Aerial photography view angle planning method and device based on optimal vision photogrammetry

Abstract

The embodiment of the invention provides an aerial photography view angle planning method based on optimal vision photogrammetry, and the method comprises the steps: determining a viewpoint sampling space of an unmanned plane based on a 3D general model of a target scene; then Poisson disk sampling is carried out on the surface of the 3D general model, and a sampling point set and a pre-selected viewpoint set corresponding to the sampling point set are generated; based on the sampling point set, carrying out reconstructability modeling on the 3D general model to generate reconstructability of each sampling point; on the basis of the pre-selected viewpoint set, viewpoint redundancy modeling is carried out, and the redundancy of each pre-selected viewpoint is generated; and based on the reconstructability of the sampling points and the redundancy of the pre-selected viewpoints, performing optimization processing on the pre-selected viewpoint set to generate an aerial photography view angle. According to the embodiment of the invention, the maximum sample point reconstructability and the minimum viewpoint redundancy are combined for the optimization processing of the pre-selected viewpoint set of the target scene, so that the precision of the three-dimensional reconstruction of the target scene is improved while the automatic aerial photography planning of the unmanned aerial vehicle is realized for the target scene.

Description

Technical Field

[0001] The present invention belongs to the field of image processing technology, and in particular relates to a method and device for aerial photography viewing angle planning based on optimal view photogrammetry. Background Art

[0002] The quality and speed of most multi-view stereo (MVS) algorithms depend heavily on the selection of input images. Reconstructing an object's surface requires sufficient views, but not all input images contribute equally to the reconstruction quality. Excessive images can lead to unnecessary processing time. Furthermore, incorrect viewpoints or scale differences can introduce noise into structure-from-motion and multi-view stereo matching, potentially degrading reconstruction quality. Therefore, view planning significantly impacts the quality of multi-view stereo reconstruction, typically requiring images from multiple viewpoints with different positions and orientations. Modern techniques often employ a coarse-to-fine approach. This involves first acquiring image data through traditional methods such as initial flights or satellite imagery. A rough initial model (scene proxy) is generated based on the 3D geometry of the object to be reconstructed. This initial model can be existing 3D scene data with high-level information or a low-precision model obtained from an initial flight. Based on this initial model, the optimal viewpoints for dense view sampling are then calculated. This allows for the design and optimization of drone aerial photography routes for the scene under test. Finally, image acquisition is performed and a detailed 3D model is reconstructed.

[0003] Existing large-scale urban scene reconstruction often requires advance aerial photography data collection. Before collecting data for urban 3D reconstruction, drone operators need to plan a flight route for the target area. The core of drone route planning is to predict whether a set of perspectives can effectively reconstruct the target area. The quality of the planning directly determines the effectiveness of the model ultimately reconstructed from the resulting aerial images. However, the flight time of conventional quadrotor drones is usually limited to 25-30 minutes by battery capacity. Therefore, the main goal of drone trajectory planning is to capture the entire 3D scene better and faster, that is, to maximize reconstruction quality while minimizing flight time. The viewpoint and path planning problem based on drones can be solved in two steps: 1) Generate the optimal shooting viewpoint to maximize reconstruction quality without considering flight time and path quality; 2) Minimize the flight cost of passing through all viewpoints generated in the first step.

[0004] The first category of existing methods, such as the ETE method, constructs a set of adaptive observation planes for viewpoint selection after exploring and generating a geometric proxy. After identifying low-quality regions, data capture is repeated until the reconstruction quality reaches a desired level. However, generating observation planes based on regional slopes can lead to gaps. This is because the observation planes are created based on the average of the normals of the patch points, and in complex areas, the observation planes may not accurately cover all areas. The second category of existing methods evaluates the quality of the geometric point cloud and selects two types of new viewpoints in a constrained view sampling space: the first type of new viewpoints is used to improve incomplete or low-quality regions, and the second type of viewpoints is added to fully cover the entire scene. Because the sample points on the geometric proxy are evenly distributed, there is a high proportion of low-quality sample points at the edges of buildings or structures with rich geometric information. The third category of existing methods finds the optimal position and orientation for each candidate camera. Based on the constraints of the drone, the optimal subset of camera positions is selected as candidate viewpoints for the optimal flight path. This path is often obtained using heuristic solutions, such as the traveling salesman problem (TSP) that is solved through all selected viewpoints. These methods separate path construction from view optimization, without considering path continuity or quality, which can significantly increase the workload of field acquisition. Furthermore, the paths obtained in this way often contain many sharp turns. When the drone passes through a turn, the necessary deceleration leads to a decrease in efficiency, while frequent acceleration consumes additional power. Summary of the Invention

[0005] In response to the above-mentioned problems existing in the prior art, an embodiment of the present invention provides an aerial photography perspective planning method, device and computer-readable medium based on optimal view photogrammetry; this method can achieve high-precision full-coverage three-dimensional reconstruction of the target scene at a relatively low cost, and improve the restoration degree of the target scene model reconstruction.

[0006] According to a first aspect of an embodiment of the present invention, a method for aerial photography perspective planning based on optimal view photogrammetry is provided, the method comprising: determining a viewpoint sampling space of a drone based on a 3D rough model of a target scene; performing Poisson disk sampling on the surface of the 3D rough model to generate a sampling point set and a preselected viewpoint set corresponding to the sampling point set; performing reconstructability modeling on the 3D rough model based on the sampling point set to generate the reconstructability of each sampling point; performing viewpoint redundancy modeling based on the preselected viewpoint set to generate the redundancy of each preselected viewpoint; and performing optimal processing on the preselected viewpoint set based on the reconstructability of the sampling points and the redundancy of the preselected viewpoints to generate an aerial photography perspective.

[0007] Optionally, the Poisson disk sampling is performed on the surface of the 3D rough model to generate a sampling point set and a preselected viewpoint set corresponding to the sampling point set; including: determining the sampling diameter of the Poisson disk based on the field of view of the camera, the overlap rate between adjacent images, and the preset distance of the viewpoints; based on the sampling diameter, Poisson disk sampling is performed on the surface of the 3D rough model to generate a sampling point set; for any sampling point in the sampling point set: obtaining the surface normal vector at the sampling point; moving the viewpoint by a preset distance in the opposite direction of the surface normal vector to obtain a preselected viewpoint corresponding to the sampling point; and obtaining the preselected viewpoint set corresponding to the sampling point set based on the preselected viewpoint corresponding to each sampling point in the sampling point set.

