Image-based UAV Swarm Target Localization Method, Device, Equipment and Medium

Abstract

The present application relates to an image-based method, device, equipment and medium for target positioning of an unmanned aerial vehicle (UAV) cluster. The method includes: detecting a target in an image captured by a UAV to obtain the position and size of the target object in the image, calculating based on the size ratio of the target object captured by different UAVs in the UAV cluster and the visual depth between the UAVs to determine the initial positioning of the target object in the working range of the UAVs, and finally performing clustering analysis on the calculated initial positioning of the target object to obtain the final positioning of the target object. By using this method, target positioning can be performed based on multiple images captured by UAVs without a reference object, reducing the limitation of the flight environment on target positioning and reducing the load carried by the UAVs. Moreover, this method performs joint positioning of the target based on the UAV cluster, effectively improving the accuracy and reliability of target positioning.

Description

Technical Field

[0001] The present application relates to the technical field of UAV swarm positioning, and particularly to an image-based UAV swarm target positioning method, device, equipment and medium. Background Art

[0002] At present, with the continuous development and popularization of UAV technology, the target positioning technology of UAVs has gradually become a popular technology in various fields. However, the existing UAV target positioning technology mainly obtains the information of the target through sensors carried by UAVs (such as infrared, lidar, etc.), and performs target positioning through its own positioning and navigation technology. However, in the above technical solutions, due to the influence of various factors such as environmental light, temperature and air pressure on the accuracy of the sensors used by UAVs, there is a problem of poor positioning accuracy. Moreover, due to the complex flight scenarios of UAVs and the changing positioning targets, in some flight scenarios with simple backgrounds and no reference objects, it is difficult for UAVs to obtain the information of the target through payload sensors, and even to obtain the precise positioning of the target object. Therefore, how to achieve the precise positioning of the target object in a flight scenario with a simple background is an urgent problem to be solved in this field. Summary of the Invention

[0003] Based on this, in view of the above technical problems, it is necessary to provide an image-based UAV swarm target positioning method, device, equipment and medium that can achieve the precise positioning of the target object in a flight scenario with a simple background.

[0004] An image-based UAV swarm target positioning method, the method includes:

[0005] Obtain the first images captured by each UAV in the UAV network and perform target detection to obtain the position and size of the target object in the first images;

[0006] By comparing the positions of the target object in different images, determine the UAV in the captured first images where the target object is at or closest to the center position of the image as the first UAV, and select a set number of UAVs close to the first UAV on the shooting direction vector of the first UAV to construct a UAV swarm;

[0007] Obtain the second images re-captured by each UAV in the UAV swarm and perform target detection to obtain the position and size of the target object in the second images;

[0008] Construct multiple groups of image pairs according to the first image captured by the first UAV and the second images captured by each UAV in the UAV swarm, and calculate according to the size ratio of the target object in each group of image pairs and the positions of the UAVs to obtain the initial positioning of the target object determined by each group of image pairs;

[0009] Obtain the initial localizations of the target objects determined by all image pairs and perform clustering analysis. Eliminate the outlier initial localizations of the target objects to obtain the initial localization clustering result, and take the average of the initial localization clustering result to obtain the final localization of the target objects.

[0010] In one embodiment, the YOLO object detection algorithm is used to perform object detection on the first image and the second image.

[0011] In one embodiment, after constructing the UAV cluster, it further includes: adjusting the shooting direction vectors of all UAVs in the UAV cluster to be consistent with the shooting direction vector of the first UAV.

[0012] In one embodiment, according to the size ratio of the target object in each group of image pairs and the positions of the UAVs, calculate the initial localizations of the target objects determined by each group of image pairs, including:

[0013] Calculate the distances between different UAVs and the target object according to the sizes of the target object in different images in each group of image pairs and the actual size of the target object;

[0014] Calculate according to the distances between different UAVs and the target object to determine the size ratio of the target object, and calculate according to the size ratio of the target object and the positions of the UAVs to obtain the initial localizations of the target objects determined by each group of image pairs.

