Unmanned aerial vehicle path online planning method and device capable of detecting scene update change in real time and computer equipment

Abstract

The invention relates to an unmanned aerial vehicle path online planning method and device for detecting scene update change in real time and computer equipment. The method comprises the following steps: acquiring a first sampling point and a first probability contained in an initial scene reconstruction model; obtaining a first importance degree of each first candidate view angle according to the first sampling point and the first probability, and obtaining a target view angle from the first candidate view angles; according to each target view angle, generating a prior flight path of the unmanned aerial vehicle; in the flight process of the unmanned aerial vehicle according to the prior flight path, if a scene change area is detected, obtaining a second sampling point and a second probability, obtaining a second importance degree of each second candidate view angle according to the second sampling point and the second probability, and obtaining a target flight view angle from the second candidate view angles; and adjusting the prior flight path according to the target flight view angle to obtain a real-time flight path. By adopting the method, the flight path length of the unmanned aerial vehicle can be reduced, so that the scene three-dimensional model updating efficiency is improved.

Description

Technical Field

[0001] The present application relates to the field of drone control technology, and in particular to a drone path online planning method, device and computer equipment for real-time detection of scene update changes. Background Art

[0002] With the development of drone control technology, a technology has emerged that uses drones to reconstruct regional three-dimensional scenes. This technology can use drones to collect complete scene data with high precision and high coverage to construct an initial scene three-dimensional model of the scene area. After that, the drone can be set up for regular flight missions to collect data from the scene areas that have changed, so as to regularly update the initial scene three-dimensional model.

[0003] The UAV flight control mission set up for periodic updates of the initial scene 3D model depends on the planning of the UAV flight path. In traditional technologies, in the UAV flight mission used to update the scene 3D model, the UAV flight process usually conducts a comprehensive exploration and reconstruction of the complete scene area.

[0004] However, this drone flight control method will cause the drone to repeatedly fly over the complete scene area, resulting in a long flight path length for the drone, which in turn reduces the efficiency of updating the scene's three-dimensional model. Summary of the invention

[0005] Based on this, it is necessary to provide a method, device, computer equipment, computer-readable storage medium and computer program product for online drone path planning that can reduce the length of the drone flight path to improve the efficiency of updating the scene three-dimensional model and detect scene update changes in real time in order to solve the above-mentioned technical problems.

[0006] In a first aspect, the present application provides a method for online planning of a drone path for detecting scene update changes in real time, comprising:

[0007] Acquire an initial scene reconstruction model of the target area, and acquire first sampling points included in the initial scene reconstruction model, and first probabilities that each of the first sampling points is located in a scene change area in the target area;

[0008] A first candidate perspective set associated with the target area is obtained, and a first importance degree of each first candidate perspective included in the first candidate perspective set is obtained according to the first sampling point and the first probability, and a target perspective is obtained from the first candidate perspectives according to the first importance degree; the perspective set composed of the target perspectives is a perspective set with the smallest number of perspectives and the largest coverage range for the target area;

[0009] Generate a priori flight path of the UAV according to each of the target viewing angles; the priori flight path is used to control the estimated flight of the UAV in the target area;

[0010] When the UAV is flying in the target area according to the prior flight path, if a scene change area is detected, obtaining second sampling points and a second probability that each of the second sampling points is located in the scene change area;

[0011] Acquire a second candidate viewing angle set associated with the target area, obtain a second importance degree of each second candidate viewing angle included in the second candidate viewing angle set according to the second sampling point and the second probability, and acquire a target flight viewing angle from the second candidate viewing angles according to the second importance degree;

[0012] The priori flight path is adjusted according to the target flight angle to obtain a real-time flight path; the real-time flight path is used to control the real-time flight of the UAV in the target area to update the initial scene reconstruction model.

[0013] In one of the embodiments, obtaining the first importance degree of each first candidate perspective included in the first candidate perspective set according to the first sampling point and the first probability includes: obtaining a current first candidate perspective, and a target first sampling point observable by the current first candidate perspective; the current first candidate perspective is any one of the first candidate perspectives; obtaining a first sample change score corresponding to each target first sampling point according to the first probability corresponding to each target first sampling point; and obtaining a first importance degree of the current first candidate perspective based on each of the first sample change scores.

[0014] In one of the embodiments, obtaining a first sample change score corresponding to each of the target first sampling points according to the first probability corresponding to each of the target first sampling points includes: obtaining a current first sampling point and the number of first candidate viewing angles from which the current first sampling point can be observed; the current first sampling point is any one of the first sampling points; and obtaining a first sample change score corresponding to the current first sampling point according to the number and the first probability corresponding to the current first sampling point.

[0015] In one embodiment, obtaining the first sampling points included in the initial scene reconstruction model and the first probability that each of the first sampling points is located in the scene change area in the target area includes: obtaining the first sampling points included in the initial scene reconstruction model and the scene area type label corresponding to each of the first sampling points; obtaining the prior probabilities corresponding to each of the scene area type labels; the prior probabilities are used to characterize the possibility of scene changes occurring in the area corresponding to each of the scene area type labels; and using the prior probabilities corresponding to each of the first sampling points as the first probabilities corresponding to each of the first sampling points.

[0016] In one of the embodiments, obtaining the second importance degree of each second candidate perspective included in the second candidate perspective set according to the second sampling point and the second probability includes: obtaining the current second candidate perspective; the current second candidate perspective is any one of the second candidate perspectives; obtaining the second sample change score of each second sampling point corresponding to the current second candidate perspective according to the second sampling point and the second probability; obtaining the second importance degree of the current second candidate perspective according to the second sample change score of each second sampling point corresponding to the current second candidate perspective.

[0017] In one of the embodiments, obtaining the second sample change score of each second sampling point corresponding to the current second candidate perspective based on the second sampling point and the second probability includes: obtaining the current second sampling point and the visited perspectives of the drone during flight in the target area; the current second sampling point is any one of the second sampling points; obtaining visibility information of the current second sampling point based on the visited perspectives and the current second candidate perspective; the visibility information is used to characterize whether the current second sampling point can be observed by the current second candidate perspective and the visited perspective; obtaining the second sample change score of the current second sampling point corresponding to the current second candidate perspective based on the second probability and the visibility information.

[0018] In one of the embodiments, the obtaining of the second sampling points and the second probability that each of the second sampling points is located in the scene change area includes: obtaining the sampling points included in the scene change area, and using the newly added sampling points included in the scene change area and the first sampling points other than the newly added sampling points as the second sampling points; setting the second probability corresponding to the newly added sampling points to 1, and using the first probability corresponding to the first sampling points other than the newly added sampling points as the second probability corresponding to the first sampling points other than the newly added sampling points.

[0019] In one of the embodiments, obtaining the first candidate perspective set associated with the target area includes: performing Poisson disk sampling on the initial scene reconstruction model to obtain the first candidate perspective set; obtaining the second candidate perspective set associated with the target area includes: eliminating the visited perspectives of the UAV during the flight of the target area from the target perspective to obtain an unvisited perspective set; performing Poisson disk sampling on the scene change area to obtain a candidate visual set of the change area; and taking the union of the unvisited perspective set and the candidate visual set of the change area as the second candidate perspective set.