[0008] Optionally, the reconstructability modeling of the 3D rough model is performed based on the sampling point set to generate the reconstructability of each sampling point, including: for any sampling point in the sampling point set: obtaining a preselected viewpoint and at least one observation viewpoint corresponding to the sampling point; determining a reconstruction quality model for the sampling point between the preselected viewpoint and the observation viewpoint; the observation viewpoint is used to indicate a viewpoint that is outside the normal vector of the sampling point and can observe the sampling point; for any observation viewpoint of the at least one observation viewpoint: determining a reconstructability metric of the sampling point between the preselected viewpoint and the observation viewpoint according to the reconstruction quality model; and determining the reconstructability of the sampling point based on the reconstruction metric corresponding to each of the at least one observation viewpoint.

[0009] Optionally, the method of performing viewpoint redundancy modeling based on the preselected viewpoint set to generate redundancy for each preselected viewpoint includes: for any preselected viewpoint in the preselected viewpoint set: determining a sampling point observable by the preselected viewpoint as a first target sampling point to obtain a plurality of first target sampling points; obtaining a plurality of reconstructivities based on the reconstructibility corresponding to each of the plurality of first target sampling points; and selecting a minimum reconstructibility value from the plurality of reconstructivities as the redundancy corresponding to the preselected viewpoint.

[0010] Optionally, based on the reconstructability of the sampling points and the redundancy of the preselected viewpoints, the preselected viewpoint set is optimized to generate an aerial photography perspective; including: obtaining the redundancy of each preselected viewpoint in the preselected viewpoint set; when it is determined that the reconstructability of the sampling point corresponding to each preselected viewpoint in the preselected viewpoint set meets a preset condition, performing redundancy minimization processing on the preselected viewpoint set to obtain a candidate viewpoint set; when it is determined that the number of candidate viewpoints in the candidate viewpoint set is unchanged, based on the preselected viewpoint set, performing reconstructability maximization processing on the sampling point corresponding to each candidate viewpoint in the candidate viewpoint set to generate a quasi-viewpoint set; and determining each quasi-viewpoint in the quasi-viewpoint set as an aerial photography perspective.

[0011] Optionally, when it is determined that the reconstructability of the sampling point corresponding to each pre-selected viewpoint in the pre-selected viewpoint set meets a preset condition, redundancy minimization processing is performed on the pre-selected viewpoint set to obtain a candidate viewpoint set; comprising: obtaining the redundancy corresponding to each pre-selected viewpoint in the pre-selected viewpoint set, and sorting the pre-selected viewpoints in the pre-selected viewpoint set in descending order of redundancy to generate a sorting result; taking the pre-selected viewpoint located first in the sorting result as the viewpoint to be deleted, and performing a deletion operation on the viewpoint to be deleted in the pre-selected viewpoint set to obtain A candidate viewpoint set is obtained; at least one second target sampling point associated with the viewpoint to be deleted is obtained, and the reconstructability of the second target sampling point is updated based on the deletion operation; whether the reconstructability of each second target sampling point in the at least one second target sampling point is greater than a preset threshold; if so, the candidate viewpoint set is used as a pre-selected viewpoint set; if not, the viewpoint to be deleted is added to the candidate viewpoint set; and the deletion operation is continued for the second pre-selected viewpoint in the sorted result until the last pre-selected viewpoint, to obtain a candidate viewpoint set.

[0012] Optionally, when it is determined that the number of candidate viewpoints in the candidate viewpoint set remains unchanged, based on the pre-selected viewpoint set, reconstructibility maximization processing is performed on the sampling points corresponding to each candidate viewpoint in the candidate viewpoint set to generate a quasi-viewpoint set; the process includes: determining a deleted viewpoint set based on the candidate viewpoint set and the pre-selected viewpoint set; for any candidate viewpoint in the candidate viewpoint set: determining the similarity between each deleted viewpoint in the deleted viewpoint set and the candidate viewpoint; based on the similarity, selecting deleted viewpoints with similarity greater than a preset threshold from the deleted viewpoint set, and using the selected deleted viewpoints as related viewpoints of the candidate viewpoint to obtain a related viewpoint set; if there is a related viewpoint in the related viewpoint set that can maximize the reconstructibility of the sampling points corresponding to the candidate viewpoint; then replacing the candidate viewpoint with the related viewpoint; if there is no related viewpoint in the related viewpoint set that can maximize the reconstructibility of the sampling points corresponding to the candidate viewpoint, then no processing is performed; performing the above operations on each candidate viewpoint in the candidate viewpoint set to generate a quasi-viewpoint set.

[0013] According to the second aspect of an embodiment of the present invention, an aerial photography perspective planning device based on optimal view photogrammetry is also provided, including: a determination module for determining the viewpoint sampling space of a drone based on a 3D rough model of a target scene; a sampling module for performing Poisson disk sampling on the surface of the 3D rough model to generate a sampling point set and a preselected viewpoint set corresponding to the sampling point set; a reconstructibility modeling module for performing reconstructibility modeling on the 3D rough model based on the sampling point set to generate the reconstructibility of each sampling point; a redundancy modeling module for performing viewpoint redundancy modeling based on the preselected viewpoint set to generate the redundancy of each preselected viewpoint; and a preferred processing module for performing preferred processing on the preselected viewpoint set based on the reconstructibility of the sampling points and the redundancy of the preselected viewpoints to generate an aerial photography perspective.

[0014] According to a third aspect of an embodiment of the present invention, an electronic device is further provided, comprising: a processor; a memory for storing instructions executable by the processor; and the processor for reading the executable instructions from the memory and executing the instructions to implement the method described in the first aspect.

[0015] According to a fourth aspect of an embodiment of the present invention, a computer-readable medium is further provided, on which a computer program is stored. When the program is executed by a processor, the method described in the first aspect is implemented.

[0016] An embodiment of the present invention provides an aerial photography perspective planning method based on optimal photogrammetry, the method comprising: first, determining the viewpoint sampling space of the drone based on a 3D rough model of the target scene; then, performing Poisson disk sampling on the surface of the 3D rough model to generate a sampling point set and a preselected viewpoint set corresponding to the sampling point set; secondly, based on the sampling point set, performing reconstructibility modeling on the 3D rough model to generate the reconstructibility of each sampling point; thereafter, performing viewpoint redundancy modeling based on the preselected viewpoint set to generate the redundancy of each preselected viewpoint; finally, based on the reconstructibility of the sampling points and the redundancy of the preselected viewpoints, performing optimal processing on the preselected viewpoint set to generate an aerial photography perspective. This embodiment combines maximizing sample point reconstructability with minimizing viewpoint redundancy for optimal processing of the preselected viewpoint set of the target scene, thereby improving the accuracy of the three-dimensional reconstruction of the target scene while realizing automatic drone aerial photography planning for the target scene. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Hereinafter, some specific embodiments of the present invention will be described in detail in an exemplary and non-limiting manner with reference to the accompanying drawings. The same reference numerals in the accompanying drawings indicate the same or similar components or parts. It should be understood by those skilled in the art that these drawings are not necessarily drawn to scale. In the accompanying drawings:

[0018] Figure 1A schematic diagram of a flow chart of an aerial photography viewing angle planning method based on optimal view photogrammetry provided by one embodiment of the present invention;

[0019] Figure 2 A schematic diagram of a process for generating an aerial photography perspective based on optimal processing according to an embodiment of the present invention;

[0020] Figure 3 A schematic structural diagram of an aerial photography viewing angle planning device based on optimal view photogrammetry provided by one embodiment of the present invention. DETAILED DESCRIPTION

[0021] In order to make the purposes, features, and advantages of the present invention more obvious and easy to understand, 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 embodiments described are only 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 those skilled in the art without making creative efforts shall fall within the scope of protection of the present invention.