[0015] In one embodiment, calculate the distances between different UAVs and the target object according to the sizes of the target object in different images in each group of image pairs and the actual size of the target object, including:

[0016] For the image pair constructed by the first image taken by the first UAV a and the second image taken by any UAV b in the UAV cluster calculate the distances between the two UAVs and the target object on the shooting direction vector of the first UAV a according to the sizes of the target object in different images and the actual size of the target object. The specific calculation expression is:

[0017]

[0018] Among them, represents the size of the target object in the first image taken by the first UAV a, represents the size of the target object in the second image taken by the UAV b, s r represents the actual size of the target object, γ1 and γ2 respectively represent the imaging parameters of the first UAV a and the UAV b, and d1 and d2 respectively represent on the shooting direction vector of the first UAV a Above, the distances between the first drone a and the drone b and the target object and respectively represent the horizontal axis coordinate, vertical axis coordinate, and vertical axis coordinate of the shooting direction vector of the first drone a in the working coordinate system of the drone.

[0019] In one embodiment, calculations are performed based on the distances between different drones and the target object to determine the size ratio of the target object. Calculations are then performed based on the size ratio of the target object and the positions of the drones to obtain the initial positioning of the target object determined for each pair of images, including:

[0020] Calculations are performed based on the distances between the first drone a and the drone b and the target object to determine that the size ratio of the target object is

[0021] Calculations are performed based on the size ratio t of the target object, the position of the first drone a, and the position of the drone b to obtain the initial positioning of the target object as (x o , y o , z o ), where x o , y o and z o respectively represent

[0022]

[0023]

[0024]

[0025] Among them, x o , y o and z o respectively represent the horizontal axis coordinate, vertical axis coordinate, and vertical axis coordinate of the initial positioning of the target object in the working coordinate system of the drone, X a , y a and z a respectively represent the horizontal axis coordinate, vertical axis coordinate, and vertical axis coordinate of the first drone a in the working coordinate system of the drone, x b , y b and z b respectively represent the horizontal axis coordinate, vertical axis coordinate, and vertical axis coordinate of the drone b in the working coordinate system of the drone.

[0026] An image-based drone swarm target positioning device, the device includes:

[0027] A first image target detection module, configured to obtain the first images captured by each drone in the drone network and perform target detection to obtain the position and size of the target object in the first images;

[0028] The UAV cluster construction module is used to determine the UAV in the first captured image where the target object is at or closest to the center position of the image by comparing the positions of the target object in different images, and select a set number of UAVs close to the first UAV on the shooting direction vector of the first UAV to construct a UAV cluster;

[0029] The second image target detection module is used to obtain the second images recaptured by each UAV in the UAV cluster and perform target detection to obtain the position and size of the target object in the second image;

[0030] The initial positioning module is used to construct multiple groups of image pairs based on the first image captured by the first UAV and the second images captured by each UAV in the UAV cluster, and calculate based on the size ratio of the target object and the positions of the UAVs in each group of image pairs to obtain the initial positioning of the target object determined by each group of image pairs;

[0031] The clustering analysis module is used to obtain the initial positioning of the target object determined by all image pairs and perform clustering analysis, remove the outlier initial positioning of the target object, obtain the clustering result of the initial positioning, and take the average of the clustering result of the initial positioning to obtain the final positioning of the target object.

[0032] A computer device includes a memory and a processor. The memory stores a computer program, and when the processor executes the computer program, the following steps are implemented:

[0033] Obtain the first images captured by each UAV in the UAV network and perform target detection to obtain the position and size of the target object in the first image;

[0034] Determine the UAV in the first captured image where the target object is at or closest to the center position of the image by comparing the positions of the target object in different images, and select a set number of UAVs close to the first UAV on the shooting direction vector of the first UAV to construct a UAV cluster;

[0035] Obtain the second images recaptured by each UAV in the UAV cluster and perform target detection to obtain the position and size of the target object in the second image;

[0036] Construct multiple groups of image pairs based on the first image captured by the first UAV and the second images captured by each UAV in the UAV cluster, and calculate based on the size ratio of the target object and the positions of the UAVs in each group of image pairs to obtain the initial positioning of the target object determined by each group of image pairs;

[0037] Obtain the initial positioning of the target object determined by all image pairs and perform clustering analysis. Eliminate the initial positioning of the outlier target objects to obtain the initial positioning clustering result, and take the average of the initial positioning clustering result to obtain the final positioning of the target object.

[0038] A computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the following steps are implemented:

[0039] Obtain the first images captured by each unmanned aerial vehicle (UAV) in the UAV network and perform target detection to obtain the position and size of the target object in the first images.

