Unmanned aerial vehicle indoor target visual positioning method and device

Abstract

The invention discloses an unmanned aerial vehicle indoor target visual positioning method and device. The method comprises the following steps: acquiring a plurality of configuration parameters of an unmanned aerial vehicle executing target equipment positioning, AOA positioning algorithm associated parameters and image data acquired by shooting equipment; the configuration parameters comprise positioning data of the unmanned aerial vehicle and shooting parameters of shooting equipment; the AOA positioning algorithm associated parameters comprise parameters of an antenna arranged at a specified position of the target equipment and the amplitude of a wireless signal transmitted by a built-in wireless signal transmitter of the unmanned aerial vehicle; performing target equipment information extraction on the image data to obtain a color component value and an imaging depth-of-field value of the target equipment in the image data; and analyzing and outputting the positioning information of the target equipment by adopting an intelligent visual positioning model based on the plurality of configuration parameters, the AOA positioning algorithm associated parameters, the color component numerical value and the imaging depth-of-field numerical value. The indoor unmanned aerial vehicle positioning processing precision can be improved.

Description

Technical Field

[0001] The present invention relates to the field of unmanned aerial vehicle positioning and computing technology, and in particular to a method and device for visually positioning an indoor target of an unmanned aerial vehicle. Background Art

[0002] This section is intended to provide a background or context to the embodiments of the invention recited in the claims. No admission is made that the description herein is prior art by inclusion in this section.

[0003] In the prior art, there are technologies that use wireless signal positioning for indoor drone positioning methods, such as the AOA positioning algorithm. When the AOA (Angle of Arrival) positioning algorithm is applied to the positioning of indoor target devices, multiple receiving antennas are deployed on the target device. In practice, the distance between each receiving antenna and the drone is different. However, the implementation of the AOA positioning algorithm in the prior art does not take into account the cumbersomeness and time-consuming nature of a series of complex processing processes, such as the measurement of the phase difference of signals received by multiple receiving antennas, the calculation of the signal arrival angle, and the calculation of the positioning information of the indoor target device. At the same time, the accuracy of the positioning information obtained by the above series of complex processing processes still has room for improvement. Summary of the invention

[0004] The embodiment of the present invention provides a method for visual positioning of indoor targets by unmanned aerial vehicles, which is used to simplify the indoor unmanned aerial vehicle positioning processing process, reduce the indoor unmanned aerial vehicle positioning processing time, and improve the indoor unmanned aerial vehicle positioning processing accuracy. The method includes:

[0005] Acquire multiple configuration parameters of a drone that performs target device positioning, AOA positioning algorithm-related parameters, and image data collected by a shooting device; the drone is equipped with a shooting device, the shooting device is electrically connected to the drone, and a pre-trained intelligent visual positioning model is configured in the drone; the configuration parameters include the positioning data of the drone and the camera parameters of the shooting device; the AOA positioning algorithm-related parameters include the parameters of an antenna arranged at a designated position of the target device and the amplitude of a wireless signal transmitted by a built-in wireless signal transmitter of the drone;

[0006] Extract target device information from the image data to obtain the color component value and imaging depth value of the target device in the image data;

[0007] An intelligent visual positioning model is adopted to parse and output the positioning information of the target device based on the multiple configuration parameters, the AOA positioning algorithm associated parameters, the color component values ​​and the imaging depth of field values; the intelligent visual positioning model uses the historical: known positioning information of the target device, multiple configuration parameters of the drone that performs the target device positioning, the AOA positioning algorithm associated parameters, and the image data collected by the shooting device to train the deep neural network.

[0008] The embodiment of the present invention also provides a UAV indoor target visual positioning device, which is used to simplify the indoor UAV positioning processing process, reduce the indoor UAV positioning processing time, and improve the indoor UAV positioning processing accuracy. The device includes:

[0009] A data acquisition module is used to acquire multiple configuration parameters of a drone that performs target device positioning, AOA positioning algorithm-related parameters, and image data collected by a shooting device; the drone is equipped with a shooting device, the shooting device is electrically connected to the drone, and a pre-trained intelligent visual positioning model is configured in the drone; the configuration parameters include the positioning data of the drone and the camera parameters of the shooting device; the AOA positioning algorithm-related parameters include the parameters of an antenna arranged at a designated position of the target device and the amplitude of a wireless signal transmitted by a built-in wireless signal transmitter of the drone;

[0010] An image processing module is used to extract target device information from image data to obtain color component values ​​and imaging depth values ​​of the target device in the image data;

[0011] The positioning analysis module is used to use an intelligent visual positioning model to parse and output the positioning information of the target device based on the multiple configuration parameters, the AOA positioning algorithm associated parameters, the color component values ​​and the imaging depth of field values; the intelligent visual positioning model uses the historical: known positioning information of the target device, multiple configuration parameters of the drone that performs the target device positioning, the AOA positioning algorithm associated parameters, and the image data collected by the shooting device to train the deep neural network.

[0012] An embodiment of the present invention also provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the above-mentioned method for visual positioning of indoor targets of unmanned aerial vehicles when executing the computer program.

[0013] An embodiment of the present invention further provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the above-mentioned method for visual positioning of indoor targets of unmanned aerial vehicles is implemented.