[0022] like Figure 1 1 is a flow chart of an aerial photography viewing angle planning method based on optimal view photogrammetry provided by one embodiment of the present invention.

[0023] A method for aerial photography viewing angle planning based on optimal photogrammetry includes at least the following steps:

[0024] S101, based on the 3D rough model of the target scene, determine the viewpoint sampling space of the UAV;

[0025] S102, performing Poisson disk sampling on the viewpoint sampling space to generate a sampling point set and a pre-selected viewpoint set corresponding to the sampling point set;

[0026] S103, based on the sampling point set, performing reconstructability modeling on the 3D rough model to generate the reconstructability of each sampling point;

[0027] S104, performing viewpoint redundancy modeling based on the preselected viewpoint set to generate redundancy for each preselected viewpoint;

[0028] S105 , performing optimization processing on the pre-selected viewpoint set based on the reconstructability of the sampling points and the redundancy of the pre-selected viewpoints to generate an aerial photography viewing angle.

[0029] In S101, in practical applications of scene 3D reconstruction, a 3D rough model of the target scene is primarily used as input for optimal photogrammetry. This rough model forms the basis for trajectory planning in optimal photogrammetry. The rough model represents the basic geometry of the target scene or object. The rough model includes geometric information about the observation site and is primarily used to determine the aerial perspective and plan a safe airspace for the drone's flight path.

[0030] The 3D rough model is obtained by the following method:

[0031] 1) Obtain image data of the target scene using oblique photogrammetry and generate a 3D rough model;

[0032] 2) First, two-dimensional plane information such as aerial survey orthophotos is used to plan the collection route of the target scene. Then, the drone flies along the collection route to collect image data of the target scene. Finally, a 3D rough model is generated based on the image data.

[0033] The safe distance d between the UAV and the target scene is determined based on the UAV's posture control capability, the precision of the 3D model, and the complexity of the target scene. min Based on the maximum shooting distance d from the drone to the target scene max And the safe distance d from the drone to the target scene min , determine the viewpoint sampling space from the drone to the target scene; among them, the farthest shooting distance d max Used to indicate the maximum observable distance corresponding to the viewpoint, which is preset based on empirical values; d max The smaller the value, the higher the reconstruction accuracy of the target scene model.

[0034] After recognizing the 3D model, min As the scale, the target scene is expanded in the horizontal and vertical directions to generate the no-fly zone for the drone. The no-fly zone is used to prohibit the drone from planning its trajectory to ensure the flight safety of the drone. After recognizing the 3D rough model, the maximum shooting distance d is used to generate the no-fly zone for the drone. max Using the scale as the baseline, the target scene is dilated horizontally and vertically to generate the drone's flight space. Based on the drone's flight space and no-fly zones, the drone's viewpoint sampling space is determined. This not only prevents collisions between the drone and buildings but also increases the drone's sampling frequency.

[0035] It should be noted that if the target scene has a large area, the target scene may be divided, and then the viewpoint sampling space of each divided small scene may be determined separately.

[0036] In S102, Poisson disk sampling involves dividing the 3D model into several small local areas, configuring each local area accordingly, and finally generating a corresponding aerial perspective, thus breaking the whole into small parts. Compared with other sampling methods, the sampling results of Poisson disk sampling can better highlight the geometric features of the 3D model surface.

[0037] Poisson disk sampling is performed on the surface of the 3D model according to the preset sampling diameter of the Poisson disk. To effectively control the density of the sampling viewpoints, the sampling diameter of the Poisson disk is determined based on the camera's field of view, the overlap ratio between adjacent images, and the preset distance between the viewpoints. The Poisson disk sampling is then performed on the surface of the 3D model based on this determined sampling diameter.

[0038] Exemplarily, based on the camera's field of view, the overlap rate between adjacent images, and the preset viewpoint distance, a sampling diameter of a Poisson disk is determined; based on the sampling diameter, Poisson disk sampling is performed on the surface of the 3D schematic model to generate a sampling point set; for any sampling point in the sampling point set: a surface normal vector at the sampling point is obtained; the viewpoint is moved a preset distance in the opposite direction of the surface normal vector to obtain a preselected viewpoint corresponding to the sampling point; based on the preselected viewpoint corresponding to each sampling point in the sampling point set, a preselected viewpoint set corresponding to the sampling point set is obtained.

[0039] For example: (1) The sampling diameter D of the Poisson disk disk Calculated by the following formula (1):

[0040] D disk =d prj ·(1-r overlap );

[0041] d prj =2·d GSD tan(θ / 2); Formula (1);

[0042] Where θ is the camera’s field of view, r overlap is the overlap ratio between adjacent images, which is a preset value obtained based on experience; d prj Represents the field of view of the camera projected on the surface of the target scene model; d GSD Indicates the viewpoint preset distance.

[0043] (2) Using the Poisson disk sampling method, the sampling diameter D obtained above is disk Poisson disk sampling is performed on the surface of the 3D rough model to uniformly generate a set of sampling points S for observation.

[0044] S={{s1,n1},{s2,n2},…,{s i ,n i}}; Formula (2);

[0045] in, Represents i sampling areas N i The average normal vector of the sampling point s i The surface normal vector at .

[0046] (3) Set the sampling point s i Along the surface normal vector n i Move in the opposite direction of d GSD Distance, thereby determining the viewpoint v corresponding to the sampling point i , viewpoint v i The shooting direction and normal vector n i The direction is opposite to that of the image. Through the analysis of stereo visual observation and observability constraints between adjacent sampling points, a corresponding viewpoint is generated for each sampling point. Based on the viewpoint v corresponding to each sampling point in the sampling point set S i , generate a pre-selected viewpoint set V, where viewpoint v i ∈V is determined by the camera position and direction R i ∈SO(3) definition; distance d GSD The overlap ratio of images captured by the preselected viewpoint set V is determined.