[0040] By comparing the positions of the target object in different images, determine the UAV that captures the first image in which the target object is at or closest to the center position of the image as the first UAV, and select a set number of UAVs close to the first UAV on the shooting direction vector of the first UAV to construct a UAV cluster.

[0041] Obtain the second images recaptured by each UAV in the UAV cluster and perform target detection to obtain the position and size of the target object in the second images.

[0042] Construct multiple groups of image pairs based on the first image captured by the first UAV and the second images captured by each UAV in the UAV cluster. Calculate according to the size ratio of the target object in each group of image pairs and the positions of the UAVs to obtain the initial positioning of the target object determined by each group of image pairs.

[0043] Obtain the initial positioning of the target object determined by all image pairs and perform clustering analysis. Eliminate the initial positioning of the outlier target objects to obtain the initial positioning clustering result, and take the average of the initial positioning clustering result to obtain the final positioning of the target object.

[0044] The above image-based UAV cluster target positioning method, device, equipment and medium obtain the position and size of the target object in the image by performing target detection on the images captured by the UAVs, and calculate according to the size ratio of the target object captured by different UAVs in the UAV cluster and the visual depth between the UAVs to determine the initial positioning of the target object in the UAV working area. Finally, by performing clustering analysis on the calculated initial positioning of the target object, the final positioning of the target object is obtained. Compared with the prior art, the present application can perform target positioning according to multiple images captured by the UAVs without a reference object, reduce the limitation of the flight environment on UAV target positioning, and reduce the UAV load. In addition, the present application reduces the influence of the single UAV shooting error on the target positioning accuracy by clustering and analyzing the initial positioning of the target object, and the joint positioning of multiple UAVs on the target effectively improves the accuracy and reliability of the target positioning. Description of the Drawings

[0045] Figure 1 It is a schematic flowchart of an image-based UAV swarm target localization method in an embodiment;

[0046] Figure 2 It is a schematic diagram for calculating the initial localization of a target object in an embodiment;

[0047] Figure 3 It is an internal structure diagram of a computer device in an embodiment. Detailed implementation manners

[0048] In order to make the objectives, technical solutions and advantages of the present application clearer and more understandable, the present application will be further described in detail below with reference to 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.

[0049] In one embodiment, as Figure 1 shown, a method for image-based UAV swarm target localization is provided, including the following steps:

[0050] Step S1: Obtain the first images captured by each UAV in the UAV network and perform target detection to obtain the position and size of the target object in the first images.

[0051] Step S2: By comparing the positions of the target object in different images, determine the UAV with the target object in the first image being at or closest to the center position of the image as the first UAV, and select a set number of UAVs close to the first UAV on the shooting direction vector of the first UAV to form a UAV swarm.

[0052] Step S3: Obtain the second images recaptured by each UAV in the UAV swarm and perform target detection to obtain the position and size of the target object in the second images.

[0053] Step S4: Construct multiple groups of image pairs based on the first image captured by the first UAV and the second images captured by each UAV in the UAV swarm, and calculate according to the size ratio of the target object and the positions of the UAVs in each group of image pairs to obtain the initial localization of the target object determined by each group of image pairs.

[0054] Step S5: Obtain the initial localizations of the target object determined by all image pairs and perform clustering analysis, remove the outlier initial localizations of the target object to obtain the clustering result of the initial localizations, and take the average of the clustering result of the initial localizations to obtain the final localization of the target object.

[0055] It can be understood that, by initially locating the target object determined for all image pairs and performing clustering analysis, the present application removes the outliers caused by the shooting errors of individual drones, effectively improving the positioning accuracy of the target object. Specifically, the clustering analysis methods adopted in the present application include BIRCH (Balanced Iterative Reducing and Clustering Using Hierarchies) and K-means (K-means clustering).

[0056] In one embodiment, the YOLO object detection algorithm is used to perform object detection on the first image and the second image.

[0057] In one embodiment, after constructing the drone cluster, it further includes: adjusting the shooting direction vectors of all drones in the drone cluster to be the same as the shooting direction vector of the first drone.

[0058] In one embodiment, according to the size ratio of the target object in each group of image pairs and the positions of the drones, the initial positioning of the target object determined for each group of image pairs is calculated, including:

[0059] According to the sizes of the target object in different images in each group of image pairs and the actual size of the target object, the distances between different drones and the target object are calculated;

[0060] According to the distances between different drones and the target object, the size ratio of the target object is determined, and according to the size ratio of the target object and the positions of the drones, the initial positioning of the target object determined for each group of image pairs is calculated.

