAV position, is the elevation correction value of the target point, t is the vertical azimuth angle of the UAV, is the vertical azimuth error value of the UAV, is the vertical azimuth correction value of the UAV; The calculating, based on the vertical azimuth correction value updated in each iteration and based on the first mapping relationship, a predicted distance value between the drone photography center point and each target photography point includes: Initialize vertical azimuth correction value , and set the initial learning rate ; The first mapping relationship is simplified into the objective function: ； Where, , is the vertical azimuth correction value; According to the vertical azimuth correction value updated for the nth time , based on the first mapping relationship, calculate the distance prediction value between the drone photography center point and each target photography point ; According to the characteristics of the target point and the drone photography point, the measurement is calibrated on the map to obtain the true value of the distance between the drone photography center point and each target photography point ; Calculate the loss value of each iteration based on the distance prediction value and the true distance value, including: in, represents the loss value of the nth iteration, is the true value of the distance corresponding to the i-th target point, is the distance prediction value corresponding to the i-th target point, and M is the number of target points; The updating of the vertical azimuth correction value for the next iteration based on the loss value and learning rate of each iteration includes: = + Where, is the derivative of the objective function for the nth iteration, is the vertical azimuth correction value of the nth iteration, is the vertical azimuth correction value of the n+1th iteration, is the learning rate of the nth iteration; Among them, the learning rate of the nth iteration is The expression is: Where, is the learning rate for the nth iteration of training in a training cycle, n is the index of the number of iterations, and N is the total number of iterations in a training cycle.

2. The method for calibrating the attitude angle error of a UAV according to claim 1, wherein: UAV position elevation correction value and the elevation correction value of the target point The calculation method is: According to the digital elevation model, a second mapping relationship between geographic coordinates and elevation correction values ​​is established: in, is the elevation correction value, is the elevation correction function of x and y coordinates, x and y represent geographic coordinates respectively, i is the column index of the digital elevation model converted according to the x coordinate, m(x) is the column index conversion function, j is the row index of the digital elevation model converted according to the y coordinate, n(y) is the row index conversion function, h is the ground elevation of the current geographic coordinate, and h(i,j) represents the elevation value of the i-th column and j-th row in the digital elevation model; ； ； Among them, x and y are the geographic horizontal and vertical coordinates respectively. 、 are the starting point abscissa and ordinate of the digital elevation model respectively. 、 are the end point abscissa and ordinate of the digital elevation model, respectively; W is the width of the digital elevation model, and H is the height of the digital elevation model; Calculate the elevation correction value of the drone position based on the geographic coordinates of the drone photography position and the second mapping relationship ; According to the characteristic relationship of the center point of the photographic target, the corresponding point is found on the map, the geographic coordinates of the target point are obtained by measuring on the map, and the elevation correction value of the target point is calculated based on the geographic coordinates of the target point and the second mapping relationship. .

3. The method for calibrating the attitude angle error of a UAV according to claim 1, wherein: Calculating the vertical azimuth angle error value of the UAV according to the optimal vertical azimuth angle correction value and the vertical azimuth angle of the UAV includes: Where, is the vertical azimuth error value, is the optimal vertical azimuth angle correction value obtained after iteration, and t is the current vertical azimuth angle of the UAV.

4. The method for calibrating the attitude angle error of a UAV according to claim 1, wherein: Also includes: Establishing a third mapping relationship between the position coordinates of the drone's photography center point and the current horizontal azimuth correction value of the drone based on the drone's position coordinates, the distance between the drone's photography center point and the target photography point, and the drone's current horizontal azimuth correction value, wherein the distance between the drone's photography center point and the target photography point is calculated based on the first mapping relationship; According to the characteristics of the center point of drone photography, the real coordinates of the center points of multiple drone photography are collected on the map; Calculate the predicted coordinates of the center point of the drone photography based on the third mapping relationship according to the horizontal azimuth correction value updated in each iteration; Calculate the loss value of each iteration based on the predicted coordinates and the actual coordinates of the center point of the drone photography; Based on the loss value and learning rate of each iteration, the horizontal azimuth correction value of the next iteration is updated, and the iteration is repeated until the loss value meets the preset conditions, thereby obtaining the optimal horizontal azimuth correction value; Calculate the horizontal azimuth error value of the UAV according to the optimal horizontal azimuth correction value and the current horizontal azimuth of the UAV.

5. The method for calibrating the attitude angle error of a UAV according to claim 4, characterized in that: The third mapping relationship is: in, The mathematical position converted from the current horizontal azimuth correction value of the drone. is the current horizontal azimuth of the UAV, is the horizontal azimuth error value of the UAV, is the current horizontal azimuth correction value of the UAV, 、 They represent the horizontal and vertical coordinates of the drone’s position, r is the distance between the drone’s photography center point and the target photography point, and x and y are the coordinates of the drone’s photography center point.

6. The method for calibrating the attitude angle error of a UAV according to claim 4, characterized in that: Also includes: Get the current horizontal azimuth angle p, current vertical azimuth angle t and current viewing angle z of the drone; According to the current horizontal azimuth angle p of the UAV and the horizontal azimuth angle error value of the UAV , calculate the current horizontal azimuth correction value of the drone; and according to the current vertical azimuth t of the drone and the vertical azimuth error value of the drone , calculate the current vertical azimuth correction value of the UAV; Determine whether the current vertical azimuth correction value of the drone is , and whether the current horizontal azimuth correction value of the drone is 0; If yes, establishing a fourth mapping relationship between the current viewing angle correction value of the drone and the boundary distance of the drone's photographic range based on the drone's elevation value; According to the characteristics of the boundary points of the drone photography range, the positions of multiple drone photography range boundary points are measured on the map, and the distances between multiple drone photography range boundaries are calculated; Substituting the boundary distances of the multiple drone photography ranges into the fourth mapping relationship, constructing a regression equation, and solving the regression equation to obtain the current optimal viewing angle correction value of the drone; The perspective error value of the drone is calculated based on the current optimal perspective correction value of the drone and the current perspective of the drone.

7. The method for calibrating the attitude angle error of a UAV according to claim 6, characterized in that: The fourth mapping relationship between the current viewing angle correction value of the drone and the boundary distance of the drone's photographic range is established based on the drone's elevation value, including: in, is the half angle of the horizontal component of the drone’s viewing angle, is the error value of the lateral component, is the half angle of the longitudinal component of the drone’s viewing angle, is the error value of the longitudinal component, h is the altitude value of the UAV, 、 The horizontal and vertical distances to the boundary points of the drone's photography range; Measure multiple items via map and multiple , multiple and multiple Substituting the fourth mapping relationship into the equation, the half-angle correction value of the horizontal component of the current optimal viewing angle of the drone and the half-angle correction value of the longitudinal component of the current optimal viewing angle of the drone are calculated through a regression method.

8. The method for calibrating the attitude angle error of a UAV according to claim 7, characterized in that: The calculating the viewing angle error value of the drone according to the current optimal viewing angle correction value of the drone and the current viewing angle of the drone includes: According to the half-angle correction value of the current optimal perspective lateral component of the drone and the half-angle value of the perspective lateral component of the drone, the error value of the perspective lateral component of the drone is calculated. ; and calculate the error value of the longitudinal component of the drone's perspective based on the half-angle correction value of the drone's current optimal perspective longitudinal component and the half-angle value of the drone's perspective longitudinal component .

Images