[0047] In S103 , no limitation is imposed on the method of reconstructability modeling.

[0048] Based on the sampling points and the pre-selected viewpoints corresponding to the sampling points, a quantitative method is applied to pre-evaluate the accuracy and completeness of the target scene reconstruction results. To this end, a quantitative metric called reconstructability is proposed based on the theory of dual-view geometry analysis.

[0049] Exemplarily, for any sampling point in the sampling point set: a preselected viewpoint and at least one observation viewpoint corresponding to the sampling point are obtained; a reconstruction quality model of the sampling point between the preselected viewpoint and the observation viewpoint is determined; the observation viewpoint is used to indicate a viewpoint that is located outside the normal vector of the sampling point and can observe the sampling point; for any observation viewpoint in the at least one observation viewpoint: a reconstruction metric of the sampling point between the preselected viewpoint and the observation viewpoint is determined based on the reconstruction quality model; and the reconstructability of the sampling point is determined based on the reconstruction metric corresponding to each observation viewpoint in the at least one observation viewpoint.

[0050] For example: Reconstructibility q(s,v i ,v j ) is used to indicate the preselected viewpoint v corresponding to any sampling point s at that sampling point i and observation point v j The reconstruction quality model under the observation of is shown in the following formula (3).

[0051] q(s,v i ,v j )=w1(α)w2(d m )w3(α)cos(θ m )

[0052] Formula (3);

[0053] In the formula, α represents the two viewpoints (v i ,v j ) to the sampling point s respectively; It represents two viewpoints (v i ,v j ) are the larger values of the distances to the sampling point s; θ m =max(θ i ,θ j ), which represents two viewpoints (v i ,v j ) angle, the viewpoint angle is used to indicate the preselected viewpoint v i Or observation point v j are the angles between the sampling point vector and the sampling point normal vector n; w1, w2 and w3 are all weight terms, and their definitions are as shown in the following formula (4).

[0054]

[0055] Among them, α1 represents the pre-selected viewpoint v i The corresponding preset angle, α3 represents the observation viewpoint v j The corresponding preset angle. k3 and k1 represent coefficients.

[0056] Each sampling point can be observed from multiple viewpoints. For any sampling point s: obtain the viewpoint set V that can observe the sampling point s = {v i} i＝1,…M ; Obtain a preselected viewpoint and at least one observation viewpoint corresponding to the sampling point from the viewpoint set; For any observation viewpoint in the at least one observation viewpoint: Calculate the reconstructability between the preselected viewpoint and the observation viewpoint based on formula (3); Then, sum the reconstructability corresponding to each observation viewpoint in the at least one observation viewpoint to generate the reconstructability h(s, V) of the sampling point. The specific calculation formula is shown in the following formula (5):

[0057]

[0058] Among them, δ(s,v i ) and δ(s,v j ) is the visibility function, which represents the pre-selected viewpoint v i and observation point v jVisibility to the sampling point s.

[0059] In S104, illustratively, for any preselected viewpoint in the preselected viewpoint set: sampling points observable by the preselected viewpoint are determined as first target sampling points to obtain a plurality of first target sampling points; a plurality of reconstructivities are obtained based on the reconstructibility corresponding to each of the plurality of first target sampling points; and a minimum reconstructibility value is selected from the plurality of reconstructivities as the redundancy corresponding to the preselected viewpoint.

[0060] For example: at viewpoint v i Several sampling points can be observed at the location, and the reconstruction h(s, V) corresponding to each of the several sampling points is obtained to obtain several reconstructions; the several reconstructions are sorted in order; the minimum value of the reconstruction h(s, V) in the sorting result is determined as the redundancy r(v) of the viewpoint, that is:

[0061] r(v,V)=min{h(s,V)|s∈S,δ(s,v)=1} Formula (6);

[0062] It should be noted that if at viewpoint v i All observed sampling points have high reconstructibility, then the viewpoint v i It may be redundant and needs to be optimized.

[0063] In S105 , based on the reconstructability of the sampling points and the redundancy of the pre-selected viewpoints, the pre-selected viewpoint set is optimized using a preset rule or model to generate an aerial photography perspective.

[0064] The preselected viewpoint set generated by dense sampling of the 3D rough model ensures sufficient coverage of the target scene, but it also introduces redundant viewpoints, making it unsuitable for direct photogrammetric trajectory planning and requiring viewpoint optimization. The optimal approach is to find a quasi-viewpoint set from the preselected viewpoint set that is small enough to fully observe all sample points in the target scene, maximizing the reconstructability of all sample points while minimizing redundancy in the quasi-viewpoint set. In particular, incomplete and low-quality areas of the scene require focused observation.

[0065] Based on the optimization process, this embodiment can select a quasi-viewpoint set from the pre-selected viewpoint set, so that the sampling points can obtain sufficient observations with minimal viewpoint redundancy, that is, maximize the reconstructability of the model sampling points while minimizing the redundancy of all viewpoints; thereby, not only can the viewpoints with low quality scores of the target scene with more geometric information be effectively eliminated, but the restoration degree of the model reconstruction can also be improved.

[0066] like Figure 2FIG. 1 is a flow chart of generating an aerial photography perspective based on a preferred process according to an embodiment of the present invention.

[0067] Based on the reconstructability of the sampling points and the redundancy of the pre-selected viewpoints, the pre-selected viewpoint set is optimized to generate an aerial photography perspective. The process includes at least the following steps:

[0068] S201, obtaining the redundancy of each pre-selected viewpoint in the pre-selected viewpoint set;

[0069] S202, when it is determined that the reconstructability of the sampling point corresponding to each pre-selected viewpoint in the pre-selected viewpoint set meets a preset condition, performing redundancy minimization processing on the pre-selected viewpoint set to obtain a candidate viewpoint set;

[0070] S203, when it is determined that the number of candidate viewpoints in the candidate viewpoint set remains unchanged, based on the pre-selected viewpoint set, performing a reconstruction maximization process on the sampling points corresponding to each candidate viewpoint in the candidate viewpoint set to generate a quasi-viewpoint set;

[0071] S204: Determine each quasi-viewpoint in the quasi-viewpoint set as an aerial photography viewing angle.

[0072] In S201 , for each pre-selected viewpoint in the pre-selected viewpoint set: the redundancy of each pre-selected viewpoint is determined based on formula (6).