[0061] Specifically, as Figure 2 shown, for the image pair constructed from the first image taken by the first drone a and the second image taken by any drone b in the drone cluster First, according to the sizes of the target object in different images and the actual size of the target object, the distances between the two drones and the target object on the shooting direction vector of the first drone a are calculated. The specific calculation expression is:

[0062]

[0063] where represents the size of the target object in the first image taken by the first drone a, represents the size of the target object in the second image taken by the drone b, s rrepresents the actual size of the target object, γ1 and γ2 respectively represent the imaging parameters of the first drone a and the drone b, and d1 and d2 respectively represent the distances between the first drone a and the drone b and the target object in the shooting direction vector of the first drone a on which, and respectively represent the horizontal axis coordinate, vertical axis coordinate, and vertical axis coordinate of the shooting direction vector of the first drone a in the working coordinate system of the drone

[0064] Then, calculate according to the distances between the first drone a and the drone b and the target object to determine that the size ratio of the target object is and according to Figure 2 it can be known that d2 = d1 - ab’, where ab’ represents the distance between the first drone a and the drone b in the shooting direction vector of the first drone a on which, Figure 2 b’ in represents the projection of the drone b on the shooting direction vector of the first drone a

[0065]

[0066] According to ab’, it can be deduced and calculated to obtain

[0067] d1 * t = d1 - ab’

[0068]

[0069] Finally, calculate according to the size ratio t of the target object, the position of the first drone a, and the position of the drone b to obtain the initial positioning of the target object as (x o , y o , z o ), where x o , y o and z o respectively represent as

[0070]

[0071]

[0072]

[0073] where x o , y o and z o respectively represent the horizontal axis coordinate, vertical axis coordinate, and vertical axis coordinate of the initial positioning of the target object in the working coordinate system of the drone, x a , y a and z arespectively represent the horizontal axis coordinate, vertical axis coordinate, and vertical axis coordinate of the first drone a in the drone working coordinate system, x b , y b and z b respectively represent the horizontal axis coordinate, vertical axis coordinate, and vertical axis coordinate of the drone b in the drone working coordinate system.

[0074] It should be understood that although Figure 1 the steps in the flowchart of Figure 1 are shown in sequence according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise clearly stated in this article, there is no strict order restriction for the execution of these steps, and these steps can be executed in other orders. Moreover,

[0075] In one embodiment, an image-based drone swarm target positioning device is provided, including: a first image target detection module, a drone swarm construction module, a second image target detection module, an initial positioning module, and a clustering analysis module, where:

[0076] The first image target detection module is used to obtain the first images captured by each drone in the drone network and perform target detection to obtain the position and size of the target object in the first images;

[0077] The drone swarm construction module is used to determine, by comparing the positions of the target objects in different images, that the drone with the target object in the first image being at or closest to the center position of the image is the first drone, and select a set number of drones close to the first drone on the shooting direction vector of the first drone to construct a drone swarm;

[0078] The second image target detection module is used to obtain the second images re-captured by each drone in the drone swarm and perform target detection to obtain the position and size of the target object in the second images;

[0079] The initial positioning module is used to construct multiple groups of image pairs based on the first image captured by the first drone and the second images captured by each drone in the drone swarm, and calculate based on the size ratio of the target object in each group of image pairs and the positions of the drones to obtain the initial positioning of the target object determined by each group of image pairs;

[0080] The clustering analysis module is used to obtain the initial positioning of the target objects determined by all image pairs, perform clustering analysis, eliminate the initial positioning of the outlier target objects, obtain the clustering result of the initial positioning, and take the average of the clustering result of the initial positioning to obtain the final positioning of the target objects.

[0081] For the specific limitations of the image-based UAV swarm target positioning device, reference can be made to the limitations of the image-based UAV swarm target positioning method in the above text, which will not be elaborated here. Each module in the above image-based UAV swarm target positioning device can be implemented in whole or in part by software, hardware, and their combinations. The above-mentioned modules can be embedded in the processor of the computer device in hardware form or independent of it, or stored in the memory of the computer device in software form, so that the processor can call and execute the operations corresponding to the above-mentioned modules.