[0020] In a second aspect, the present application also provides a drone path online planning device for detecting scene update changes in real time, including:

[0021] A first sampling acquisition module, used to acquire an initial scene reconstruction model of a target area, and acquire first sampling points included in the initial scene reconstruction model, and a first probability that each of the first sampling points is located in a scene change area in the target area;

[0022] a target perspective screening module, configured to obtain a first candidate perspective set associated with the target area, obtain a first importance degree of each first candidate perspective included in the first candidate perspective set according to the first sampling point and the first probability, and obtain a target perspective from the first candidate perspectives according to the first importance degree; the perspective set composed of the target perspectives is a perspective set with the smallest number of perspectives and the largest coverage range for the target area;

[0023] A priori path generation module, used to generate a priori flight path of the UAV according to each of the target viewing angles; the priori flight path is used to control the estimated flight of the UAV in the target area;

[0024] A second sampling acquisition module is used to acquire second sampling points and a second probability that each of the second sampling points is located in the scene change area if a scene change area is detected during the UAV flying in the target area according to the prior flight path;

[0025] a flight perspective screening module, configured to obtain a second candidate perspective set associated with the target area, obtain a second importance degree of each second candidate perspective included in the second candidate perspective set according to the second sampling point and the second probability, and obtain a target flight perspective from the second candidate perspectives according to the second importance degree;

[0026] A real-time path generation module is used to adjust the prior flight path according to the target flight angle to obtain a real-time flight path; the real-time flight path is used to control the real-time flight of the UAV in the target area to update the initial scene reconstruction model.

[0027] In a third aspect, the present application further provides a computer device, comprising a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of the method described in any one of the embodiments of the first aspect when executing the computer program.

[0028] In a fourth aspect, the present application further provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps of the method described in any one of the embodiments of the first aspect.

[0029] In a fifth aspect, the present application further provides a computer program product, comprising a computer program, which, when executed by a processor, implements the steps of the method described in any one of the embodiments of the first aspect.

[0030] The above-mentioned method, device, computer equipment, storage medium and computer program product for online planning of drone paths for real-time detection of scene update changes obtain an initial scene reconstruction model of the target area, and obtain the first sampling points contained in the initial scene reconstruction model, and the first probability that each first sampling point is located in the scene change area in the target area; obtain a first candidate perspective set associated with the target area, and obtain the first importance degree of each first candidate perspective contained in the first candidate perspective set according to the first sampling point and the first probability, and obtain the target perspective from the first candidate perspective according to the first importance degree; the perspective set composed of the target perspective is the perspective set with the smallest number of perspectives and the largest coverage range for the target area; generate a A priori flight path of a UAV; the priori flight path is used to control the estimated flight of the UAV in a target area; when the UAV is flying in the target area according to the priori flight path, if a scene change area is detected, a second sampling point and a second probability that each second sampling point is located in the scene change area are obtained; a second candidate perspective set associated with the target area is obtained, and according to the second sampling point and the second probability, a second importance degree of each second candidate perspective included in the second candidate perspective set is obtained, and according to the second importance degree, a target flight perspective is obtained from the second candidate perspective; the priori flight path is adjusted according to the target flight perspective to obtain a real-time flight path; the real-time flight path is used to control the real-time flight of the UAV in the target area to update the initial scene reconstruction model.The present application can obtain a first probability that each first sampling point belongs to a scene change area according to an initial scene reconstruction model of the target area before the UAV flies, and obtain a first importance degree of each first candidate perspective in a first candidate perspective set associated with the target area in combination with the first probability, thereby screening out the target perspective according to the first importance degree, and the perspective set composed of the screened target perspectives can meet the conditions of a minimum number of perspectives and a maximum coverage range for the target area, and then generate a priori flight path according to the target perspective, so that the UAV can fly in the target area according to the prior flight path, and if a scene change area is detected during the flight, the second sampling point and the second probability that each second sampling point belongs to the scene change area can be obtained in real time, and then the second importance degree of each second candidate perspective in a second candidate perspective set associated with the target area can be obtained in combination with the first probability, thereby screening out the target flight perspective according to the second importance degree, Then, the prior flight path is adjusted according to the target flight perspective to obtain the real-time flight path, thereby controlling the real-time flight of the drone. Compared with the prior art drone flight that requires a comprehensive exploration of the complete scene area, the drone of the present application can first obtain the importance of each first candidate perspective based on the first probability that the first sampling point belongs to the scene change area before flying, thereby generating a priori flight path, and during the flight, when the scene change area is detected, the importance of each second candidate perspective can be obtained according to the second probability that the second sampling point belongs to the scene change area, thereby obtaining the target flight perspective to adjust the prior flight path to obtain the real-time flight path. In this way, the flight path can be generated according to the probability that the sampling point belongs to the scene change area, thereby avoiding the drone from repeatedly flying the complete scene area, and only conducting targeted detection on the area with a high probability of change, thereby reducing the length of the drone flight path and improving the efficiency of updating the scene three-dimensional model. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the related technologies, the drawings required for use in the embodiments or the related technical descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0032] Figure 1 A schematic diagram of a flow chart of an online drone path planning method for detecting scene update changes in real time in one embodiment;

[0033] Figure 2 A schematic diagram of a process for obtaining a first importance level in one embodiment;

[0034] Figure 3is a schematic diagram of a flow chart for obtaining a second importance level in one embodiment;

[0035] Figure 4 is a schematic diagram of a process for obtaining a second sample change score in one embodiment;

[0036] Figure 5 A schematic diagram of a viewing angle generation area in one embodiment;

[0037] Figure 6 A basic flow chart of a method for online planning of a drone path in one embodiment;

[0038] Figure 7 A flow chart of change region detection and extraction in one embodiment;

[0039] Figure 8 A structural block diagram of an online drone path planning device for detecting scene update changes in real time in one embodiment;

[0040] Fig. 9 FIG. 4 is a diagram showing the internal structure of a computer device in one embodiment. DETAILED DESCRIPTION

[0041] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0042] In one embodiment, Figure 1 As shown, a method for online planning of a drone path for real-time detection of scene update changes is provided. This embodiment uses the method applied to a terminal as an example. In this embodiment, the method includes the following steps:

[0043] Step S101 : obtaining an initial scene reconstruction model of a target area, and obtaining first sampling points included in the initial scene reconstruction model, and first probabilities that each first sampling point is located in a scene change area in the target area.

[0044] Among them, the target area refers to the area that needs to be reconstructed into a three-dimensional scene, and the initial scene reconstruction model refers to the scene reconstruction model generated by the current three-dimensional scene reconstruction for the target area. For example, it can be a scene reconstruction model generated by data collected by the previous UAV flight mission. The first sampling point refers to the sampling point obtained by sampling from the initial scene reconstruction model, and the first probability refers to the probability that the first sampling point is located in the scene change area in the target area, which can be the prior probability of the first sampling point for the scene change area before the UAV flies.

[0045] Specifically, before the UAV performs the current flight mission, the terminal can pre-acquire the initial scene reconstruction model of the target area, and sample the initial scene reconstruction model to obtain the first sampling point. At the same time, the probability that each first sampling point is located in the scene change area in the target area can be calculated as the first probability corresponding to each first sampling point.

[0046] Step S102, obtaining a first candidate perspective set associated with the target area, obtaining a first importance degree of each first candidate perspective contained in the first candidate perspective set according to the first sampling point and the first probability, and obtaining a target perspective from the first candidate perspectives according to the first importance degree; the perspective set composed of the target perspective is a perspective set with the smallest number of perspectives and the largest coverage range for the target area.

[0047] The first candidate perspective set refers to a perspective set composed of the first candidate perspectives, which refers to the candidate flight perspectives generated before the drone performs the current flight mission, and the first importance level refers to the importance of each first candidate perspective in the current drone flight mission. The greater the importance level, the higher the priority of being selected as the target perspective, and the target perspective is the perspective selected by the first candidate perspective to reduce the redundancy of the perspective. In other words, the screening goal of the target perspective is to reduce the number of perspectives as much as possible while ensuring maximum coverage, that is, the perspective set composed of the target perspective has the smallest number of perspectives and the largest coverage of the target area.

[0048] Specifically, the terminal may also obtain a first candidate perspective set associated with the target area, and may combine the first sampling points acquired in step S101 and the first probabilities corresponding to the first sampling points to calculate the importance of each first candidate perspective contained in the first candidate perspective set, that is, the first importance. Afterwards, the terminal may further obtain the target perspective from the first candidate perspectives according to the importance of the first candidate perspectives, so that the perspective set composed of the screened target perspectives can meet the minimum number of perspectives and the maximum coverage of the target area, thereby effectively reducing the redundancy of perspectives.

