Unmanned aerial vehicle oblique photography flight path planning method and device

Abstract

This application relates to the field of unmanned aerial vehicle (UAV) technology and discloses a method and device for planning UAV oblique photography flight paths. The method includes: acquiring boundary information of a target area, imaging parameters of the camera device mounted on the UAV, and flight parameters of the UAV; determining a first area in a geographic coordinate system based on the boundary information; determining multiple second areas, exposure intervals, and flight strip intervals; generating a set of candidate grids covering each second area based on the exposure intervals and flight strip intervals, resulting in multiple sets of candidate grids; filtering multiple sets of candidate grids to obtain multiple sets of effective grids; generating a continuous shooting path by traversing the adjacency relationships of each effective grid, resulting in multiple shooting paths; and determining the actual flight waypoints of the UAV at each shooting point based on the shooting path, flight altitude, and pitch angle, to generate multiple flight paths corresponding one-to-one with the multiple shooting paths. This application can rationally plan flight paths.

Description

Technical Field

[0001] This application relates to the field of unmanned aerial vehicle (UAV) technology, specifically to a method and device for planning UAV oblique photography flight paths. Background Technology

[0002] Before performing 3D modeling on a target area (such as a building), data about the target area must first be collected to facilitate 3D modeling. Taking visual acquisition as an example, a camera is typically used to photograph the target area, and the resulting images are then used for 3D modeling. If the target area has a certain height, to ensure comprehensive image acquisition, a camera mounted on a drone can be used to photograph the target area. By controlling the drone to fly along a flight path, the camera can capture images from different spatial positions, obtaining comprehensive images of the target area and ensuring the accuracy of the 3D modeling. How to rationally plan the drone's flight path is a problem that needs to be solved. Summary of the Invention

[0003] In view of the above problems, this application provides a method and device for planning flight paths for oblique photography of unmanned aerial vehicles (UAVs) to solve the problem of the lack of reasonable flight path planning methods in the prior art.

[0004] According to one aspect of the embodiments of this application, a method for planning oblique photography flight paths of a UAV is provided. The method includes: acquiring boundary information of a target area, imaging parameters of a camera device mounted on the UAV, and flight parameters of the UAV, wherein the imaging parameters include at least a horizontal field of view, and the flight parameters include at least flight altitude, pitch angle, directional overlap rate, and lateral overlap rate; determining a first area corresponding to the target area in a geographic coordinate system based on the boundary information; determining a second area corresponding to the first area in each of multiple flight path coordinate systems, thereby obtaining multiple second areas corresponding one-to-one with the multiple flight path coordinate systems; and determining the exposure interval based on the imaging parameters and the flight parameters. The exposure distance and the flight strip spacing are used to generate a set of candidate grids covering each of the second regions, resulting in multiple sets of candidate grids corresponding one-to-one with the multiple second regions. For each set of candidate grids, effective grids are filtered based on the second region to obtain multiple sets of effective grids corresponding one-to-one with the multiple sets of candidate grids. For each set of effective grids, a continuous shooting path is generated by traversing the adjacency relationship of the effective grids to obtain multiple shooting paths corresponding one-to-one with the multiple sets of effective grids. Based on the shooting path, the flight altitude, and the pitch angle, the actual flight waypoint of the UAV at each shooting point is determined to generate multiple flight paths corresponding one-to-one with the multiple shooting paths.

[0005] In one optional approach, obtaining the boundary information of the target area includes: obtaining the latitude and longitude coordinates of N boundary points of the target area, where N ≥ 3; determining the first region corresponding to the target area in a geographic coordinate system based on the boundary information includes: determining N endpoints corresponding one-to-one with the N boundary points in the geographic coordinate system based on the latitude and longitude coordinates of the N boundary points; determining the geometric center of the N endpoints to obtain a center point; determining the polar angle of the line connecting each endpoint to the center point to obtain N polar angles corresponding one-to-one with the N endpoints; and connecting the N endpoints in order of magnitude of the N polar angles to obtain the first region.

[0006] In an optional approach, before determining the second region corresponding to the first region in each of the multiple route coordinate systems to obtain multiple second regions corresponding one-to-one with the multiple route coordinate systems, the method further includes: determining M route directions, wherein the included angle between any two of the M route directions is an integer multiple of 180° / M, and M is an integer greater than 1; establishing a coordinate system with each of the route directions as the horizontal axis to obtain M route coordinate systems corresponding one-to-one with the M route directions; the step of determining the second region corresponding to the first region in each of the multiple route coordinate systems to obtain multiple second regions corresponding one-to-one with the multiple route coordinate systems includes: converting the first region into regions in each of the route coordinate systems to obtain M second regions corresponding one-to-one with the M route coordinate systems.

[0007] In one alternative approach, the M route directions are determined based on a reference route angle θ, where M=4, and the four route directions are θ, θ+90°, θ+180°, and θ+270°.

[0008] In one optional approach, generating a set of candidate meshes covering each of the second regions based on the exposure interval and the flight strip spacing to obtain multiple sets of candidate meshes corresponding one-to-one with the multiple second regions includes: generating a minimum bounding polygon covering the second region in each of the flight path coordinate systems to obtain multiple minimum bounding polygons corresponding one-to-one with the multiple second regions; within each minimum bounding polygon, uniformly laying out candidate mesh center points in a row and column manner with the exposure interval as the horizontal axis step size and the flight strip spacing as the vertical axis step size, and generating the candidate mesh based on the candidate mesh center points.

[0009] In one optional approach, the step of performing effective mesh filtering based on the second region for each group of candidate meshes to obtain multiple groups of effective meshes corresponding one-to-one with the multiple groups of candidate meshes includes: determining whether the candidate meshes satisfy any of the following conditions: condition 1, at least one corner point of the candidate mesh is located inside the second region; condition 2, the center point of the candidate mesh is located inside the second region; condition 3, at least one endpoint of the second region is located inside the enclosing area of ​​the candidate mesh; and determining the candidate meshes that satisfy any one of conditions 1, 2, and 3 as effective meshes to obtain the multiple groups of effective meshes.