[0082] In one embodiment, a computer device is provided. The computer device can be a terminal, and its internal structure diagram can be as Figure 3 shown. The computer device includes a processor, a memory, a network interface, a display screen, and an input device connected through a system bus. Among them, 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 network interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, it realizes an image-based UAV swarm target positioning method. The display screen of the computer device can be a liquid crystal display screen or an electronic ink display screen. The input device of the computer device can be a touch layer covering the display screen, or a button, a trackball, or a touchpad provided on the shell of the computer device, or an external keyboard, a touchpad, or a mouse, etc.

[0083] Those skilled in the art can understand that Figure 3 the structure shown in

[0084] is only a block diagram of some structures related to the solution of this application, and does not constitute a limitation on the computer device to which the solution of this application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine some components, or have different component arrangements.

[0085] Obtain the first images captured by each UAV in the UAV network and perform target detection to obtain the position and size of the target object in the first image;

[0086] By comparing the positions of the target object in different images, determine the drone in the first captured image where the target object is at or closest to the center position of the image as the first drone, and select a set number of drones close to the first drone on the shooting direction vector of the first drone to form a drone cluster;

[0087] Obtain the second images recaptured by each drone in the drone cluster and perform target detection to obtain the position and size of the target object in the second images;

[0088] Construct multiple groups of image pairs based on the first image captured by the first drone and the second images captured by each drone in the drone cluster, and calculate according to the size ratio of the target object and the positions of the drones in each group of image pairs to obtain the initial positioning of the target object determined by each group of image pairs;

[0089] Obtain the initial positionings of the target object determined by all image pairs and perform clustering analysis, eliminate the outlier initial positionings of the target object, obtain the clustering result of the initial positionings, and take the average of the clustering result of the initial positionings to obtain the final positioning of the target object.

[0090] 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 following steps are implemented:

[0091] Obtain the first images captured by each drone in the drone network and perform target detection to obtain the position and size of the target object in the first images;

[0092] By comparing the positions of the target object in different images, determine the drone in the first captured image where the target object is at or closest to the center position of the image as the first drone, and select a set number of drones close to the first drone on the shooting direction vector of the first drone to form a drone cluster;

[0093] Obtain the second images recaptured by each drone in the drone cluster and perform target detection to obtain the position and size of the target object in the second images;

[0094] Construct multiple groups of image pairs based on the first image captured by the first drone and the second images captured by each drone in the drone cluster, and calculate according to the size ratio of the target object and the positions of the drones in each group of image pairs to obtain the initial positioning of the target object determined by each group of image pairs;

[0095] Obtain the initial positionings of the target object determined by all image pairs and perform clustering analysis, eliminate the outlier initial positionings of the target object, obtain the clustering result of the initial positionings, and take the average of the clustering result of the initial positionings to obtain the final positioning of the target object.

[0096] Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above embodiments can be completed by instructing relevant hardware through a computer program. 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 methods. Among them, any reference to a memory, storage, database, or other medium used in the embodiments provided in the present application can include non-volatile and / or volatile memories. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDR SDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and Rambus dynamic RAM (RDRAM), etc.

[0097] The technical features of the above embodiments can be combined arbitrarily. For the sake of brevity of description, 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, it should be considered as the scope described in this specification.

[0098] The above-described embodiments merely represent several implementation manners of the present application. The description is relatively specific and detailed, but it should not be construed as a limitation on the scope of the present application. It should be noted that for those of ordinary skill in the art, without departing from the concept of the present application, several modifications and improvements can still be made, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the appended claims.

Claims

1. An image-based method for target localization of UAV swarms, characterized in that, The method includes: Obtaining first images captured by each unmanned aerial vehicle (UAV) in the UAV network and performing object detection to obtain the position and size of the target object in the first images; By comparing the positions of the target object in different images, determining the UAV in the captured first image in which the target object is at or closest to the center position of the image as the first UAV, and selecting a set number of UAVs close to the first UAV on the shooting direction vector of the first UAV to construct a UAV cluster; Obtaining second images re-captured by each UAV in the UAV cluster and performing object detection to obtain the position and size of the target object in the second images; Constructing multiple groups of image pairs based on the first image captured by the first UAV and the second images captured by each UAV in the UAV cluster, and calculating according to the size ratio of the target object and the positions of the UAVs in each group of image pairs to obtain the initial positioning of the target object determined by each group of image pairs; Obtaining the initial positioning of the target object determined by all image pairs and performing clustering analysis, removing the outlier initial positioning of the target object to obtain the clustering result of the initial positioning, and averaging the clustering result of the initial positioning to obtain the final positioning of the target object.