[0049] Step S103, generating a priori flight path of the UAV according to each target viewing angle; the priori flight path is used to control the estimated flight of the UAV in the target area.

[0050] The prior flight path refers to the flight path generated in advance before the drone performs the current flight mission. The flight path can represent the estimated flight trajectory of the drone in the target area. Specifically, after obtaining multiple target perspectives, the multiple target perspectives can be connected to generate the prior flight path for the drone to perform the current flight mission.

[0051] Step S104 , when the UAV is flying in the target area according to the prior flight path, if a scene change area is detected, second sampling points and a second probability that each second sampling point is located in the scene change area are obtained.

[0052] The second sampling point is the sampling point used to generate the real-time flight path when the drone detects a scene change area during the current flight mission. The second probability refers to the probability that the second sampling point is located in the scene change area in the target area. Specifically, after generating the prior flight path, the drone can fly in the target area according to the prior flight path. If a scene change area is detected during the flight, the terminal can also correct the prior flight path, that is, re-acquire the second sampling point and the second probability of each second sampling point being located in the scene change area.

[0053] Step S105, obtaining a second candidate perspective set associated with the target area, obtaining a second importance degree of each second candidate perspective contained in the second candidate perspective set according to the second sampling point and the second probability, and obtaining the target flight perspective from the second candidate perspective according to the second importance degree.

[0054] The second candidate perspective set refers to a perspective set composed of second candidate perspectives. The second candidate perspective refers to the candidate flight perspective generated after the scene change area is detected when the UAV performs the current flight mission. The second importance level refers to the importance of each second candidate perspective in the current UAV flight mission. The greater the importance level, the higher the priority of being selected as the target flight perspective. The target flight perspective is the perspective selected by the second candidate perspective, which can be used as the next perspective for the UAV's real-time flight.

[0055] Specifically, after detecting a scene change area during the flight of the drone, the terminal can also obtain a second candidate perspective set associated with the target area, and can combine the second sampling points collected in step S104 and the second probability corresponding to each second sampling point to calculate the importance of each second candidate perspective included in the second candidate perspective set, that is, the second importance. Afterwards, the terminal can further obtain the next perspective of the drone's real-time flight from the second candidate perspective according to the importance of the second candidate perspective, that is, obtain the target flight perspective.

[0056] Step S106, adjusting the prior flight path according to the target flight viewing angle to obtain a real-time flight path; the real-time flight path is used to control the real-time flight of the UAV in the target area to update the initial scene reconstruction model.

[0057] The real-time flight path refers to the flight path that is corrected from the prior flight path when the scene change area is detected during the flight of the drone. This path can be used to control the real-time flight of the drone, thereby updating the initial scene reconstruction model. Specifically, after obtaining the target flight angle, the prior flight path of the drone can be adjusted to obtain the real-time flight path. The drone can then fly according to the corrected real-time flight path, thereby updating the initial scene reconstruction model.

[0058] For example, before the drone flies, the generated prior flight path can be: perspective A-perspective B-perspective C-perspective D. When the drone flies according to the prior flight path, it can be after flying from perspective A to perspective B. If a scene change area is detected in perspective B, the target flight perspective can be obtained as the next flight perspective, such as perspective E. At this time, the flight path will be corrected to the real-time flight path, that is, perspective A-perspective B-perspective E.

[0059] In the above-mentioned drone path online planning method for real-time detection of scene update changes, an initial scene reconstruction model of the target area is obtained, and the first sampling points contained in the initial scene reconstruction model and the first probability that each first sampling point is located in the scene change area in the target area are obtained; a first candidate perspective set associated with the target area is obtained, and according to the first sampling point and the first probability, a first importance degree of each first candidate perspective contained in the first candidate perspective set is obtained, and according to the first importance degree, a target perspective is obtained from the first candidate perspective; the perspective set composed of the target perspective is a perspective set with the smallest number of perspectives and the largest coverage range for the target area; according to each target perspective, a priori flight path of the drone is generated; The prior flight path is used to control the estimated flight of the UAV in the target area; when the UAV flies in the target area according to the prior flight path, if a scene change area is detected, a second sampling point and a second probability that each second sampling point is located in the scene change area are obtained; a second candidate perspective set associated with the target area is obtained, and a second importance degree of each second candidate perspective included in the second candidate perspective set is obtained according to the second sampling point and the second probability, and a target flight perspective is obtained from the second candidate perspective according to the second importance degree; the prior flight path is adjusted according to the target flight perspective to obtain a real-time flight path; the real-time flight path is used to control the real-time flight of the UAV in the target area to update the initial scene reconstruction model.The present application can obtain a first probability that each first sampling point belongs to a scene change area according to an initial scene reconstruction model of the target area before the UAV flies, and obtain a first importance degree of each first candidate perspective in a first candidate perspective set associated with the target area in combination with the first probability, thereby screening out the target perspective according to the first importance degree, and the perspective set composed of the screened target perspectives can meet the conditions of a minimum number of perspectives and a maximum coverage range for the target area, and then generate a priori flight path according to the target perspective, so that the UAV can fly in the target area according to the prior flight path, and if a scene change area is detected during the flight, the second sampling point and the second probability that each second sampling point belongs to the scene change area can be obtained in real time, and then the second importance degree of each second candidate perspective in a second candidate perspective set associated with the target area can be obtained in combination with the first probability, thereby screening out the target flight perspective according to the second importance degree, Then, the prior flight path is adjusted according to the target flight perspective to obtain the real-time flight path, thereby controlling the real-time flight of the drone. Compared with the prior art drone flight that requires a comprehensive exploration of the complete scene area, the drone of the present application can first obtain the importance of each first candidate perspective based on the first probability that the first sampling point belongs to the scene change area before flying, thereby generating a priori flight path, and during the flight, when the scene change area is detected, the importance of each second candidate perspective can be obtained according to the second probability that the second sampling point belongs to the scene change area, thereby obtaining the target flight perspective to adjust the prior flight path to obtain the real-time flight path. In this way, the flight path can be generated according to the probability that the sampling point belongs to the scene change area, thereby avoiding the drone from repeatedly flying the complete scene area, and only conducting targeted detection on the area with a high probability of change, thereby reducing the length of the drone flight path and improving the efficiency of updating the scene three-dimensional model.

[0060] In one embodiment, Figure 2 As shown, step S102 may further include:

[0061] Step S201, obtaining a current first candidate viewing angle and a target first sampling point observable by the current first candidate viewing angle; the current first candidate viewing angle is any one of the first candidate viewing angles.

[0062] The current first candidate perspective refers to any one of the first candidate perspectives included in the first candidate perspective set, and the target first sampling point refers to the first sampling point observable by the current first candidate perspective. In this embodiment, the terminal may pre-acquire the visibility relationship between each first candidate perspective and the first sampling point. The visibility relationship may characterize whether each first candidate perspective may observe a first sampling point. Therefore, based on the above visibility relationship, the first sampling point observable by the current first candidate perspective may be determined as the target first sampling point.

[0063] Step S202, obtaining a first sample change score corresponding to each target first sampling point according to the first probability corresponding to each target first sampling point;

[0064] Step S203: obtaining the first importance degree of the current first candidate perspective based on the first sample change scores.

[0065] The first sample change score refers to the sample change score corresponding to the first sampling point of each target, which can be used to characterize the possibility of sample change. In this embodiment, after the first probability corresponding to the first sampling point of each target, the first sample change score corresponding to each first sampling point of each target can also be calculated based on the above first probability, and then the first sample change score is used to obtain the first importance degree of the current first candidate perspective.