[0010] In an optional approach, before generating a continuous shooting path by traversing the adjacency relationships of each group of effective grids to obtain multiple shooting paths corresponding one-to-one with the multiple groups of effective grids, the method further includes: iteratively executing the following steps until the continuous coverage constraint is satisfied between adjacent grid rows and adjacent grid columns in each group of effective grids: taking adjacent columns as units, obtaining the minimum row coordinates and maximum row coordinates of the effective grids in two columns respectively to determine the comparison range of each row index level; within the comparison range, checking whether the effective grid exists in two columns row by row index level; when the effective grid exists in a column of a certain row index level, but the effective grid does not exist in the corresponding row index level of the adjacent column or in the row index level that differs from the level within a preset threshold range, supplementing the adjacent column with a new effective grid at the coordinate position corresponding to the missing row index level.

[0011] In one optional approach, determining the actual flight waypoint of the UAV at each shooting point based on the shooting path, the flight altitude, and the pitch angle to generate multiple flight paths corresponding one-to-one with the multiple shooting paths includes: for each shooting point on the shooting path, calculating the horizontal offset distance of the shooting point relative to a point directly below the UAV based on the flight altitude and the pitch angle; for each shooting path, moving the shooting point in the shooting path along the opposite direction of the horizontal axis of the flight path coordinate system corresponding to the shooting path by the horizontal offset distance to obtain the actual flight waypoint of the UAV at the shooting point.

[0012] In one alternative approach, after determining the actual flight waypoints of the UAV at each shooting point based on the shooting path, the flight altitude, and the pitch angle to generate multiple flight paths corresponding one-to-one with the multiple shooting paths, the method further includes: determining each of the actual flight waypoints as shooting waypoints and generating a task file, the task file being used to instruct the UAV to fly along the flight path and trigger the camera device to perform a single shot when it reaches the shooting waypoint; or, generating a task file based on the flight path, the task file being used to instruct the UAV to fly along the flight path and continuously record video.

[0013] According to another aspect of the embodiments of this application, a UAV oblique photography flight path planning device is provided, including a memory, a processor, and a computer program stored in the memory, wherein the processor executes the computer program to implement the UAV oblique photography flight path planning method as described above.

[0014] In this embodiment, after determining the first region corresponding to the target area in the geographic coordinate system, a second region corresponding to the first region is determined in each of the multiple flight path coordinate systems. A set of candidate grids is determined based on each second region, and effective grids are selected from each set of candidate grids to obtain multiple sets of effective grids that correspond one-to-one with the multiple flight path coordinate systems. Then, a flight path is reasonably planned based on each set of effective grids to obtain multiple flight paths that correspond one-to-one with the multiple flight path coordinate systems. This allows the camera device mounted on the UAV to take pictures of the target area from different directions when the UAV flies along each flight path, enriching the perspective of taking pictures of the target area and obtaining a full-range tilted image of the target area so that the target area can be accurately modeled in three dimensions based on these images.

[0015] The above description is merely an overview of the technical solutions of the embodiments of this application. In order to better understand the technical means of the embodiments of this application and to implement them in accordance with the contents of the specification, and to make the above and other objects, features and advantages of the embodiments of this application more obvious and understandable, specific implementation methods of this application are described below. Attached Figure Description

[0016] The accompanying drawings are for illustrative purposes only and are not intended to limit the scope of this application. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings: Figure 1 A flowchart illustrating the UAV oblique photography flight path planning method provided in an embodiment of this application is shown. Figure 2 This illustration shows a schematic diagram of the four endpoints corresponding to the four boundary points of the target region, provided in an embodiment of this application. Figure 3This application provides an embodiment of oblique photogrammetry projection geometry diagram. Figure 4 A schematic diagram of an effective mesh provided in an embodiment of this application is shown; Figure 5 A schematic diagram of the camera points, shooting waypoints, and flight routes provided in the embodiments of this application is shown; Figure 6 A schematic diagram of the structure of the UAV oblique photography flight path planning device provided in an embodiment of this application is shown. Detailed Implementation

[0017] Exemplary embodiments of the present application will now be described in more detail with reference to the accompanying drawings. Although exemplary embodiments of the present application are shown in the drawings, it should be understood that the present application may be implemented in various forms and should not be limited to the embodiments set forth herein.

[0018] Before creating a 3D model of a target area (such as a building), data about that area must first be collected to facilitate the 3D modeling process. Taking visual acquisition as an example, a camera is typically used to photograph the target area, and the resulting images are then used for 3D modeling. If the target area has a certain height, to ensure comprehensive image acquisition, a camera mounted on a drone can be used to photograph the area. By controlling the drone to fly along a flight path, the camera can capture images from different spatial positions, obtaining comprehensive images of the target area and ensuring the accuracy of the 3D modeling.

[0019] The above shooting methods are also known as oblique photography. Oblique photography is a type of photogrammetry technique that uses a camera device mounted on a flight platform (such as a drone) to acquire multi-angle images of a target area from various perspectives, such as forward view, backward view, side view, or fixed pitch angle. The resulting images are then used for aerial triangulation, dense matching, texture mapping, and 3D model reconstruction.

[0020] Understandably, the comprehensiveness of image acquisition of the target area directly determines the accuracy of 3D modeling. Since the camera device is mounted on a drone, the rationality of the drone's flight path planning becomes a key factor affecting the image coverage. Improper flight path planning may result in insufficient image acquisition, failing to fully cover the target area; on the other hand, it may generate a large number of highly similar redundant images, wasting computing and storage resources.

[0021] Moreover, in 3D modeling tasks, the quality of the final model does not depend solely on the drone's flight execution status, but is also constrained by the initial flight path planning and shooting pose design. If the flight path planning is unreasonable, even if mature modeling algorithms are used in the backend, the model may still have problems such as missing facade textures, broken dense point clouds, local model defacement, and holes.

[0022] Currently, drone flight path planning is mostly based on regular rectangular areas or simple parallel flight strips, typically based on the following assumptions: first, the target area boundary is relatively regular; second, the flight strip direction is approximately consistent with the main direction of the target area; third, waypoints can be directly used as drone shooting points; and fourth, there is no need to specifically consider ground projection offset caused by pitch imaging. However, for irregular target areas such as building complexes, industrial parks, mining areas, slopes, and road junctions, the above assumptions are often difficult to meet.

[0023] In summary, the currently planned drone flight paths may have the following problems: too many invalid shots outside the target area, leading to an increase in drone sorties; insufficient adaptability to the boundaries of irregular target areas, easily resulting in missed shots at the edges; failure to couple and model the horizontal field of view, flight altitude, and pitch angle of the camera device, resulting in unreasonable planned flight paths; failure to compensate for discontinuities between adjacent flight paths, resulting in unstable image connectivity; and lack of a unified spatial description between the ground shooting center and the actual drone position, making it difficult to directly form executable waypoints for engineering purposes.