[0073] In S202, illustratively, a redundancy corresponding to each pre-selected viewpoint in the pre-selected viewpoint set is obtained, and the pre-selected viewpoints in the pre-selected viewpoint set are sorted in descending order of redundancy to generate a sorting result; a pre-selected viewpoint ranked first in the sorting result is used as a viewpoint to be deleted, and a deletion operation is performed on the viewpoint to be deleted in the pre-selected viewpoint set to obtain a candidate viewpoint set; at least one second target sampling point associated with the viewpoint to be deleted is obtained, and the reconstructability of the second target sampling point is updated based on the deletion operation; it is determined whether the reconstructability of each second target sampling point in the at least one second target sampling point is greater than a preset threshold; if so, the candidate viewpoint set is used as the pre-selected viewpoint set; if not, the viewpoint to be deleted is added to the candidate viewpoint set; and the deletion operation is continued for the pre-selected viewpoint ranked second in the sorting result until the last pre-selected viewpoint to obtain a candidate viewpoint set.

[0074] For example:

[0075] Step 1: Determine the redundancy r(v) of each pre-selected viewpoint in the pre-selected viewpoint set V, and sort the pre-selected viewpoints in the pre-selected viewpoint set in descending order of redundancy to generate a sorting result;

[0076] Step 2: The pre-selected viewpoint v with the highest redundancy in the sorting results ias the viewpoints to be deleted, and perform a deletion operation on the viewpoints to be deleted in the pre-selected viewpoint set to obtain a candidate viewpoint set W;

[0077] Step 3: Get and preselect viewpoint v i At least one second target sampling point associated with the preselected viewpoint v i The sampling points that can be observed; and based on formula (5), the reconstructibility of each second target sampling point is updated;

[0078] Step 4: Determine whether the reconstructability of each second target sampling point after the update is greater than the threshold t h If yes, return to step 2 and process the second target sampling viewpoint v according to the sorting result. i+1 ; If not, the viewpoint v to be deleted i Add it back to the candidate viewpoint set and return to step 2 to process the second target sampling viewpoint v according to the sorting result i+1 .

[0079] In S203, exemplarily, a deleted viewpoint set is determined based on the candidate viewpoint set and the pre-selected viewpoint set; for any candidate viewpoint in the candidate viewpoint set: the similarity between each deleted viewpoint in the deleted viewpoint set and the candidate viewpoint is determined; based on the similarity, deleted viewpoints having a similarity greater than a preset threshold are selected from the deleted viewpoint set, and the selected deleted viewpoints are used as related viewpoints of the candidate viewpoint to obtain a related viewpoint set; if a related viewpoint exists in the related viewpoint set that can maximize the reconstructability of the sampling point corresponding to the candidate viewpoint, the candidate viewpoint is replaced by the related viewpoint; if a related viewpoint does not exist in the related viewpoint set that can maximize the reconstructability of the sampling point corresponding to the candidate viewpoint, no processing is performed; the above operations are performed for each candidate viewpoint in the candidate viewpoint set to generate a quasi-viewpoint set.

[0080] For example:

[0081] Step 1: Determine the candidate viewpoint v in the candidate viewpoint set i The similarity between it and each deleted viewpoint in the deleted viewpoint set. The similarity calculation formula is as shown in formula (7), which is used to find the closest candidate viewpoint v from the deleted viewpoint set. i All deleted viewpoints.

[0082]

[0083] Among them, i o j is the dot product that measures the similarity between viewpoint directions; represents the distance between the centers of the two cameras; ε is a small constant used to avoid integer division by zero.

[0084] Step 2: Find each candidate viewpoint v in the candidate viewpoint set W from the deleted viewpoint set V i Similar related viewpoint set Ω(v i );

[0085] Step 3: From the relevant viewpoint set Ω(v i ) query makes the candidate viewpoint v i The corresponding viewpoints that maximize the reconstructability H(S,U) of the sampling points

[0086] Step 4: If the query result representation exists, then the relevant viewpoint Replace v i , and the next candidate viewpoint v in the candidate viewpoint set i+1 Continue to step 1; if the query result representation does not exist, no processing is performed, and the next candidate viewpoint v in the candidate viewpoint set is selected. i+1 Continue with step 1.

[0087] This embodiment combines minimizing viewpoint redundancy with maximizing reconstructability. By optimizing the aerial photography viewing angle, it can achieve the goal of maximizing model reconstruction accuracy with the least number of images. This not only realizes automatic drone aerial photography planning but also reduces aerial photography costs.

[0088] It should be noted that the viewpoint mentioned in each embodiment of the present invention refers to the image acquisition viewpoint when the drone performs aerial photography of the target scene.

[0089] The following describes in detail an aerial photography viewing angle planning method based on optimal view photogrammetry provided by this embodiment in conjunction with specific application scenarios.

[0090] A method for aerial photography viewing angle planning based on optimal photogrammetry includes at least the following steps:

[0091] S1, based on the 3D rough model of the target scene, determines the UAV’s viewpoint sampling space.

[0092] S2, based on the camera's field of view, the overlap rate between adjacent images, and the preset viewpoint distance, determine a sampling diameter of a Poisson disk; based on the sampling diameter, perform Poisson disk sampling on the surface of the 3D schematic model to generate a sampling point set; for any sampling point in the sampling point set: obtain a surface normal vector at the sampling point; move the viewpoint a preset distance in the opposite direction of the surface normal vector to obtain a preselected viewpoint corresponding to the sampling point; and obtain a preselected viewpoint set corresponding to the sampling point set based on the preselected viewpoint corresponding to each sampling point in the sampling point set.

[0093] S3, for any sampling point in the sampling point set: obtaining a preselected viewpoint and at least one observation viewpoint corresponding to the sampling point; determining a reconstruction quality model for the sampling point between the preselected viewpoint and the observation viewpoint; the observation viewpoint is used to indicate a viewpoint that is outside the normal vector of the sampling point and can observe the sampling point; for any observation viewpoint in the at least one observation viewpoint: determining a reconstructability metric of the sampling point between the preselected viewpoint and the observation viewpoint based on the reconstruction quality model; and determining the reconstructibility of the sampling point based on the reconstruction metric corresponding to each observation viewpoint in the at least one observation viewpoint.

[0094] S4. For any preselected viewpoint in the preselected viewpoint set, sampling points observable by the preselected viewpoint are determined as first target sampling points to obtain a plurality of first target sampling points; a plurality of reconstructivities are obtained based on the reconstructibility corresponding to each of the plurality of first target sampling points; and a minimum reconstructibility value is selected from the plurality of reconstructivities as the redundancy corresponding to the preselected viewpoint.