[0066] For example, the candidate perspective With sampling point The visibility relationship between can be expressed by the following formula:

[0067]

[0068] Current first candidate perspective The first importance It can be calculated by the following formula:

[0069]

[0070] in, represents the first sample change fraction corresponding to each first sampling point, and Indicates the first sampling point The set of sample points composed of . When it is 1, Can be When the first sample change score is observed, the first importance degree is calculated using the first sample change score, that is, only the first sample change score corresponding to the target first sampling point observable by the current first candidate perspective is used to calculate the importance degree of the current first candidate perspective.

[0071] In this embodiment, the target first sampling point observable by each first candidate perspective can also be determined, and the corresponding first sample change score can be further obtained according to the first probability corresponding to the target first sampling point, so as to obtain the first importance degree of each first candidate perspective. In this way, the accuracy of obtaining the first importance degree can be improved.

[0072] Furthermore, step S202 may further include: obtaining the current first sampling point and the number of first candidate perspectives from which the current first sampling point can be observed; the current first sampling point is any one of the first sampling points; and obtaining a first sample change score corresponding to the current first sampling point according to the number and the first probability corresponding to the current first sampling point.

[0073] The current first sampling point refers to any one of the first sampling points. In this embodiment, since the terminal can obtain the visibility relationship between each first candidate perspective and the first sampling point in advance, the number of first candidate perspectives that can observe the current first sampling point can also be obtained based on the visibility relationship. For example, if the first sampling point 1 can be observed by the first candidate perspective A and the first candidate perspective B, then the number of first candidate perspectives that can observe the first sampling point 1 is 2. Afterwards, the first sample change score corresponding to the current first sampling point can be calculated using this number and the first probability corresponding to the current first sampling point.

[0074] For example, the current first sampling point The corresponding first sample change score It can be calculated by the following formula:

[0075]

[0076] in, Indicates that the current first sampling point is observed The number of first candidate perspectives, and It can be calculated by the following formula:

[0077]

[0078] in, represents the scaling parameter, Indicates the current first sampling point The corresponding first probability is Indicates the current first sampling point With each first candidate perspective The visibility relationship between them.

[0079] In this embodiment, the first sample change score corresponding to the current first sampling point can also be obtained according to the number of first candidate viewing angles of the current first sampling point and the first probability corresponding to the current first sampling point. This method can ensure that the possibility of sample change changes with the number of viewing angles at which the sample is observed. For example, the possibility of sample change gradually decreases as the sample is observed by more viewing angles, thereby further improving the accuracy of obtaining the first sample change score.

[0080] In addition, step S101 may further include: obtaining the first sampling points included in the initial scene reconstruction model, and the scene area type labels corresponding to each first sampling point; obtaining the prior probabilities corresponding to each scene area type label; the prior probabilities are used to characterize the possibility of scene changes occurring in the areas corresponding to each scene area type label; and using the prior probabilities corresponding to each first sampling point as the first probabilities corresponding to each first sampling point.

[0081] The scene area type label can characterize the type of the scene area where each first sampling point is located, such as a road or a building, etc. The prior probability refers to the possibility probability of a scene change occurring in the area corresponding to the scene area type label. In this embodiment, a correspondence between each scene type label and the prior probability can be pre-constructed, and the correspondence can be used to obtain the first probability corresponding to each first sampling point.

[0082] Specifically, the terminal can first sample the initial scene reconstruction model to obtain the first sampling point contained in the initial scene reconstruction model, and then identify the scene area type label corresponding to the first sampling point, and combine the prior probabilities corresponding to each scene area type label to obtain the prior probability corresponding to each first sampling point, and use it as the first probability corresponding to each first sampling point.

[0083] For example, a priori probability 1 corresponding to a road label and a priori probability 2 corresponding to a building label may be pre-constructed. When the scene area type label corresponding to the first sampling point 1 is a road label, the first probability corresponding to the first sampling point is the priori probability 1. If the scene area type label corresponding to the first sampling point 1 is a building label, the first probability corresponding to the first sampling point is the priori probability 2.

[0084] In this embodiment, the correspondence between the scene area type label and the prior probability can also be pre-constructed, so that the first probability corresponding to the first sampling point can be obtained according to the scene area type label corresponding to the first sampling point. In this way, the accuracy of the prior probability calculation can be improved.

[0085] In one embodiment, Figure 3 As shown, step S105 may further include:

[0086] Step S301, obtaining a current second candidate viewing angle; the current second candidate viewing angle is any one of the second candidate viewing angles;

[0087] Step S302: Obtain, according to the second sampling points and the second probability, a second sample change score of each second sampling point corresponding to the current second candidate viewing angle.

[0088] The current second candidate perspective refers to any one of the second candidate perspectives included in the second candidate perspective set, and the second sample change score refers to the sample change score of each second sampling point corresponding to the current second candidate perspective, which can be used to characterize the possibility of sample change.

[0089] Specifically, the terminal may also use any one of the second candidate perspectives as the current second candidate perspective, and calculate the second sample change score of each second sampling point corresponding to the current second candidate perspective by combining each second sampling point and the corresponding second probability.

[0090] Step S303: obtaining a second importance degree of the current second candidate perspective according to a second sample change score of each second sampling point corresponding to the current second candidate perspective.

[0091] After obtaining the second sample change score corresponding to the current second candidate perspective for each second sampling point, the second sample change score may be further summed to obtain the second importance degree of the current second candidate perspective.

[0092] Current second candidate perspective The second level of importance It can be calculated by the following formula:

[0093]

[0094] in, Indicates the second sampling point Corresponding to the current second candidate perspective The second sample change fraction, while Indicates the second sampling point It can be seen that in the embodiment, the second importance degree of the current second candidate perspective can be obtained by summing up the second sample change scores of each second sampling point corresponding to the current second candidate perspective.

[0095] In this embodiment, the second sample change scores of each second sampling point corresponding to the current second candidate perspective can also be calculated through each second sampling point and the second probability, so as to obtain the second importance degree of the current second candidate perspective using each second sample change score. In this way, the accuracy of the second importance degree calculation can be improved.

[0096] Furthermore, if Figure 4 As shown, step S302 may further include:

[0097] Step S401, obtaining the current second sampling point and the visited viewing angles of the UAV during the flight of the target area; the current second sampling point is any one of the second sampling points.

[0098] The current second sampling point refers to any one of the second sampling points, and the visited perspective refers to the perspective that the drone has visited during the flight of the target area. Taking the drone's prior flight path as perspective A-perspective B-perspective C-perspective D as an example, when the drone is flying according to the prior flight path, if it flies from perspective A to perspective B and detects a scene change area, then the visited perspectives can include perspective A and perspective B. If the target flight perspective, that is, the next flight perspective, is perspective E, the flight path will be corrected to the real-time flight path, that is, perspective A-perspective B-perspective E, then the visited perspectives include perspective A, perspective B and visual E.

[0099] Specifically, the terminal may use any one of the second sampling points as the current second sampling point, and may determine the visited viewing angles of the drone during the flight of the target area.

[0100] Step S402: Obtain visibility information of the current second sampling point according to the visited viewing angle and the current second candidate viewing angle; the visibility information is used to indicate whether the current second sampling point can be observed by the current second candidate viewing angle and the visited viewing angle.

[0101] The visibility information of the current second sampling point refers to information used to characterize whether the current second sampling point can be observed by the current second candidate view and the visited view. The information can be obtained based on the visibility relationship between the current second sampling point and the candidate view, that is, it can include the visibility relationship between the current second sampling point and the current second candidate view, and the visibility relationship between the current second sampling point and each visited view.

[0102] Specifically, after each visited viewing angle and the current second candidate viewing angle are determined, the visibility relationship between the viewing angle and the current second sampling point can be obtained, thereby obtaining the visibility information of the current second sampling point.

[0103] Step S403: Obtain a second sample change score of the current second sampling point corresponding to the current second candidate viewing angle according to the second probability and the visibility information.