[0024] To enable efficient planning of UAV flight paths, this application proposes a UAV oblique photography flight path planning method. First, the boundary information of the target area, the flight parameters of the UAV, and the imaging parameters of the camera device mounted on the UAV are acquired. Then, based on the boundary information, a first region corresponding to the target area is determined in a geographic coordinate system, converting the three-dimensional target area into a planar first region. Subsequently, the first region is transformed into regions under multiple flight path coordinate systems through coordinate mapping, resulting in multiple second regions. Next, the exposure interval and flight path interval are calculated based on parameter coupling, and a set of candidate grids covering each second region is generated based on the exposure interval and flight path interval. Finally, effective grids are selected from each set of candidate grids, and multiple shooting paths are determined based on the effective grids. Multiple flight paths are then generated based on these multiple shooting paths.

[0025] Figure 1This diagram illustrates a flowchart of a UAV oblique photogrammetry flight path planning method provided in an embodiment of this application. The method is executed by a UAV oblique photogrammetry flight path planning device, which can be a terminal device including one or more processors, such as a server, computer, touchscreen phone, smartphone, tablet computer, portable electronic device, or other electronic device. The processor may be a central processing unit (CPU), an application-specific integrated circuit (ASIC), or one or more integrated circuits configured to implement the embodiments of this application; no limitation is made here. The one or more processors included in the UAV oblique photogrammetry flight path planning device can be of the same type, such as one or more CPUs; or they can be of different types, such as one or more CPUs and one or more ASICs; no limitation is made here. Figure 1 As shown, the method includes the following steps: Step 110: Obtain the boundary information of the target area, the imaging parameters of the camera device on the UAV, and the flight parameters of the UAV.

[0026] The target area refers to the object that needs to be modeled in 3D, such as a building.

[0027] The boundary information of the target area is information in a geographic coordinate system. The geographic coordinate system can be a latitude and longitude coordinate system, a Universal Transverse Mercator (UTM) coordinate system, etc. This embodiment of the application uses a latitude and longitude coordinate system as an example. The boundary information obtained in this step includes the latitude and longitude coordinates of N boundary points of the target area. The boundary points of the target area refer to spatial location feature points selected around the actual physical contour edge of the target area that can accurately define its horizontal projection range and geometric shape, such as the outer corner of a building, the turning point of a convex or concave structure, and the extreme extension of the area's edge. N is a positive integer, and N≥3. After obtaining the latitude and longitude coordinates of the boundary points, the area corresponding to the target area can be determined in the latitude and longitude coordinate system based on the latitude and longitude coordinates of the boundary points.

[0028] Imaging parameters include at least the horizontal field of view, and flight parameters include at least the flight altitude, pitch angle, directional overlap, and lateral overlap.

[0029] Forward overlap, also known as longitudinal overlap, refers to the degree of overlap between two adjacent images taken along the same flight path in the direction of flight. Lateral overlap, also known as horizontal overlap, refers to the degree of overlap between two adjacent images along two adjacent flight paths in the direction perpendicular to flight. Forward overlap is fundamental to ensuring data quality. Only when the forward overlap meets the requirements can sufficient corresponding feature points be provided for subsequent 3D time modeling based on images, thereby generating continuous and dense point cloud data, avoiding model breaks, deformations, or holes, and ultimately constructing a high-precision and highly complete 3D model. However, if only the forward overlap meets the requirements, while the lateral overlap does not, although the model quality within a single flight path can be guaranteed, it will lead to a lack of sufficient overlap between adjacent flight paths. This will cause cross-flight image matching to fail, resulting in the inability to achieve robust spatial stitching of multi-flight data, ultimately leading to structural misalignment, texture tearing, and pervasive data blind spots at the boundaries of the overall 3D model. Different modeling software has different requirements for forward and lateral overlap when using images for 3D modeling.

[0030] Three-dimensional reconstruction software based on images typically has specific value ranges for the forward and lateral overlap rates. Therefore, if a certain three-dimensional reconstruction software is needed for subsequent three-dimensional reconstruction, the forward overlap rate and lateral overlap rate can be determined based on the three-dimensional reconstruction software. After determining the forward overlap rate and lateral overlap rate, they can be provided to the device executing the embodiments of this application.

[0031] Step 120: Based on the boundary information, determine the first region corresponding to the target region in the geographic coordinate system.

[0032] Specifically, in this step, the first region can be determined through the following steps a1 to a4.

[0033] Step a1: Based on the latitude and longitude coordinates of the N boundary points, determine the N endpoints that correspond one-to-one with the N boundary points in the geographic coordinate system.

[0034] Specifically, after obtaining the latitude and longitude coordinates of N boundary points and establishing a latitude and longitude coordinate system, the endpoints with the same latitude and longitude as the boundary points can be determined under this coordinate system.

[0035] Step a2: Determine the geometric center of the N endpoints to obtain the center point.

[0036] Here, we take N=4 as an example. Figure 2 This diagram illustrates the four endpoints corresponding to the four boundary points of the target region provided in an embodiment of this application. For example... Figure 2 As shown, the four endpoints are endpoint A, endpoint B, endpoint C and endpoint D, and the center point is point E.

[0037] Step a3: Determine the polar angle of the line connecting each endpoint to the center point to obtain N polar angles that correspond one-to-one with the N endpoints.

[0038] like Figure 2 As shown, after determining multiple endpoints, different methods of connecting the endpoints will result in different boundaries for the first region. For example, the boundary of the target region and... Figure 2 The boundaries shown in (a) are consistent. If the endpoints are connected in the wrong order, the final geometry may look like this. Figure 2 The region shown in (b) cannot accurately fit the actual boundary contour of the target region. Therefore, in order to connect the endpoints in the correct topological order and construct a polygonal boundary that matches the true contour of the target region, in this step, the polar angle of the line connecting each endpoint to the center point is first determined, and then the endpoints can be connected based on the polar angle in subsequent steps.