2. The method according to claim 1, wherein Using the YOLO object detection algorithm to perform object detection on the first image and the second image.

3. The method according to claim 1, characterized in that, After constructing the UAV cluster, it further includes: adjusting the shooting direction vectors of all UAVs in the UAV cluster to be consistent with the shooting direction vector of the first UAV.

4. The method according to claim 1, wherein Calculating according to the size ratio of the target object and the positions of the UAVs in each group of image pairs to obtain the initial positioning of the target object determined by each group of image pairs, including: Calculating the distances between different UAVs and the target object according to the sizes of the target object in different images and the actual size of the target object in each group of image pairs; Calculating according to the distances between different UAVs and the target object to determine the size ratio of the target object, and calculating according to the size ratio of the target object and the positions of the UAVs to obtain the initial positioning of the target object determined by each group of image pairs.

5. The method according to claim 4, wherein Calculating the distances between different UAVs and the target object according to the sizes of the target object in different images and the actual size of the target object in each group of image pairs, including: For the image pair constructed from the first image captured by the first drone a and the second image captured by any drone b in the drone cluster Based on the sizes of the target object in different images and the actual size of the target object, calculate the distances between the two drones and the target object in the shooting direction vector of the first drone a. The specific calculation formula is as follows: Among them, represents the size of the target object in the first image captured by the first drone a, represents the size of the target object in the second image captured by the drone b, s r represents the actual size of the target object, γ1 and γ2 respectively represent the imaging parameters of the first drone a and the drone b, and d1 and d2 respectively represent the distances between the first drone a and the drone b and the target object in the shooting direction vector of the first drone a, and respectively represent the horizontal axis coordinate, vertical axis coordinate, and vertical axis coordinate of the shooting direction vector of the first drone a in the working coordinate system of the drone.

6. The method according to claim 5, characterized in that, Calculating according to the distances between different UAVs and the target object to determine the size ratio of the target object, and calculating according to the size ratio of the target object and the positions of the UAVs to obtain the initial positioning of the target object determined by each group of image pairs, including: Based on the distances between the first drone a and the drone b and the target object, calculate and determine that the size ratio of the target object is Calculate based on the size ratio t of the target object, the position of the first drone a, and the position of drone b, and obtain the initial positioning of the target object as (x o , y o , z o ), where x o , y o and z o respectively represent Among them, x o , y o and z o respectively represent the horizontal axis coordinate, vertical axis coordinate, and vertical axis coordinate of the target object's initial position in the UAV working coordinate system. x a , y a and z a respectively represent the horizontal axis coordinate, vertical axis coordinate, and vertical axis coordinate of the first UAV a in the UAV working coordinate system. x b , y b and z b respectively represent the horizontal axis coordinate, vertical axis coordinate, and vertical axis coordinate of the UAV b in the UAV working coordinate system.

7. An image-based target positioning device for an unmanned aerial vehicle cluster, characterized in that, The device includes: A first image object detection module, configured to obtain first images captured by each UAV in the UAV network and perform object detection to obtain the position and size of the target object in the first images; A UAV cluster construction module, configured to determine the UAV in the captured first image in which the target object is at or closest to the center position of the image as the first UAV by comparing the positions of the target object in different images, and select a set number of UAVs close to the first UAV on the shooting direction vector of the first UAV to construct a UAV cluster; The second image target detection module is used to obtain the second images re-taken by each drone in the drone cluster and perform target detection to obtain the position and size of the target object in the second images; The initial positioning module is used to construct multiple groups of image pairs based on the first image taken by the first drone and the second images taken by each drone in the drone cluster, calculate according to the size ratio of the target object and the position of the drone in each group of image pairs, and obtain the initial positioning of the target object determined by each group of image pairs; The clustering analysis module is used to obtain the initial positioning of the target object determined by all image pairs and perform clustering analysis, eliminate the outlier initial positioning of the target object, obtain the initial positioning clustering result, and take the average of the initial positioning clustering result to obtain the final positioning of the target object.

8. A computer device, comprising a memory and a processor, the memory storing a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the method according to any one of claims 1 to 6.

Images