[0104] Finally, the terminal may combine the second probability corresponding to the current second sampling point and the above visibility information to calculate a second sample change score of the current second sampling point corresponding to the current second candidate viewing angle.

[0105] For example, the current second sampling point Corresponding to the current second candidate perspective The second sample change score It can be calculated by the following formula:

[0106]

[0107] in, Indicates the current second sampling point Corresponding to the current second candidate perspective that has not been visited The sample change fraction of Indicates the current second sampling point Corresponding to each visited perspective The sample change score is composed of these two parts.

[0108] and, It can be calculated by the following formula:

[0109]

[0110] in, represents the scaling parameter, Indicates the current second sampling point The corresponding second probability is Indicates the current second sampling point With the current second candidate perspective The visibility relationship between them.

[0111] and It can be calculated by the following formula:

[0112]

[0113] in, represents the scaling parameter, and Indicates the current second sampling point With any visited perspective The visibility relationship between them.

[0114] In this embodiment, the visibility relationship between the current second sampling point and each visited viewpoint, and the visibility relationship between the current second sampling point and the current second candidate viewpoint can be obtained as the visibility information of the current second sampling point, so as to combine the visibility information of the current second sampling point with the second probability corresponding to the current second sampling point to calculate the second sample change score of the current second sampling point corresponding to the current second candidate viewpoint. In this way, the accuracy of the calculation of the second sample change score can be further improved.

[0115] In addition, step S104 may further include: acquiring sampling points included in the scene change area, and using the newly added sampling points included in the scene change area and the first sampling points other than the newly added sampling points as second sampling points; setting the second probability corresponding to the newly added sampling points to 1, and using the first probability corresponding to the first sampling points other than the newly added sampling points as the second probability corresponding to the first sampling points other than the newly added sampling points.

[0116] In this embodiment, the second sampling point may be composed of two parts, namely, a newly added sampling point obtained by sampling the scene change area after the scene change area is detected during the flight of the drone, and a part of the first sampling point other than the newly added sampling point, and the second probabilities corresponding to the two parts of the sampling points are different, wherein since the newly added sampling point must be located in the scene change area, the second probability corresponding to the part of the sampling points is set to 1, and for the first sampling point other than the newly added sampling point, the first probability corresponding to the first sampling point is used as the second probability of the second sampling point of the part.

[0117] For example, the second sampling point The second probability It can be represented by the following formula:

[0118]

[0119] in, is the prior probability, that is, the first probability.

[0120] In this embodiment, the newly added sampling points included in the scene change area and the first sampling points other than the newly added sampling points can be used as second sampling points, and the second probability of the newly added sampling points for is 1, while the second probability of the first sampling points other than the newly added sampling points is maintained as the original first probability. In this way, the accuracy of obtaining the second sampling points and the second probability can be improved.

[0121] In one embodiment, step S102 may further include: performing Poisson disk sampling on the initial scene reconstruction model to obtain a first candidate perspective set; step S105 may further include: eliminating the visited perspectives of the drone during the flight of the target area from the target perspective to obtain an unvisited perspective set; performing Poisson disk sampling on the scene change area to obtain a candidate visual set of the change area; and taking the union of the unvisited perspective set and the candidate visual set of the change area as the second candidate perspective set.

[0122] In this embodiment, the first candidate perspective set can be obtained by performing Poisson disk sampling on the initial scene reconstruction model, and the second candidate perspective set can be composed of the union of two sets, namely, the unvisited perspective set and the candidate visual set of the changed area. Among them, the unvisited perspective set refers to a set composed of unvisited perspectives, which can be obtained by eliminating the visited perspectives of the drone during the flight of the target area from the target perspective. The candidate visual set of the changed area is obtained by performing Poisson disk sampling on the scene change area.

[0123] Continuing to take the drone's prior flight path as perspective A-perspective B-perspective C-perspective D as an example, when the drone is flying according to the prior flight path, if it flies from perspective A to perspective B and detects a scene change area, the target perspective may include perspective A, perspective B, perspective C, and perspective D, where the visited perspectives are perspective A and perspective B, and the unvisited perspective set includes perspective C and perspective D. In addition, the terminal may also perform Poisson disk sampling on the scene change area to obtain a candidate visual set of the change area, for example, it may include perspective C and perspective E, then the second candidate perspective set may be the union of the above two perspective sets, that is, the included second candidate perspectives may include perspective C, perspective D, and perspective E.

[0124] In this embodiment, Poisson disk sampling can be performed on the initial scene reconstruction model to obtain a first candidate perspective set, and the union of the unvisited perspective set and the changed area candidate visual set can be used as the second candidate perspective set. In this way, the integrity of the candidate perspective screening can be ensured.

[0125] In one embodiment, a method for online planning of drone paths for scene updates is also provided. This method introduces a heuristic variability method, based on previous scene reconstruction data and change probability statistics, and gives priority to exploring areas that may change. This can effectively avoid repeated exploration of static areas and significantly reduce flight time and computational overhead. Rapid detection and response to changes in the urban environment is achieved. By combining the detection of changed areas with real-time path planning, the system can dynamically adjust the flight trajectory to ensure that only the changed areas are explored and updated, thereby improving the real-time and accuracy of change detection. In addition, a dual strategy of prior paths and real-time paths is adopted, combined with statistical prior data and real-time change detection, to ensure that when updating three-dimensional scenes, the efficiency of information acquisition can be maximized and unnecessary data collection can be reduced. The principle of this method is as follows:

[0126] First, the core of this embodiment is to introduce a changeability heuristic to guide trajectory planning. By analyzing previously reconstructed data (such as T 1 The system can evaluate the possibility of change in each area based on the three-dimensional scene at the moment of time and the probability of change statistics. This heuristic variability calculates the probability of change in each area based on historical data and scene statistics, thus providing a basis for trajectory planning. The system divides the trajectory into two parts: the prior path and the real-time path. The prior path is based on T 1 The reconstructed model is used for planning, which reduces unnecessary flight time and computing resource consumption by prioritizing the exploration of areas with a higher probability of change and avoiding repeated exploration of static areas.

[0127] When the drone flies to the prior path, if a change area is found, real-time path planning will be triggered. Real-time path planning can dynamically adjust the path according to the image differences captured during the flight, give priority to exploring the change area, and continue to adjust the path through online updates to ensure full coverage of the change area. In the process of exploring the change area, the system combines image feature matching and 3D reconstruction methods to generate accurate point clouds of the change area, and uses the convex hull algorithm to extract the geometric information of the change area for subsequent 3D reconstruction and update.

[0128] To further improve efficiency, real-time path planning also takes into account the redundancy of explored areas, and ensures that changing areas are fully explored without wasting time on static areas by reducing the acquisition of repeated perspectives. During the real-time path planning process, the system continuously updates the target area and uses heuristic variability to select the next optimal perspective to achieve efficient exploration of changing areas.

[0129] This embodiment can be specifically implemented through the following process:

[0130] The candidate viewpoints of the prior path and the real-time path are designed to fully observe all areas of the target surface T on the predefined safe height plane h. For the prior path, T corresponds to T 1 For the real-time path, T represents the convex hull of M detected change areas, which is empty at the beginning and gradually expands if there is a change.

[0131] The region for generating new candidate views is generated by extending a padding margin on the target T This filling ensures sufficient coverage of all sampling points along the edge of the target convex hull. Within this extended area, Poisson disk sampling is used to generate candidate viewing angles. , where each candidate perspective Contains its 3D position and orientation. To accommodate drones equipped with multiple cameras (e.g., five cameras) or single-camera systems, five candidate views are generated at the same location, each with a different orientation. However, this introduces redundancy, so it is necessary to prioritize and streamline these views. For the prior path, the goal is to reduce redundancy while maintaining coverage. For the real-time path, the focus is on selecting the next best view to explore the entire changing area. To achieve this goal, this embodiment defines a heuristic method to quantify the importance of each view to guide the two path planning tasks.