[0039] In this step, when determining the polar angle of the line connecting each endpoint to the center point, the center point is used as the pole of the polar coordinate system, and a two-dimensional rectangular coordinate system is established with the pole as the origin. A specific coordinate axis (such as the coordinate axis containing the due east or due north directions) is set as the reference starting direction of the polar axis. Then, for each endpoint, the two-dimensional coordinate difference (i.e., the difference between the horizontal and vertical coordinates) relative to the pole is calculated. Based on the coordinate difference, the absolute angle value corresponding to the quadrant where each endpoint is located is solved using inverse trigonometric functions (such as the arctangent function). Then, combined with the quadrant where it is located, it is converted into a polar angle measured with the polar axis as the starting direction and along a predetermined rotation direction (such as clockwise or counterclockwise).

[0040] Step a4: Connect the N endpoints in order of the magnitude of the N polar angles to obtain the first region.

[0041] In this step, specifically, the N polar angles calculated in the previous steps can be sorted in ascending or descending order to obtain an ordered polar angle sequence. Then, the N endpoints corresponding to the polar angle sequence are rearranged. Finally, according to the rearranged sequence, adjacent endpoints are connected in pairs to form continuous edge segments, and a closed connection is made between the first and last endpoints to obtain the first region that is consistent with the boundary of the target region.

[0042] For example, Figure 2In (a) of the diagram, the polar angle corresponding to endpoint A is -135°, endpoint B is -45°, endpoint C is 45°, and endpoint D is 135°. Connecting the four endpoints in ascending or descending order of polar angle will yield the rectangle ABCD (i.e., the first region) shown in the diagram. This method effectively avoids the problem of graphic self-intersection or distortion caused by disordered connection of endpoints, ensuring that the boundary contour of the first region conforms to the geometric characteristics of the target region.

[0043] Step 130: For multiple route coordinate systems, determine the second region corresponding to the first region under each route coordinate system, thus obtaining multiple second regions that correspond one-to-one with the multiple route coordinate systems.

[0044] Before performing this step, M route coordinate systems can be determined through the following steps b1 to b2. Here, M is a positive integer, and M ≥ 2.

[0045] Step b1: Determine the directions of M flight routes.

[0046] Understandably, when photographing a target area, the more diverse the shooting angles, the more complete the resulting image, and the higher the accuracy of the 3D model constructed from that image. Therefore, to avoid the limitations of a single flight path perspective and to eliminate blind spots in the target area (especially building facades and complex structural blind spots) as much as possible, thereby obtaining omnidirectional images, this step determines M flight path directions, where M ≥ 2, and the angle between any two of the M flight path directions is an integer multiple of 180° / M.

[0047] If the value of M is too large, it will lead to an increase in the number of planned flight paths. While this helps to collect richer images to improve the accuracy of 3D modeling, it will significantly increase the operation time and computational resource consumption. Therefore, to balance the accuracy and efficiency of 3D modeling, in some embodiments, M is set to 4. In order to determine the M flight path directions, the flight parameters obtained in step 110 also include the reference flight path angle θ. Accordingly, the flight path directions determined in step b1 are θ, θ+90°, θ+180°, and θ+270°. The reference flight path angle θ can be determined based on the overall tilt angle of the target area. For example, the overall tilt angle of the target area can be directly used as the reference flight path angle. By determining four different and complementary flight path directions, after the four flight paths are generated based on the four headings, when the UAV flies along the four flight paths, its onboard camera device can collect data from the target area from different directions, thereby collecting images of the various sides and top edges of the target area.

[0048] Step b2: Establish a coordinate system with each route direction as the horizontal axis to obtain M route coordinate systems that correspond one-to-one with the M route directions.

[0049] In this step, by establishing an independent route coordinate system for each route direction, the absolute positional relationship of the target area under a unified geographic coordinate system can be transformed to each route coordinate system, thereby enabling route planning under each route coordinate system.

[0050] After determining the route coordinate system through the above steps, in step 130, by converting the first region into regions under each route coordinate system, M second regions corresponding one-to-one with the M route coordinate systems can be obtained.

[0051] Specifically, through coordinate system rotation and translation transformations, the coordinates of each boundary point of the first region in the original geographic coordinate system are mapped one by one to local coordinates in each flight path coordinate system, thereby reconstructing the corresponding second region in each flight path coordinate system. Since the horizontal axis of each flight path coordinate system is parallel to the corresponding flight path direction, the spatial attitude of each second region obtained by the transformation can intuitively reflect the relative orientation of the target region with respect to the current flight path direction, enabling subsequent reasonable planning of the UAV's flight path based on the second region.

[0052] Step 140: Determine the exposure interval and flight strip interval based on the imaging parameters and flight parameters.

[0053] In this embodiment, the flight parameters obtained in step 110 include the UAV's flight altitude h, pitch angle, lateral overlap ratio Rf, and lateral overlap ratio Rs. The imaging parameters obtained in step 110 include the horizontal field of view. The image resolution of the image captured by the camera device determines the vertical field of view, which can be derived from the aspect ratio of the image resolution. . Figure 3 A geometric relationship diagram of oblique photogrammetry projection provided in an embodiment of this application is shown. For example... Figure 3 As shown, the UAV is located at point F, and the corresponding projection center is point G. The horizontal offset distance of the projection center relative to the UAV is d. Lines FH and FI are the pitch imaging center lines of the camera device mounted on the UAV, and the horizontal coverage width of the ground is W. The exposure spacing Sf and the flight strip spacing Ss can be calculated using the following formulas (1) to (5).

[0054] (1) (2) (3) (4) (5) Where L is the heading coverage length.

[0055] Step 150: Based on the exposure spacing and flight strip spacing, generate a set of candidate grids covering each second region, resulting in multiple sets of candidate grids that correspond one-to-one with the multiple second regions.

[0056] Specifically, in this step, multiple sets of candidate meshes can be generated through the following steps c1 to c2.

[0057] Step c1: In each route coordinate system, generate the minimum bounding polygon covering the second region, and obtain multiple minimum bounding polygons that correspond one-to-one with multiple second regions.

[0058] In this step, a minimum bounding polygon is generated so that the flight path can be reasonably planned based on this minimum bounding polygon in the future. The minimum bounding polygon can be a rectangle or a rhombus, etc.

[0059] Step c2: Within each minimum bounding polygon, candidate grid center points are evenly laid out in rows and columns with the exposure spacing as the horizontal axis step size and the flight strip spacing as the vertical axis step size, and candidate grids are generated based on the candidate grid center points.