[0095] S5, obtaining redundancy of each pre-selected viewpoint in the pre-selected viewpoint set; obtaining redundancy corresponding to each pre-selected viewpoint in the pre-selected viewpoint set, and sorting the pre-selected viewpoints in the pre-selected viewpoint set in descending order of redundancy to generate a sorting result; selecting the pre-selected viewpoint ranked first in the sorting result as a viewpoint to be deleted, and performing a deletion operation on the viewpoint to be deleted in the pre-selected viewpoint set to obtain a candidate viewpoint set; obtaining at least one second target sampling point associated with the viewpoint to be deleted, and updating the reconstructability of the second target sampling point based on the deletion operation; determining whether the reconstructability of each second target sampling point in the at least one second target sampling point is greater than a preset threshold; if so, selecting the candidate viewpoint set as the pre-selected viewpoint set; if not, adding the viewpoint to be deleted to the candidate viewpoint set; and continuing to perform the deletion operation on the pre-selected viewpoint ranked second in the sorting result until the last pre-selected viewpoint to obtain a candidate viewpoint set.

[0096] S6, determining a deleted viewpoint set based on the candidate viewpoint set and the pre-selected viewpoint set; for any candidate viewpoint in the candidate viewpoint set: determining the similarity between each deleted viewpoint in the deleted viewpoint set and the candidate viewpoint; based on the similarity, selecting deleted viewpoints with similarity greater than a preset threshold from the deleted viewpoint set, and using the selected deleted viewpoints as related viewpoints of the candidate viewpoint to obtain a related viewpoint set; if there is a related viewpoint in the related viewpoint set that can maximize the reconstruction of the sampling point corresponding to the candidate viewpoint; then replacing the candidate viewpoint with the related viewpoint; if there is no related viewpoint in the related viewpoint set that can maximize the reconstruction of the sampling point corresponding to the candidate viewpoint, then no processing is performed; performing the above operation for each candidate viewpoint in the candidate viewpoint set to generate a quasi-viewpoint set.

[0097] This embodiment proposes a method for aerial photography perspective planning based on optimal photogrammetry. This method generates a rough 3D box model of the target scene and uses this as the planning basis to constrain the generation and selection of a preselected viewpoint set. The sampling point set is then evaluated and optimized based on a criterion that combines minimizing viewpoint redundancy and maximizing reconstructability, selecting the optimal drone aerial photography perspective that matches the target scene's geometry and meets the requirements for 3D reconstruction. This method achieves automated drone aerial photography planning and maximizes model reconstruction accuracy with a minimum number of images.

[0098] like Figure 3 , which is a structural diagram of an aerial photography viewing angle planning device based on optimal view photogrammetry provided by one embodiment of the present invention.

[0099] A device for planning aerial photography perspectives based on optimal view photogrammetry, the device 300 comprising: a determination module 301 for determining a viewpoint sampling space of a drone based on a 3D rough model of a target scene; a sampling module 302 for performing Poisson disk sampling on the surface of the 3D rough model to generate a sampling point set and a preselected viewpoint set corresponding to the sampling point set; a reconstructibility modeling module 303 for performing reconstructibility modeling on the 3D rough model based on the sampling point set to generate the reconstructibility of each sampling point; a redundancy modeling module 304 for performing viewpoint redundancy modeling based on the preselected viewpoint set to generate the redundancy of each preselected viewpoint; and a preferred processing module 305 for performing preferred processing on the preselected viewpoint set based on the reconstructibility of the sampling points and the redundancy of the preselected viewpoints to generate an aerial photography perspective.

[0100] In a preferred implementation of this embodiment, the sampling module includes: a determination unit for determining a sampling diameter of a Poisson disk based on a camera's field of view, an overlap ratio between adjacent images, and a preset viewpoint distance; a sampling unit for performing Poisson disk sampling on the surface of the 3D schematic model based on the sampling diameter to generate a sampling point set; an acquisition unit for, for any sampling point in the sampling point set: obtaining a surface normal vector at the sampling point; moving the viewpoint a preset distance in the opposite direction of the surface normal vector to obtain a preselected viewpoint corresponding to the sampling point; and an acquisition unit for obtaining a preselected viewpoint set corresponding to the sampling point set based on the preselected viewpoint corresponding to each sampling point in the sampling point set.

[0101] In a preferred implementation of this embodiment, the reconstructability modeling module is further configured to: for any sampling point in the sampling point set: obtain a preselected viewpoint and at least one observation viewpoint corresponding to the sampling point; determine a reconstruction quality model for the sampling point between the preselected viewpoint and the observation viewpoint; the observation viewpoint is used to indicate a viewpoint that is outside the normal vector of the sampling point and can observe the sampling point; for any observation viewpoint in the at least one observation viewpoint: determine a reconstructability metric of the sampling point between the preselected viewpoint and the observation viewpoint based on the reconstruction quality model; and determine the reconstructability of the sampling point based on the reconstruction metric corresponding to each observation viewpoint in the at least one observation viewpoint.

[0102] In a preferred implementation of this embodiment, the redundancy modeling module includes: a determining unit configured to, for any preselected viewpoint in the preselected viewpoint set, determine a sampling point observable by the preselected viewpoint as a first target sampling point, thereby obtaining a plurality of first target sampling points; an obtaining unit configured to obtain a plurality of reconstructivities based on the reconstructibility corresponding to each of the plurality of first target sampling points; and a selecting unit configured to select a minimum reconstructibility value from the plurality of reconstructivities as the redundancy corresponding to the preselected viewpoint.

[0103] In a preferred implementation of this embodiment, the preferred processing module includes: an acquisition unit for acquiring the redundancy of each pre-selected viewpoint in the pre-selected viewpoint set; a redundancy processing unit for performing redundancy minimization processing on the pre-selected viewpoint set to obtain a candidate viewpoint set when it is determined that the reconstructibility of the sampling point corresponding to each pre-selected viewpoint in the pre-selected viewpoint set meets a preset condition; a reconstructibility processing unit for performing reconstructibility maximization processing on the sampling point corresponding to each candidate viewpoint in the candidate viewpoint set based on the pre-selected viewpoint set to generate a quasi-viewpoint set when it is determined that the number of candidate viewpoints in the candidate viewpoint set remains unchanged; and a determination unit for determining each quasi-viewpoint in the quasi-viewpoint set as an aerial photography perspective.

[0104] In a preferred implementation of this embodiment, the redundancy processing unit includes: a sorting subunit, configured to obtain the redundancy corresponding to each pre-selected viewpoint in the pre-selected viewpoint set, and sort the pre-selected viewpoints in the pre-selected viewpoint set in descending order of redundancy to generate a sorting result; a deleting subunit, configured to select the pre-selected viewpoint ranked first in the sorting result as a viewpoint to be deleted, and perform a deleting operation on the viewpoint to be deleted in the pre-selected viewpoint set to obtain a candidate viewpoint set; an updating subunit, configured to obtain at least one second target sampling point associated with the viewpoint to be deleted, and update the reconstructability of the second target sampling point based on the deleting operation; a judging subunit, configured to judge whether the reconstructability of each second target sampling point in the at least one second target sampling point is greater than a preset threshold; if so, select the candidate viewpoint set as the pre-selected viewpoint set; if not, add the viewpoint to be deleted to the candidate viewpoint set; and continue to perform the deleting operation on the pre-selected viewpoint ranked second in the sorting result until the last pre-selected viewpoint to obtain a candidate viewpoint set.