[0132] like Figure 5 As shown in Figure 2, given that the main goal is to detect changes rather than perform complete reconstruction, in the experiment, all candidate viewpoints are generated on a safe height plane set to h = 120 meters. The viewpoint generation area is determined by extending a padding margin around the target T, as shown in Figure 2. Figure 5 This padding ensures that the generated viewpoints can fully cover all sample points on the edge of the target T. .Padding size The camera's field of view and And the safety height h is calculated, the formula is , where d is a small constant to ensure adequate coverage. In the configuration of this embodiment, d is set to the sampling radius used in the Poisson disk sampling step.

[0133] The goal of this heuristic method is to prioritize the detection of surface areas with higher probability of change. The target surface T is uniformly discretized into N sample points , where each sample point With a score are related, and:

[0134]

[0135] If a sample is confirmed to be part of the change area in real-time path planning, its score , indicating that its change possibility is the highest. Otherwise, are the prior probabilities. These probabilities are calculated using the statistics of the WUSU dataset, which includes 12 categories of urban structures. By analyzing the change rates of these categories, they are mapped to the 7 semantic labels in the UrbanBIS dataset.

[0136] Visibility Function Determine a sample point From the perspective Visible in:

[0137]

[0138] A sample point The coverage of is the number of perspectives from which the point is observed:

[0139]

[0140] For real-time path planning, the view V is divided into a visit sequence and unvisited sequences The next best view is from In order to maximize the possibility of change, all viewpoints are initially unvisited. .

[0141] A certain perspective The importance of is defined by the probability of variation of the sample it observes. To do this, we first define the sample Relative to viewing angle The possible changes are:

[0142]

[0143] in and are scaling parameters that control the contribution of unvisited and visited perspectives, respectively. The negative factor associated with visited perspectives reflects diminishing marginal returns: as a sample is observed by more visited perspectives, its probability of change gradually stabilizes. This is consistent with the idea that once a sample is observed, its "changed" or "unchanged" status is gradually confirmed, and each additional observation provides less new information. Therefore, the probability of change decreases, highlighting the inefficiency of redundant observations.

[0144] The overall process can be as follows Figure 6 As shown, the input is a labeled T 1Reconstruction model, where each semantic label is associated with a prior probability. First, the prior path is planned by minimizing redundancy while maximizing potential change possibility and coverage. When the drone flies along the prior path and detects a change area, real-time path planning is triggered to explore the entire target change area by analyzing the difference image. The output of the system is a convex hull set containing all detected change areas. Specifically, it may include the following parts:

[0145] (1) Prior path planning:

[0146] From T 1 Reconstruction scene To begin, export the sample set , which represents the discretized target point associated with the probability of change. Using these inputs, a dense set of candidate views V is generated as described above. The main goal of a priori path planning is to identify a minimal subset of views , while reducing redundancy and ensuring maximum coverage . This goal is inspired by the max-min optimization framework. The trajectories connecting these selected views form the prior aerial path. To this end, an iterative optimization process is adopted to optimize the view set by removing redundant views and maximizing the target coverage. This optimization problem can be formulated as:

[0147]

[0148] The first term minimizes redundancy and the second term maximizes coverage. The perspective set obtained after solving this optimization problem is It is also used as the initial candidate set for real-time path planning. In the whole process, all viewpoints are considered unvisited, i.e. ,and .

[0149] Redundancy is defined based on the probability of variation. A perspective is considered redundant if the samples observed by the perspective (1) are already well covered by other perspectives, or (2) have a low probability of variation, resulting in less importance of its observation. Therefore, the redundancy of a perspective can be quantified as the negative of its heuristic importance, i.e. , the higher the redundancy, the lower the importance. The heuristic importance of a perspective is derived based on the likelihood of change of the samples it observes. Sample The possibility of change is based on the set of perspectives from which it is observed. The change probability is defined as:

[0150]

[0151] Among them, for any have This formula ensures that the probability of a sample changing decreases as it is observed from more perspectives. Then, the perspective The heuristic importance of is calculated based on the samples it observes:

[0152]

[0153] To construct a continuous trajectory, visit All selected perspectives in the problem are framed as the Traveling Salesman Problem (TSP). The solution ensures an efficient flight path that preferentially explores areas with higher probability of change based on previous statistics.

[0154] (2) Change area detection and extraction:

[0155] like Figure 7 As shown in Figure 1, this is an example of a real-time path, applied to a scene in UrbanBIS. 1 The reconstruction model (top left) shows three existing buildings as a priori, while T 2 The model (bottom left) serves as ground truth, where one of the buildings has been demolished. When the drone detects changes in the scene, it dynamically adjusts its path (center) by identifying differences in the captured 2D images. Subsequently, a cloud of detected points (bottom right) and a convex hull of the changed area (top right) are generated based on the updated real-time path.

[0156] In T 2 When the drone follows the planned path Execute in sequence. For each specified view in this sequence The drone will look in the direction of Capturing a 2D image Then, each captured image is compared with the 1 Corresponding rendered images generated by the reconstructed model Compare to ensure that perspective and direction are consistent.

[0157] In order to detect and identify the changed regions at the image level, a learning-based image feature matching model DKM is used and enhanced by a generalized self-training framework GIM. This framework is pre-trained using large-scale Internet videos to ensure robustness in a variety of scenarios. , the area without feature matching is identified as the change area .

[0158] To extract the corresponding 3D information, a pre-trained multi-view stereo (MVS) reconstruction method DUSt3R is applied. Used as a mask to exclude unchanged areas, resulting in a 3D point cloud representing the identified changed areas When multiple changed regions are detected, they are added to a processing queue and processed sequentially.

[0159] To ensure the performance of using DUSt3R during path planning, The reconstruction includes only the sequences from the most recent view However, this limitation may result in incomplete point clouds. In order to use the newly generated point cloud data Enhance the current target area , their overlap is evaluated by the Intersection over Union (IoU) metric. ,in is a configurable similarity threshold, then the target region is updated to Otherwise, the target area will be replaced by the new point cloud. .

[0160] At each viewing angle, extract the detected point cloud The convex hull is used as the detected change area. This convex hull is used as the target area for further exploration , providing information for real-time path planning decisions and dynamically determining the next best viewing angle.

[0161] (3) Real-time path planning in changing areas

[0162] Once the change area is identified , the system will switch to real-time planning mode to dynamically and comprehensively explore the area. If multiple change areas are detected, the system will prioritize the change area closest to the current position of the drone as the next exploration target. Given a sequence of visited viewpoints and its corresponding change targets The goal of real-time planning is to choose the next best view (if any). , maximize the As the exploration progresses, the process dynamically updates the target area to .

[0163] Similar to prior path planning, first the change area Take samples and get ,in .set up is the new unvisited viewpoint to be added to the track. In order to evaluate the contribution of a candidate viewpoint, we first consider the possibility of changes in all samples that can be observed by this viewpoint. , relative to the candidate next perspective , considering the visited view sequence , and its change possibility is defined as:

[0164]

[0165] Candidate perspectives The total change likelihood gain (or heuristic importance) of is calculated as the additional information it provides beyond that obtained from the sequence of visited views:

[0166]

[0167] Candidate perspectives From two sources: (1) From the a priori path unvisited perspectives; (2) a real-time perspective set generated using Poisson disk sampling . Combine these two sets and filter out the perspectives where at least one sample can be observed :

[0168]

[0169] For each , calculate its change probability gain and process it in parallel. In order to prevent extreme trajectory deviation, the system selects the top K candidate perspectives with the highest change probability gain and selects the first K candidate perspectives that are closest to the current perspective. The closest one is taken as the next best view. In the experiment, K is set to 10. Then, the real-time path is constructed by sequentially connecting the visited paths with the selected next best view. All visited perspectives When observed, If multiple change regions are detected, the system will proceed to the next closest target. , and repeat the process.