[0060] Specifically, starting from any endpoint of the smallest bounding polygon or the origin of the heading coordinate system, the pre-calculated exposure spacing is set as a single-step advance distance along the horizontal axis (X-axis), and the flight strip spacing is set as a single-step advance distance along the vertical axis (Y-axis). Then, using a double loop traversal, the horizontal and vertical step sizes are accumulated row by row and column by column within the smallest bounding polygon, thus calculating all planar coordinate points that meet the conditions as candidate mesh center points. Finally, each center point is used as the geometric center to determine and generate the candidate mesh. The candidate mesh can be a rectangular mesh or a rhomboid mesh. Taking a rectangular mesh as an example, after determining the center point of the candidate mesh, the exposure spacing and flight strip spacing are used as the length and width of the rectangular mesh, respectively, to determine and generate the candidate mesh.

[0061] Step 160: For each group of candidate grids, perform effective grid filtering based on the second region to obtain multiple groups of effective grids that correspond one-to-one with the multiple groups of candidate grids.

[0062] Since the candidate mesh is generated within the smallest bounding polygon with a fixed step size, it will inevitably extend outward beyond the second area that actually needs to be photographed. Therefore, it is necessary to eliminate redundant and invalid meshes through geometric positional relationships so that the flight path can be reasonably planned based on the effective meshes.

[0063] If a candidate mesh falls within the second region, it indicates that it overlaps with the second region. This ensures that when the camera on the UAV performs its imaging task at the location corresponding to the mesh, the camera's field of view can effectively cover the target area, thus guaranteeing that the captured image is valid data for subsequent 3D reconstruction. Therefore, if at least one corner point of the candidate mesh is located within the second region, or if the center point of the candidate mesh is located within the second region, it indicates that it overlaps with the second region, and the candidate mesh is then determined to be a valid mesh and retained.

[0064] If at least one endpoint of the second region is located within the enclosed area of ​​the candidate grid, it means that although the center and edge of the candidate grid may be outside the second region, when the camera device performs the shooting task at the position corresponding to the grid, the field of view of the camera device can still capture the key boundary features of the target region, thereby ensuring that the captured image data is effective data for subsequent 3D reconstruction.

[0065] Based on this, in step 160, different conditions can be set to determine whether a candidate mesh falls within the second region, or whether at least one endpoint of the second region is within the enclosing area of ​​the candidate mesh, thus determining that the candidate mesh is a valid mesh and is retained. Specifically, by judging one by one whether the candidate meshes in each group meet any of the following conditions one to three, the candidate meshes that meet any condition are retained, and all the retained meshes in a group of candidate meshes are the group of valid meshes corresponding to that group of candidate meshes. Among them, condition one is that at least one corner point of the candidate mesh is located inside the second region; condition two is that the center point of the candidate mesh is located inside the second region; and condition three is that at least one endpoint of the second region is located inside the enclosing area of ​​the candidate mesh.

[0066] By using the three complementary judgment conditions mentioned above, all grids that substantially intersect with the second region can be screened out more comprehensively. Subsequently, flight paths can be rationally planned based on the effective grids so that when the UAV flies along the flight path, the target area can be fully captured by the camera device mounted on the UAV to the greatest extent.

[0067] Step 170: For each group of valid grids, generate a continuous shooting path by traversing the adjacency relationship of the valid grids, and obtain multiple shooting paths that correspond one-to-one with multiple groups of valid grids.

[0068] Specifically, the adjacency relationships of valid grids can be traversed based on a preset traversal strategy. For example, for a set of valid grids, the corner point of the leftmost edge grid can be selected as the starting point of the path. The path is traversed row by row or column by column in an ordered manner within the matrix structure formed by the valid grids according to a preset direction of travel (e.g., a bow-shaped or serpentine path). During the traversal, the next adjacent grid is directly obtained along the current direction of travel and sequentially added to the path sequence. When the path extends to the boundary of the current row or column and it is no longer possible to obtain adjacent grids in the same direction, it automatically jumps to the edge grid of the next row or column, which serves as the path's turning point. This process continues until all valid grids in the set have been traversed and added to the sequence, thereby connecting the points of the valid grids end to end to obtain a continuous shooting path without any breaks.

[0069] Step 180: Based on the shooting path, flight altitude, and pitch angle, determine the actual flight waypoints of the UAV at each shooting point to generate multiple flight paths that correspond one-to-one with the multiple shooting paths.

[0070] Specifically, step 180 can be achieved through the following steps d1 to d2.

[0071] Step d1: For each shooting point on the shooting path, calculate the horizontal offset distance of the shooting point relative to the point directly below the drone, based on the flight altitude and pitch angle.

[0072] The shooting point is the center point of the effective grid. In this step, the horizontal offset distance d can be calculated using the above formula (1).

[0073] Step d2: For each shooting path, move the shooting point in the shooting path in the opposite direction of the horizontal axis of the corresponding flight path coordinate system by a horizontal offset distance to obtain the actual flight waypoint of the UAV at the shooting point.

[0074] Since each shooting path is determined based on a set of effective grids, and a set of effective grids corresponds to a flight path coordinate system, and as described in steps b1 to b2 above, the horizontal axis of each flight path coordinate system is determined based on a heading, that is, a flight path coordinate system corresponds to a heading. Therefore, in this step, after determining the horizontal offset distance d based on the above formula (1), for each shooting path, after moving the shooting point in the shooting path in the opposite direction of the heading corresponding to the shooting path by the horizontal offset distance d, the actual flight waypoint corresponding to the shooting point in the shooting path can be obtained.

[0075] Figure 4 A schematic diagram of an effective mesh provided in an embodiment of this application is shown. For example... Figure 4As shown in the figure, the positional relationship between the effective grid and the first region is illustrated in the latitude and longitude coordinate system. In other words, the positional relationship between the effective grid and the first region is illustrated after the effective grid in a certain route coordinate system is converted to the latitude and longitude coordinate system.

[0076] Figure 5 A schematic diagram illustrating the camera points, shooting waypoints, and flight routes provided in an embodiment of this application is shown. For example... Figure 5 As shown in the figure, the positional relationship between the actual flight waypoints and the flight path is illustrated in the latitude and longitude coordinate system. In other words, after converting the effective grid in a certain flight path coordinate system to the latitude and longitude coordinate system, the positional relationship between the shooting point, the actual flight waypoints, the flight path, and the first area is shown. Figure 5 In the diagram, the path formed by connecting all the shooting points with dashed lines is a shooting path, and the path formed by connecting all the actual flight waypoints with solid lines is a corresponding flight route. Actual flight waypoint P1 is the starting point of the flight route, and actual flight waypoint P2 is the ending point. Taking two sets of points as an example... Figure 5 In this process, the actual flight waypoint corresponding to the shooting point P11 determined in this step is P21, and the actual flight waypoint corresponding to the shooting point P12 is P22.