[0105] In a preferred implementation of this embodiment, the reconstructability processing unit includes: a determination subunit, configured to determine a deleted viewpoint set based on the candidate viewpoint set and the pre-selected viewpoint set; a similarity selection subunit, configured to determine, for any candidate viewpoint in the candidate viewpoint set, a similarity between each deleted viewpoint in the deleted viewpoint set and the candidate viewpoint; based on the similarity, select deleted viewpoints having a similarity greater than a preset threshold from the deleted viewpoint set, and use the selected deleted viewpoints as related viewpoints of the candidate viewpoint to obtain a related viewpoint set; if a related viewpoint exists in the related viewpoint set that maximizes the reconstructability of the sampling point corresponding to the candidate viewpoint, then the candidate viewpoint is replaced with the related viewpoint; if a related viewpoint does not exist in the related viewpoint set that maximizes the reconstructability of the sampling point corresponding to the candidate viewpoint, then no processing is performed; and a generation subunit, configured to perform the above operations on each candidate viewpoint in the candidate viewpoint set to generate a quasi-viewpoint set.

[0106] The above-described device can execute the method for planning an aerial photography perspective based on optimal view photogrammetry provided in one embodiment of the present invention, and has the corresponding functional modules and beneficial effects for executing the method. For technical details not fully described in this embodiment, please refer to the method for planning an aerial photography perspective based on optimal view photogrammetry provided in one embodiment of the present invention.

[0107] The present invention also provides an electronic device, comprising: a processor; a memory for storing executable instructions of the processor; the processor is used to read the executable instructions from the memory and execute the instructions to implement the aerial photography perspective planning method based on optimal view photogrammetry described in the present invention.

[0108] In addition to the above-mentioned methods and devices, an embodiment of the present application may also be a computer program product, which includes computer program instructions, which, when executed by a processor, enable the processor to execute the steps of the method according to various embodiments of the present application described in the above-mentioned "Exemplary Method" section of this specification.

[0109] The computer program product may be written in any combination of one or more programming languages to implement the program code for performing the operations of the embodiments of the present application, including object-oriented programming languages such as Java, C++, and conventional procedural programming languages such as "C" or similar programming languages. The program code may be executed entirely on the user's computing device, partially on the user's computing device, as a standalone software package, partially on the user's computing device and partially on a remote computing device, or entirely on a remote computing device or server.

[0110] In addition, an embodiment of the present application may also be a computer-readable storage medium having computer program instructions stored thereon, which, when executed by a processor, enable the processor to execute the steps of the method according to the following embodiments of the present application described in the above "Exemplary Method" section of this specification.

[0111] The computer-readable storage medium can adopt any combination of one or more readable media. The readable medium can be a readable signal medium or a readable storage medium. The readable storage medium can, for example, include but is not limited to a system, device or component of electricity, magnetism, light, electromagnetic, infrared, or semiconductor, or any combination thereof. More specific examples (non-exhaustive list) of readable storage media include: an electrical connection with one or more wires, a portable disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof.

[0112] The basic principles of the present application have been described above in conjunction with specific embodiments. However, it should be noted that the advantages, strengths, and effects mentioned in this application are merely illustrative and not restrictive, and it should not be assumed that these advantages, strengths, and effects are required of each embodiment of this application. In addition, the specific details disclosed above are merely illustrative and facilitating understanding, and are not restrictive. The above details do not limit this application to necessarily being implemented using the above specific details.

[0113] The block diagrams of the devices, devices, equipment, and systems involved in this application are merely illustrative examples and are not intended to require or imply that they must be connected, arranged, or configured in the manner shown in the block diagrams. As will be appreciated by those skilled in the art, these devices, devices, equipment, and systems can be connected, arranged, or configured in any manner. Words such as "include," "comprise," "have," and the like are open-ended words, meaning "including but not limited to," and can be used interchangeably therewith. The words "or" and "and" used herein refer to the words "and / or" and can be used interchangeably therewith, unless the context clearly indicates otherwise. The word "such as" used herein refers to the phrase "such as but not limited to," and can be used interchangeably therewith.

[0114] It should also be noted that in the apparatus, device, and method of the present application, each component or each step can be decomposed and / or recombined, and such decomposition and / or recombination should be regarded as equivalent solutions of the present application.

[0115] The above description of the disclosed aspects is provided to enable any person skilled in the art to make or use the present application. Various modifications to these aspects will be readily apparent to those skilled in the art, and the general principles defined herein may be applied to other aspects without departing from the scope of the present application. Therefore, the present application is not intended to be limited to the aspects shown herein, but rather to be accorded the widest scope consistent with the principles and novel features disclosed herein.

[0116] The above description has been provided for the purpose of illustration and description. Furthermore, this description is not intended to limit the embodiments of the present application to the forms disclosed herein. Although a number of example aspects and embodiments have been discussed above, those skilled in the art will recognize certain variations, modifications, alterations, additions, and sub-combinations thereof.

[0117] In the description of this specification, reference to the terms "one embodiment," "some embodiments," "examples," "specific examples," or "some examples" means that the specific features, structures, materials, or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. Moreover, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples. In addition, those skilled in the art may combine and integrate different embodiments or examples described in this specification, as well as features of different embodiments or examples, unless otherwise inconsistent.

[0118] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of the technical features being referred to. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of such features. In the description of the present invention, "plurality" means two or more, unless otherwise specifically defined.

[0119] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any modifications or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in the present invention should be included in the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be based on the scope of protection of the claims.

Claims

1. A method for aerial photography viewing angle planning based on optimal photogrammetry, characterized in that: include: Determine the drone's viewpoint sampling space based on the 3D rough model of the target scene; Performing Poisson disk sampling on the surface of the 3D schematic model to generate a sampling point set and a preselected viewpoint set corresponding to the sampling point set; Based on the sampling point set, reconstructability modeling is performed on the 3D rough model to generate the reconstructability of each sampling point; Based on the preselected viewpoint set, viewpoint redundancy modeling is performed to generate redundancy of each preselected viewpoint; Based on the reconstructability of the sampling points and the redundancy of the pre-selected viewpoints, the pre-selected viewpoint set is optimized to generate an aerial photography viewing angle.