[0170] After all detected change regions have been explored, the system will revisit any remaining unobserved regions. Unvisited perspectives from prior paths The views are filtered to include only those that can observe unvisited samples. These views are then connected into a continuous trajectory through the Traveling Salesman Problem (TSP) formulation, just like in the a priori path planning stage, to ensure effective coverage of the remaining scenes. By dynamically exploring the identified changed areas and revisiting the unexplored areas, the real-time path planning strategy ensures the comprehensiveness and efficiency of scene updates.

[0171] This embodiment solves the inefficiency problem of existing technologies in urban scene updating by combining prior probability and real-time detection of change areas. The key idea is to use T in prior path planning. 1The reconstruction model and change probability statistics of each moment are used to evaluate the possibility of change in each area, and the initial flight path is planned based on this. In real-time path planning, when a change area is detected, a new flight path is dynamically generated by analyzing the difference image to thoroughly explore the target change area. This method achieves more efficient and accurate urban scene updates by reducing unnecessary flight time and computational overhead. Specifically, the prior path planning optimizes the initial flight path by minimizing redundancy and maximizing coverage of potential change areas, while the real-time path planning ensures a comprehensive exploration of the detected change areas by selecting the next best viewing angle. In addition, by combining statistical priors with real-time decision-making, the present invention achieves high-quality scene updates while significantly reducing resource consumption, making it more scalable and effective in large-scale complex urban environments.

[0172] It should be understood that, although the various steps in the flowcharts involved in the above-mentioned embodiments are displayed in sequence according to the indication of the arrows, these steps are not necessarily executed in sequence according to the order indicated by the arrows. Unless there is a clear explanation in this article, the execution of these steps does not have a strict order restriction, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above-mentioned embodiments can include multiple steps or multiple stages, and these steps or stages are not necessarily executed at the same time, but can be executed at different times, and the execution order of these steps or stages is not necessarily to be carried out in sequence, but can be executed in turn or alternately with other steps or at least a part of the steps or stages in other steps.

[0173] Based on the same inventive concept, the embodiment of the present application also provides a drone path online planning device for real-time detection of scene update changes for implementing the drone path online planning method for real-time detection of scene update changes involved in the above. The implementation scheme for solving the problem provided by the device is similar to the implementation scheme recorded in the above method, so the specific limitations in one or more drone path online planning devices for real-time detection of scene update changes provided below can be found in the above limitations on the drone path online planning method for real-time detection of scene update changes, and will not be repeated here.

[0174] In one embodiment, Figure 8 As shown, a drone path online planning device for real-time detection of scene update changes is provided, comprising: a first sampling acquisition module 801, a target perspective screening module 802, a priori path generation module 803, a second sampling acquisition module 804, a flight perspective screening module 805 and a real-time path generation module 806, wherein:

[0175] A first sampling acquisition module 801 is used to acquire an initial scene reconstruction model of the target area, and acquire first sampling points included in the initial scene reconstruction model, and a first probability that each first sampling point is located in a scene change area in the target area;

[0176] The target perspective screening module 802 is used to obtain a first candidate perspective set associated with the target area, obtain a first importance degree of each first candidate perspective included in the first candidate perspective set according to the first sampling point and the first probability, and obtain a target perspective from the first candidate perspectives according to the first importance degree; the perspective set composed of the target perspective is a perspective set with the smallest number of perspectives and the largest coverage range for the target area;

[0177] The prior path generation module 803 is used to generate a priori flight path of the UAV according to each target viewing angle; the priori flight path is used to control the estimated flight of the UAV in the target area;

[0178] The second sampling acquisition module 804 is used to acquire second sampling points and a second probability that each second sampling point is located in the scene change area if a scene change area is detected during the UAV flying in the target area according to the prior flight path;

[0179] The flight perspective screening module 805 is used to obtain a second candidate perspective set associated with the target area, obtain a second importance degree of each second candidate perspective included in the second candidate perspective set according to the second sampling point and the second probability, and obtain a target flight perspective from the second candidate perspective according to the second importance degree;

[0180] The real-time path generation module 806 is used to adjust the prior flight path according to the target flight angle to obtain the real-time flight path; the real-time flight path is used to control the real-time flight of the UAV in the target area to update the initial scene reconstruction model.

[0181] In one embodiment, the target perspective screening module 802 is further used to obtain the current first candidate perspective and the target first sampling point observable by the current first candidate perspective; the current first candidate perspective is any one of the first candidate perspectives; according to the first probability corresponding to each target first sampling point, the first sample change score corresponding to each target first sampling point is obtained; based on each first sample change score, the first importance degree of the current first candidate perspective is obtained.

[0182] In one embodiment, the target perspective screening module 802 is further used to obtain the current first sampling point and the number of first candidate perspectives at which the current first sampling point can be observed; the current first sampling point is any one of the first sampling points; and according to the number and the first probability corresponding to the current first sampling point, a first sample change score corresponding to the current first sampling point is obtained.

[0183] In one embodiment, the first sampling acquisition module 801 is further used to obtain the first sampling points included in the initial scene reconstruction model, and the scene area type labels corresponding to each first sampling point; obtain the prior probabilities corresponding to each scene area type label; the prior probabilities are used to characterize the possibility of scene changes occurring in the areas corresponding to each scene area type label; and use the prior probabilities corresponding to each first sampling point as the first probabilities corresponding to each first sampling point.

[0184] In one embodiment, the flight perspective screening module 805 is further used to obtain the current second candidate perspective; the current second candidate perspective is any one of the second candidate perspectives; according to the second sampling point and the second probability, the second sample change score of each second sampling point corresponding to the current second candidate perspective is obtained; according to the second sample change score of each second sampling point corresponding to the current second candidate perspective, the second importance degree of the current second candidate perspective is obtained.

[0185] In one embodiment, the flight perspective screening module 805 is further used to obtain the current second sampling point and the visited perspectives of the drone during its flight in the target area; the current second sampling point is any one of the second sampling points; according to the visited perspectives and the current second candidate perspectives, the visibility information of the current second sampling point is obtained; the visibility information is used to characterize whether the current second sampling point can be observed by the current second candidate perspective and the visited perspectives; according to the second probability and the visibility information, a second sample change score of the current second sampling point corresponding to the current second candidate perspective is obtained.

[0186] In one embodiment, the second sampling acquisition module 804 is further used to acquire sampling points included in the scene change area, and use the newly added sampling points included in the scene change area and the first sampling points other than the newly added sampling points as second sampling points; set the second probability corresponding to the newly added sampling points to 1, and use the first probability corresponding to the first sampling points other than the newly added sampling points as the second probability corresponding to the first sampling points other than the newly added sampling points.

[0187] In one embodiment, the target perspective screening module 802 is further used to perform Poisson disk sampling on the initial scene reconstruction model to obtain a first candidate perspective set; the flight perspective screening module 805 is further used to eliminate the visited perspectives of the drone during the flight of the target area from the target perspective to obtain an unvisited perspective set; perform Poisson disk sampling on the scene change area to obtain a candidate visual set of the change area; and use the union of the unvisited perspective set and the candidate visual set of the change area as the second candidate perspective set.

[0188] Each module in the above-mentioned drone path online planning device for real-time detection of scene update changes can be implemented in whole or in part by software, hardware, and a combination thereof. Each of the above-mentioned modules can be embedded in or independent of a processor in a computer device in the form of hardware, or can be stored in a memory in a computer device in the form of software, so that the processor can call and execute operations corresponding to each of the above modules.