[0077] In this embodiment, after determining the first region corresponding to the target area in a geographic coordinate system, a second region corresponding to the first region is determined in each of the multiple flight path coordinate systems. A set of candidate grids is determined based on each second region, and valid grids are selected from each set of candidate grids, resulting in multiple sets of valid grids corresponding one-to-one with the multiple flight path coordinate systems. Then, a flight path is rationally planned based on each set of valid grids, resulting in multiple flight paths corresponding one-to-one with the multiple flight path coordinate systems. This allows the camera mounted on the UAV to capture images of the target area from different directions when the UAV flies along each flight path, enriching the viewing angles of the target area and obtaining a comprehensive oblique image of the target area. This allows for accurate 3D modeling of the target area based on these images. Furthermore, this application's embodiments determine the exposure spacing and flight strip spacing based on the imaging geometry and overlap rate indices (i.e., forward overlap rate and lateral overlap rate) of the UAV's wide-angle camera device. This fully considers the UAV's flight altitude, the camera device's horizontal field of view, the image ratio of the images captured by the camera device, and the UAV's pitch angle. Compared to manually setting the flight strip spacing based on experience, this approach is more technically sound and repeatable. By employing a "local coordinate transformation + flight path rotation alignment" method, navigation can be planned for target areas of any polygon. Moreover, by setting up a four-way joint supplementary acquisition mechanism, the elevation information of the target area in different directions can be mutually supplemented and acquired. Furthermore, the method of filtering effective grids based on grid corners, grid centers, and second region endpoints is more suitable for planning flight paths for target areas with narrow edges and irregular sharp corners compared to filtering effective grids using a single judgment method.

[0078] After planning the drone's flight path, in order to collect images of the target area, the camera device mounted on the drone can be controlled to record video of the target area. Specifically, in Figure 1 Based on the provided embodiments, after step 180, the method further includes the following step e1.

[0079] Step e1: Generate a mission file based on the flight path. The mission file is used to make the UAV fly along the flight path and continuously record video.

[0080] Specifically, the mission file encapsulates the flight path and corresponding flight control commands, including parameters such as flight speed, altitude, heading angle, and camera pitch angle. During the execution of this mission file, the UAV's flight control system is linked with the camera device, allowing the UAV to continuously trigger the camera to enter recording mode while flying along the flight path. This acquires a continuous video stream covering the entire target area with stable frame overlap, enabling rapid and coherent acquisition of target area image data without increasing latency from frequent photo captures or attitude disturbances.

[0081] After capturing video of the target area using a camera device, images need to be extracted from the video for 3D modeling, which takes time and reduces the efficiency of 3D modeling. Therefore, to improve the efficiency of 3D modeling, in some embodiments, the camera device mounted on the drone can be controlled to capture images of the target area instead of continuously recording video. Specifically, in Figure 1 Based on the provided embodiments, after step 180, the method further includes the following step f1.

[0082] Step f1: Determine each actual flight waypoint as the shooting waypoint and generate a task file. The task file is used to make the UAV fly along the flight path and trigger the camera device to take a single shot when it reaches the shooting waypoint.

[0083] like Figure 5 As shown, in this step, the actual flight waypoints corresponding to the shooting points are determined as shooting waypoints. Specifically, after each actual flight waypoint is determined as a shooting waypoint, a task file including the flight path and control commands is generated, so that when the UAV executes the task file and flies along the flight path, it can take pictures of the target area at the corresponding shooting waypoint, thereby acquiring images of the target area.

[0084] It is worth noting that, for the target area of ​​the alien, through Figure 1 The effective grids selected in the illustrated embodiment may have discontinuities. Planning flight paths based on such discontinuous effective grids may prevent the UAV from effectively covering and acquiring images of the areas corresponding to these discontinuous effective grids during flight photography. To ensure the effective grids are continuous, in Figure 1 Based on the provided embodiments, in some embodiments, prior to step 170, the UAV oblique photography flight path planning method further iteratively executes the following step g1 until the continuous coverage constraint is satisfied between adjacent grid rows and adjacent grid columns in each group of valid grids.

[0085] Step g1: Using adjacent columns as units, obtain the minimum and maximum row coordinates of the valid grid in each column to determine the comparison range of each row index level; within the comparison range, check whether there is a valid grid in each column for each row index level; when a row index level has a valid grid in one column, but there is no valid grid in the corresponding row index level of the adjacent column or in the row index level that differs from this level within a preset threshold range, add a new valid grid to the coordinate position corresponding to the missing row index level in the adjacent column.

[0086] Because the aforementioned generation and selection of candidate meshes in various route coordinate systems only considered the containment relationship between a single mesh and the target area boundary, without taking into account the topological adjacency between meshes, the selected effective mesh set is prone to discontinuities or isolated phenomena in the arrangement of irregular areas with depressions, narrow corners, or complex fragments. Therefore, by iteratively executing step g1, a two-way spatial interpolation constraint based on rows (along the route extension direction) and columns (along the route interval direction) is essentially introduced in the local coordinate system: using two adjacent columns of meshes as the basic detection unit, the vertical boundary (i.e., the comparison range) of the actual distribution of the two columns of meshes is defined using the minimum and maximum row coordinates, and then a row-by-row scan is performed within this boundary; once a column is found to have an effective mesh at a specific row level, while the adjacent column has a mesh gap, a new effective mesh is immediately added at the missing position. Although the center of this supplementary grid may not fall exactly inside the original polygon boundary of the target area, it can fill the image coverage gap between adjacent effective grids, thus physically ensuring that when the UAV flies level along the flight path generated by the effective grid, the images captured by the camera device can cover the target area as comprehensively as possible.