2. The method according to claim 1, characterized in that The step of performing Poisson disk sampling on the surface of the 3D schematic model to generate a sampling point set and a preselected viewpoint set corresponding to the sampling point set comprises: The sampling diameter of the Poisson disk is determined based on the camera's field of view, the overlap rate between adjacent images, and the preset distance between viewpoints; Based on the sampling diameter, performing Poisson disk sampling on the surface of the 3D schematic model to generate a sampling point set; For any sampling point in the sampling point set: obtaining a surface normal vector at the sampling point; moving the viewpoint by a preset distance in the opposite direction of the surface normal vector to obtain a preselected viewpoint corresponding to the sampling point; Based on the pre-selected viewpoint corresponding to each sampling point in the sampling point set, a pre-selected viewpoint set corresponding to the sampling point set is obtained.

3. The method according to claim 1, characterized in that The method of performing reconstructability modeling on the 3D rough model based on the sampling point set to generate the reconstructability of each sampling point comprises: For any sampling point in the sampling point set: obtaining a preselected viewpoint and at least one observation viewpoint corresponding to the sampling point; determining a reconstruction quality model for the sampling point between the preselected viewpoint and the observation viewpoint; the observation viewpoint is used to indicate a viewpoint that is located outside the normal vector of the sampling point and can observe the sampling point; for any observation viewpoint in the at least one observation viewpoint: determining a reconstruction metric of the sampling point between the preselected viewpoint and the observation viewpoint based on the reconstruction quality model; and determining the reconstructibility of the sampling point based on the reconstruction metric corresponding to each observation viewpoint in the at least one observation viewpoint.

4. The method according to claim 1, wherein The method of performing viewpoint redundancy modeling based on the preselected viewpoint set to generate redundancy of each preselected viewpoint comprises: For any preselected viewpoint in the preselected viewpoint set: determining a sampling point observable by the preselected viewpoint as a first target sampling point, to obtain a plurality of first target sampling points; Obtaining a plurality of reconstructivities based on the reconstructibility corresponding to each of the plurality of first target sampling points; A minimum reconstructability value is selected from the plurality of reconstructivities as the redundancy corresponding to the preselected viewpoint.

5. The method according to claim 1, wherein The method of performing optimal processing on the pre-selected viewpoint set based on the reconstructability of the sampling points and the redundancy of the pre-selected viewpoints to generate an aerial photography viewing angle comprises: Obtaining the redundancy of each of the preselected viewpoints in the preselected viewpoint set; When it is determined that the reconstructability of the sampling point corresponding to each of the pre-selected viewpoints in the pre-selected viewpoint set meets a preset condition, performing redundancy minimization processing on the pre-selected viewpoint set to obtain a candidate viewpoint set; When it is determined that the number of candidate viewpoints in the candidate viewpoint set remains unchanged, based on the pre-selected viewpoint set, performing a reconstructability maximization process on the sampling points corresponding to each candidate viewpoint in the candidate viewpoint set to generate a quasi-viewpoint set; Each quasi-viewpoint in the quasi-viewpoint set is determined as an aerial photography viewing angle.

6. The method according to claim 5, characterized in that The method of performing redundancy minimization processing on the preselected viewpoint set to obtain a candidate viewpoint set when it is determined that the reconstructability of the sampling point corresponding to each preselected viewpoint in the preselected viewpoint set meets a preset condition comprises: Obtaining the redundancy corresponding to each pre-selected viewpoint in the pre-selected viewpoint set, and sorting the pre-selected viewpoints in the pre-selected viewpoint set in descending order of redundancy to generate a sorting result; taking the preselected viewpoint ranked first in the sorting result as the viewpoint to be deleted, and performing a deletion operation on the viewpoint to be deleted in the preselected viewpoint set to obtain a candidate viewpoint set; Acquire at least one second target sampling point associated with the viewpoint to be deleted, and update the reconstructability of the second target sampling point based on the deletion operation; Determining whether the reconstructability of each of the at least one second target sampling point is greater than a preset threshold; if so, using the candidate viewpoint set as a pre-selected viewpoint set; if not, adding the viewpoint to be deleted to the candidate viewpoint set; and continuing to delete the pre-selected viewpoint ranked second in the sorted result until the last pre-selected viewpoint, thereby obtaining a candidate viewpoint set.

7. The method according to claim 5, characterized in that The method of generating a quasi-viewpoint set by performing a reconstruction maximization process on the sampling points corresponding to each candidate viewpoint in the candidate viewpoint set based on the pre-selected viewpoint set when it is determined that the number of candidate viewpoints in the candidate viewpoint set remains unchanged comprises: Determining a deleted viewpoint set based on the candidate viewpoint set and the pre-selected viewpoint set; For any candidate viewpoint in the candidate viewpoint set: determining a similarity between each deleted viewpoint in the deleted viewpoint set and the candidate viewpoint; based on the similarity, selecting deleted viewpoints from the deleted viewpoint set whose similarity is greater than a preset threshold, and using the selected deleted viewpoints as related viewpoints of the candidate viewpoint to obtain a related viewpoint set; if a related viewpoint exists in the related viewpoint set that can maximize the reconstruction of the sampling point corresponding to the candidate viewpoint, replacing the candidate viewpoint with the related viewpoint; if no related viewpoint exists in the related viewpoint set that can maximize the reconstruction of the sampling point corresponding to the candidate viewpoint, no processing is performed; The above operation is performed on each candidate viewpoint in the candidate viewpoint set to generate a quasi viewpoint set.

8. An aerial photography viewing angle planning device based on optimal photogrammetry, characterized in that: include: A determination module is used to determine the UAV's viewpoint sampling space based on the 3D rough model of the target scene; a sampling module, configured to perform Poisson disk sampling on the surface of the 3D schematic model to generate a sampling point set and a preselected viewpoint set corresponding to the sampling point set; A reconstructability modeling module, configured to perform reconstructability modeling on the 3D schematic model based on the sampling point set, and generate the reconstructability of each sampling point; A redundancy modeling module, configured to perform viewpoint redundancy modeling based on the preselected viewpoint set and generate redundancy for each preselected viewpoint; The optimization processing module is used to perform optimization processing on the pre-selected viewpoint set based on the reconstructability of the sampling points and the redundancy of the pre-selected viewpoints to generate an aerial photography perspective.

9. An electronic device, comprising: processor; a memory for storing instructions executable by the processor; The processor is configured to read the executable instructions from the memory and execute the instructions to implement the method according to any one of claims 1 to 7.

10. A computer-readable medium having a computer program stored thereon, wherein when the program is executed by a processor, the method according to any one of claims 1 to 7 is implemented.

Images