[0189] In an exemplary embodiment, a computer device is provided. The computer device may be a terminal, and its internal structure diagram may be as shown in FIG. Fig. 9 As shown. The computer device includes a processor, a memory, an input / output interface, a communication interface, a display unit and an input device. The processor, the memory and the input / output interface are connected through a system bus, and the communication interface, the display unit and the input device are connected to the system bus through the input / output interface. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The input / output interface of the computer device is used to exchange information between the processor and the external device. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner can be realized through WIFI, a mobile cellular network, NFC (near field communication) or other technologies. When the computer program is executed by the processor, a method for online planning of a drone path for real-time detection of scene update changes is realized. The display unit of the computer device is used to form a visually visible picture, which can be a display screen, a projection device or a virtual reality imaging device. The display screen can be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device can be a touch layer covering the display screen, or a button, trackball or touchpad set on the computer device shell, or an external keyboard, touchpad or mouse.

[0190] Those skilled in the art will understand that Fig. 9 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0191] In one embodiment, a computer device is further provided, including a memory and a processor, wherein a computer program is stored in the memory, and the processor implements the steps in the above method embodiments when executing the computer program.

[0192] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps in the above-mentioned method embodiments are implemented.

[0193] In one embodiment, a computer program product is provided, including a computer program, which implements the steps in the above method embodiments when executed by a processor.

[0194] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with relevant regulations.

[0195] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetoresistive random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. Non-relational databases may include distributed databases based on blockchains, etc., but are not limited to this. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, etc., but are not limited to this.

[0196] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0197] The above-described embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the present application. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the attached claims.

Claims

1. A method for online planning of drone paths for real-time detection of scene updates and changes, characterized in that: The method comprises: Acquire an initial scene reconstruction model of the target area, and acquire first sampling points included in the initial scene reconstruction model, and first probabilities that each of the first sampling points is located in a scene change area in the target area; A first candidate perspective set associated with the target area is obtained, and a first importance degree of each first candidate perspective included in the first candidate perspective set is obtained according to the first sampling point and the first probability, and a target perspective is obtained from the first candidate perspectives according to the first importance degree; the perspective set composed of the target perspectives is a perspective set with the smallest number of perspectives and the largest coverage range for the target area; Generate a priori flight path of the UAV according to each of the target viewing angles; the priori flight path is used to control the estimated flight of the UAV in the target area; When the UAV is flying in the target area according to the prior flight path, if a scene change area is detected, obtaining second sampling points and a second probability that each of the second sampling points is located in the scene change area; Acquire a second candidate viewing angle set associated with the target area, obtain a second importance degree of each second candidate viewing angle included in the second candidate viewing angle set according to the second sampling point and the second probability, and acquire a target flight viewing angle from the second candidate viewing angles according to the second importance degree; The priori flight path is adjusted according to the target flight angle to obtain a real-time flight path; the real-time flight path is used to control the real-time flight of the UAV in the target area to update the initial scene reconstruction model.

2. The method according to claim 1, characterized in that The obtaining, according to the first sampling point and the first probability, a first importance degree of each first candidate perspective included in the first candidate perspective set includes: Acquire a current first candidate viewing angle and a target first sampling point observable by the current first candidate viewing angle; the current first candidate viewing angle is any one of the first candidate viewing angles; Obtaining a first sample change score corresponding to each of the first sampling points of the target according to the first probability corresponding to each of the first sampling points of the target; Based on each of the first sample change scores, a first importance degree of the current first candidate perspective is obtained.

3. The method according to claim 2, characterized in that The obtaining, according to the first probability corresponding to each of the first sampling points of the target, a first sample change score corresponding to each of the first sampling points of the target includes: Acquire a current first sampling point and the number of first candidate viewing angles from which the current first sampling point can be observed; the current first sampling point is any one of the first sampling points; A first sample change score corresponding to the current first sampling point is obtained according to the quantity and the first probability corresponding to the current first sampling point.

4. The method according to claim 2 or 3, characterized in that: The obtaining of the first sampling points included in the initial scene reconstruction model and the first probability that each of the first sampling points is located in the scene change area in the target area includes: Acquire the first sampling points included in the initial scene reconstruction model, and the scene area type labels corresponding to the first sampling points; Acquire a priori probabilities corresponding to each of the scene area type labels; the priori probabilities are used to characterize the possibility of scene changes occurring in the areas corresponding to each of the scene area type labels; The priori probabilities corresponding to the first sampling points are used as the first probabilities corresponding to the first sampling points.

5. The method according to claim 1, characterized in that The obtaining, according to the second sampling point and the second probability, a second importance degree of each second candidate perspective included in the second candidate perspective set includes: Acquire a current second candidate viewing angle; the current second candidate viewing angle is any one of the second candidate viewing angles; Obtaining, according to the second sampling points and the second probability, a second sample change score of each second sampling point corresponding to the current second candidate perspective; A second importance degree of the current second candidate perspective is obtained according to a second sample change score of each of the second sampling points corresponding to the current second candidate perspective.

6. The method according to claim 5, characterized in that The obtaining, according to the second sampling points and the second probability, a second sample change score of each second sampling point corresponding to the current second candidate perspective includes: Acquire a current second sampling point and a visited viewing angle of the UAV during the flight of the target area; the current second sampling point is any one of the second sampling points; Obtain visibility information of the current second sampling point according to the visited viewing angle and the current second candidate viewing angle; the visibility information is used to indicate whether the current second sampling point can be observed by the current second candidate viewing angle and the visited viewing angle; A second sample change score of the current second sampling point corresponding to the current second candidate viewing angle is obtained according to the second probability and the visibility information.

7. The method according to claim 5 or 6, characterized in that: The acquiring the second sampling points and the second probability that each of the second sampling points is located in the scene change area includes: Acquire sampling points included in the scene change region, and use newly added sampling points included in the scene change region and first sampling points other than the newly added sampling points as the second sampling points; The second probability corresponding to the newly added sampling point is set to 1, and the first probability corresponding to the first sampling point except the newly added sampling point is used as the second probability corresponding to the first sampling point except the newly added sampling point.

8. The method according to claim 1, characterized in that The obtaining of a first candidate viewing angle set associated with the target area includes: Performing Poisson disk sampling on the initial scene reconstruction model to obtain the first candidate perspective set; The step of obtaining a second candidate viewing angle set associated with the target area includes: Eliminating the visited viewpoints of the UAV during the flight of the target area from the target viewpoints to obtain a set of unvisited viewpoints; Performing Poisson disk sampling on the scene change area to obtain a candidate visual set of the change area; The union of the unvisited viewing angle set and the changed area candidate viewing angle set is used as the second candidate viewing angle set.

9. An online drone path planning device for real-time detection of scene update changes, characterized in that: The device comprises: A first sampling acquisition module, used to acquire an initial scene reconstruction model of a target area, and acquire first sampling points included in the initial scene reconstruction model, and a first probability that each of the first sampling points is located in a scene change area in the target area; a target perspective screening module, configured to obtain a first candidate perspective set associated with the target area, obtain a first importance degree of each first candidate perspective included in the first candidate perspective set according to the first sampling point and the first probability, and obtain a target perspective from the first candidate perspectives according to the first importance degree; the perspective set composed of the target perspectives is a perspective set with the smallest number of perspectives and the largest coverage range for the target area; A priori path generation module, used to generate a priori flight path of the UAV according to each of the target viewing angles; the priori flight path is used to control the estimated flight of the UAV in the target area; A second sampling acquisition module is used to acquire second sampling points and a second probability that each of the second sampling points is located in the scene change area if a scene change area is detected during the UAV flying in the target area according to the prior flight path; a flight perspective screening module, configured to obtain a second candidate perspective set associated with the target area, obtain a second importance degree of each second candidate perspective included in the second candidate perspective set according to the second sampling point and the second probability, and obtain a target flight perspective from the second candidate perspectives according to the second importance degree; A real-time path generation module is used to adjust the prior flight path according to the target flight angle to obtain a real-time flight path; the real-time flight path is used to control the real-time flight of the UAV in the target area to update the initial scene reconstruction model.

10. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 8 are implemented.

Images