[0087] The embodiments of this application will be described below with reference to experimental data. In this experiment, the target area is a building with a triangular cross-section (referred to as a triangular building). This triangular building corresponds to three boundary points. The reference heading is 30°. The image resolution of the image captured by the wide-angle camera mounted on the UAV is 4032×3024, the aspect ratio is 4:3, and the horizontal field of view of the camera is 84.00°. Based on this, the vertical field of view is approximately 68.06°. The UAV's flight altitude is 50m, and the pitch angle is -45°. In the 3D modeling scene, the heading overlap rate is taken as 80.00%, and the lateral overlap rate is taken as 70.00%. According to the embodiments of this application, the exposure spacing along the flight path is approximately 13.51m, and the flight strip spacing perpendicular to the flight path is approximately 27.01m. Around the reference heading, this experiment simultaneously generates supplementary headings in four directions: 30°, 120°, 210°, and 300°, to form a multi-directional joint acquisition scheme.

[0088] For a heading of 30°: there are 35 effective grid center points, 33 waypoints for shooting, an exposure interval of approximately 13.51 m, and a flight strip interval of approximately 27.01 m.

[0089] For a heading of 120°: there are 35 effective grid center points, 35 waypoints for shooting, an exposure interval of approximately 13.51m, and a flight strip interval of approximately 27.01m.

[0090] For a heading of 210°: there are 34 effective grid center points, 34 waypoints for shooting, an exposure interval of approximately 13.51m, and a flight strip interval of approximately 27.01m.

[0091] For a heading of 300°: there are 37 effective grid center points, 37 waypoints for shooting, an exposure interval of approximately 13.51m, and a flight strip interval of approximately 27.01m.

[0092] To illustrate the difference between the flight path planned using the UAV oblique photogrammetry flight path planning method provided in this application and the manually planned flight path, under the same mission area and flight parameters, the single-direction planning result of this application is compared with the manually empirical bounded rectangle parallel flight path scheme. The manually empirical scheme uses a fixed flight path direction and directly deploys shooting points based on the bounded rectangle of the mission area, without performing polygon boundary joint overlap screening or point connection processing. The comparison results are shown in Table 1 below.

[0093] Table 1: Differences between flight paths planned using the UAV oblique photogrammetry flight path planning method provided in this application and manually planned flight paths

[0094] Furthermore, if the manual experience-based approach adopts a simplified cropping method of "only retaining the center point located within the task area," then the number of shooting points retained is 25, which is 8 fewer than the boundary transition area coverage points in this application. Therefore, it can be seen that this application can balance boundary area coverage and path continuity while controlling invalid shooting.

[0095] The above comparison is only used to illustrate the differences between the two planning methods under the same area and parameter conditions, and is not intended to limit the scope of protection. As can be seen from the above embodiments, this application can automatically form four sets of executable coverage paths around the same target area. The four directions provide supplementary coverage from different perspectives, and their joint execution helps to improve facade visibility, edge texture integrity, and the 3D reconstruction quality of corner areas.

[0096] Furthermore, the open-source UAV image processing platform (OpenDroneMap, ODM) was used to acquire images and perform 3D reconstruction of the target area mentioned in the above experiment. According to the modeling log statistics, the modeling software version was 3.5.6, the number of input images was 292, and all 292 images were used for reconstruction, with a total modeling time of approximately 36.32 minutes. During the reconstruction process, 274,249 effective trajectories, 260,880 sparse reconstruction points, and 38,469,908 dense point clouds were obtained, with an estimated point cloud spacing of approximately 0.03 m. Finally, a textured 3D model and orthorectified results were generated. The ODM modeling result indicators and values ​​can be found in Table 2 below.

[0097] Table 2: ODM Modeling Result Indicators and Values

[0098] In this experiment, textured glTF binary model files, georeferenced point clouds, and orthorectification results were also output. The orthorectification results were output in a size of approximately 9320×8037 pixels, demonstrating that the images captured by the camera device along the planned flight path in this embodiment of the application, and the UAV flying along that path, can support subsequent 3D modeling and orthorectification.

[0099] Figure 6 This illustration shows a structural schematic diagram of the UAV oblique photography flight path planning device provided in an embodiment of this application. The specific embodiments of this application do not limit the specific implementation of the UAV oblique photography flight path planning device. Figure 6 As shown, the UAV oblique photography flight path planning device 200 may include a processor 202 and a memory 204.

[0100] The memory 204 is used to store the computer program 206. The memory 204 may include high-speed RAM, and may also include non-volatile memory, such as at least one disk drive. The computer program 206 may include computer-executable instructions.

[0101] The processor 202 is used to execute the computer program 206 to implement the above-described embodiment of the UAV oblique photography flight path planning method.

[0102] The processor 202 may be a central processing unit (CPU), an application-specific integrated circuit (ASIC), or one or more integrated circuits configured to implement the embodiments of this application. The one or more processors included in the UAV oblique photography flight path planning device 200 may be processors of the same type, such as one or more CPUs; or they may be processors of different types, such as one or more CPUs and one or more ASICs.

[0103] This application provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described UAV oblique photography flight path planning method embodiment.

[0104] This application provides a computer program that can be executed by a processor to implement the above-described UAV oblique photography flight path planning method embodiment.

[0105] This application provides a computer program product, which includes a computer program that, when executed by a processor, implements the above-described UAV oblique photography flight path planning method embodiment.

[0106] In the several embodiments provided in this application, any function, if implemented as a software functional module / unit and sold or used as an independent product, can be stored in a computer-readable storage medium. Based on this understanding, all or part of the technical solution of this application can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or other electronic device) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing computer program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0107] The algorithms or displays provided herein are not inherently related to any particular computer, virtual system, or other device. Various general-purpose systems can also be used in conjunction with the teachings herein. The required structure for constructing such systems is apparent from the above description. Furthermore, the embodiments of this application are not directed to any particular programming language. It should be understood that the content of this application described herein can be implemented using various programming languages, and the above description of specific languages ​​is for the purpose of disclosing the best mode of implementation of this application.

[0108] It should be noted that the above embodiments are illustrative of this application and not restrictive, and those skilled in the art can devise alternative embodiments without departing from the scope of the appended claims. In the claims, any reference signs placed between parentheses should not be construed as limiting the claims. The word "comprising" does not exclude the presence of elements or steps not listed in the claims. The word "a" or "an" preceding an element does not exclude the presence of a plurality of such elements. This application can be implemented by means of hardware comprising several different elements and by means of a suitably programmed computer. In claims enumerating several means, several units or modules of these means may be embodied by the same item of hardware. The use of the words first, second, and third, etc., does not indicate any order. These words can be interpreted as names. The steps in the above embodiments, unless otherwise specified, should not be construed as limiting the order of execution.

[0109] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A method for planning a UAV aerial photography route, characterized in that, The method includes: The boundary information of the target area, the imaging parameters of the camera device carried by the UAV, and the flight parameters of the UAV are acquired. The imaging parameters include at least the horizontal field of view, and the flight parameters include at least the flight altitude, pitch angle, heading overlap rate, and lateral overlap rate. Based on the boundary information, a first region corresponding to the target region is determined in a geographic coordinate system; For multiple route coordinate systems, a second region corresponding to the first region is determined under each route coordinate system, resulting in multiple second regions that correspond one-to-one with the multiple route coordinate systems. The exposure interval and flight strip interval are determined based on the imaging parameters and the flight parameters. Based on the exposure spacing and the flight strip spacing, a set of candidate grids covering each of the second regions is generated, resulting in multiple sets of candidate grids that correspond one-to-one with the multiple second regions. For each group of candidate grids, effective grids are filtered based on the second region to obtain multiple groups of effective grids that correspond one-to-one with the multiple groups of candidate grids; For each group of effective grids, a continuous shooting path is generated by traversing the adjacency relationship of the effective grids, resulting in multiple shooting paths corresponding one-to-one with the multiple groups of effective grids; Based on the shooting path, the flight altitude, and the pitch angle, the actual flight waypoint of the UAV at each shooting point is determined to generate multiple flight routes that correspond one-to-one with the multiple shooting paths.

2. The method according to claim 1, characterized in that, The acquisition of the boundary information of the target area includes: Obtain the latitude and longitude coordinates of N boundary points of the target area, where N ≥ 3; The step of determining the first region corresponding to the target region in a geographic coordinate system based on the boundary information includes: Based on the latitude and longitude coordinates of the N boundary points, determine N endpoints that correspond one-to-one with the N boundary points in the geographic coordinate system; Determine the geometric centers of the N endpoints to obtain the center point; Determine the polar angle of the line connecting each endpoint to the center point to obtain N polar angles that correspond one-to-one with the N endpoints; The first region is obtained by connecting the N endpoints in order of the magnitude of the N polar angles.

3. The method according to claim 1, characterized in that, Before determining the second region corresponding to the first region in each of the multiple route coordinate systems to obtain multiple second regions corresponding one-to-one with the multiple route coordinate systems, the method further includes: Determine M flight path directions, wherein the angle between any two of the M flight path directions is an integer multiple of 180° / M, where M is an integer greater than 1; Establish a coordinate system with each of the flight directions as the horizontal axis to obtain M flight direction coordinate systems that correspond one-to-one with the M flight directions; For multiple route coordinate systems, a second region corresponding to the first region is determined in each route coordinate system, resulting in multiple second regions that correspond one-to-one with the multiple route coordinate systems, including: The first region is converted into regions under each of the M route coordinate systems to obtain M second regions that correspond one-to-one with the M route coordinate systems.

4. The method according to claim 3, characterized in that, The M route directions are determined based on the reference route angle θ, M=4, and the four route directions are θ, θ+90°, θ+180° and θ+270° respectively.

5. The method according to claim 1, characterized in that, The step involves generating a set of candidate grids covering each of the second regions based on the exposure spacing and the flight strip spacing, resulting in multiple sets of candidate grids corresponding one-to-one with the plurality of second regions, including: In each of the aforementioned route coordinate systems, a minimum bounding polygon covering the second region is generated, resulting in multiple minimum bounding polygons that correspond one-to-one with the multiple second regions. Within each of the minimum bounding polygons, candidate grid center points are evenly laid out in a row and column manner with the exposure spacing as the horizontal axis step size and the flight strip spacing as the vertical axis step size, and the candidate grid is generated based on the candidate grid center points.

6. The method according to claim 1, characterized in that, For each group of candidate grids, effective grid filtering is performed based on the second region to obtain multiple groups of effective grids that correspond one-to-one with the multiple groups of candidate grids, including: Determine whether the candidate mesh satisfies any of the following conditions: Condition 1: At least one corner point of the candidate mesh is located inside the second region; Condition 2: The center point of the candidate grid is located inside the second region; Condition 3: At least one endpoint of the second region is located within the enclosing area of ​​the candidate grid; Candidate meshes that satisfy any one of the conditions one, two, and three are determined as valid meshes, thus obtaining the multiple sets of valid meshes.

7. The method according to claim 1, characterized in that, Before generating a continuous shooting path by traversing the adjacency relationships of each group of effective grids to obtain multiple shooting paths corresponding one-to-one with the multiple groups of effective grids, the method further includes: Iteratively execute the following steps until the continuous coverage constraint is satisfied between adjacent grid rows and adjacent grid columns in each group of valid grids: Using adjacent columns as units, the minimum and maximum row coordinates of the valid grids in the two columns are obtained respectively to determine the comparison range of each row index level; within the comparison range, the validity of the valid grids in the two columns is checked row by row index level; when the valid grids exist in a column of a certain row index level, but the valid grids do not exist in the corresponding row index level of the adjacent column or in the row index level that differs from the level within a preset threshold range, a new valid grid is added to the coordinate position corresponding to the missing row index level in the adjacent column.

8. The method according to claim 1, characterized in that, The step of determining the actual flight waypoints of the UAV at each shooting point based on the shooting path, the flight altitude, and the pitch angle, to generate multiple flight paths corresponding one-to-one with the multiple shooting paths, includes: For each shooting point on the shooting path, the horizontal offset distance of the shooting point relative to the point directly below the drone is calculated based on the flight altitude and the pitch angle. For each shooting path, the shooting point in the shooting path is moved by the horizontal offset distance in the opposite direction of the horizontal axis of the flight path coordinate system to obtain the actual flight waypoint of the UAV at the shooting point.

9. The method according to claim 1, characterized in that, After determining the actual flight waypoints of the UAV at each shooting point based on the shooting path, the flight altitude, and the pitch angle to generate multiple flight paths corresponding one-to-one with the multiple shooting paths, the method further includes: Each actual flight waypoint is designated as a shooting waypoint, and a task file is generated. This task file directs the UAV to fly along the flight path and triggers the camera device to take a single shot upon reaching the shooting waypoint; or... A mission file is generated based on the flight path, and the mission file is used to make the UAV fly along the flight path and continuously record video.

10. A UAV oblique photography flight path planning device, comprising a memory, a processor, and a computer program stored in the memory, characterized in that, The processor executes the computer program to implement the UAV oblique photography flight path planning method according to any one of claims 1 to 9.

Images