[0014] An embodiment of the present invention further provides a computer program product, which includes a computer program. When the computer program is executed by a processor, the above-mentioned drone indoor target visual positioning method is implemented.

[0015] In an embodiment of the present invention, multiple configuration parameters of a drone that performs target device positioning, AOA positioning algorithm associated parameters, and image data collected by a shooting device are obtained; the drone is equipped with a shooting device, the shooting device is electrically connected to the drone, and a pre-trained intelligent visual positioning model is configured in the drone; the configuration parameters include the positioning data of the drone and the camera parameters of the shooting device; the AOA positioning algorithm associated parameters include the parameters of the antenna deployed at the specified position of the target device and the amplitude of the wireless signal emitted by the built-in wireless signal transmitter of the drone; the target device information is extracted from the image data to obtain the color component value and the imaging depth of field value of the target device in the image data; the intelligent visual positioning model is used to parse and output the positioning information of the target device based on the multiple configuration parameters, the AOA positioning algorithm associated parameters, the color component value and the imaging depth of field value; the intelligent visual positioning model uses the historical: known positioning information of the target device, multiple configuration parameters of the drone that performs target device positioning, AOA positioning algorithm associated parameters, and image data collected by the shooting device to train a deep neural network. Compared with the technical solution of applying the AOA positioning algorithm in the prior art, the embodiment of the present invention combines the visual image data, the parameters of the antenna deployed at the specified position of the target device, and the AOA positioning algorithm associated parameters of the amplitude of the wireless signal transmitted by the built-in wireless signal transmitter of the drone, and adopts a deep neural network for positioning analysis, thereby avoiding the complex processing process of the existing AOA positioning algorithm, simplifying the indoor drone positioning processing process, reducing the indoor drone positioning processing time, and improving the high-precision positioning of indoor target devices. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the prior art descriptions. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work. In the drawings:

[0017] Figure 1 Schematic diagram of the process of the indoor target visual positioning method of a UAV in an embodiment of the present invention;

[0018] Figure 2 This is a schematic diagram of the structure of a drone and a photographing device in an embodiment of the present invention;

[0019] Figure 3 A specific example diagram of the indoor target visual positioning method of a UAV in an embodiment of the present invention;

[0020] Figure 4 Schematic diagram of a visual positioning device for indoor targets of a UAV in an embodiment of the present invention;

[0021] Figure 5 Schematic diagram of a computer device in an embodiment of the present invention. DETAILED DESCRIPTION

[0022] To make the purpose, technical solution and advantages of the embodiments of the present invention more clear, the embodiments of the present invention are further described in detail below in conjunction with the accompanying drawings. Here, the exemplary embodiments of the present invention and their descriptions are used to explain the present invention, but are not intended to limit the present invention.

[0023] Indoor buildings usually block GPS signals. Therefore, indoor drone positioning methods generally use other alternative technologies, such as visual positioning, which uses cameras to capture target devices for positioning, but this method has poor positioning accuracy. Another example is positioning through lidar, but this method is expensive, or combined with inertial measurement units, but long-term use will accumulate errors.

[0024] There are also technologies that use wireless signal positioning in the prior art, such as the AOA positioning algorithm. When the AOA positioning algorithm is applied to the positioning of indoor target devices, the following working mechanism is adopted:

[0025] There is a single wireless transmitter on a drone flying in an indoor environment;

[0026] The indoor target device has multiple receiving antennas, which form an antenna array;

[0027] At this time, a single wireless transmitter on the drone sends a signal to a target device in the room, and the signal is received by multiple receiving antennas on the target device respectively;

[0028] Although the distance from the drone to the target device is fixed, the distances from each receiving antenna to the drone are different. These differences result in different signal path lengths for the sent signal to reach each antenna, which will produce phase differences. By measuring these phase differences, the angle of signal arrival can be calculated, that is, the angle formed by each two different receiving antennas relative to the two directions of the target device, that is, the angle formed between the signal paths of each two receiving antennas. Based on this arrival angle and the positioning information of the drone, a triangulation algorithm is used (the three angles are the target device, two adjacent receiving antennas, the known data are the distance between the two adjacent receiving antennas, the angle at the vertex where the target device is located, i.e. the arrival angle, and the remaining two angles of the triangle. By calculating the length of the signal path of the two receiving antennas, the indoor positioning of the target device can be completed).

[0029] It can be seen that the implementation process of the AOA positioning algorithm in the prior art is relatively complicated, the processing time is long, and the positioning accuracy still has room for improvement. To this end, an embodiment of the present invention proposes a method for visual positioning of indoor targets of unmanned aerial vehicles.

[0030] Figure 1 FIG. 1 is a flow chart of a method for visually locating an indoor target of a UAV in an embodiment of the present invention. Figure 1 As shown, the method includes:

[0031] Step 101, obtaining multiple configuration parameters of a drone that performs target device positioning, AOA positioning algorithm associated parameters, and image data collected by a shooting device; the drone is equipped with a shooting device, the shooting device is electrically connected to the drone, and a pre-trained intelligent visual positioning model is configured in the drone; the configuration parameters include the positioning data of the drone and the camera parameters of the shooting device; the AOA positioning algorithm associated parameters include the parameters of the antenna deployed at the designated position of the target device and the amplitude of the wireless signal transmitted by the built-in wireless signal transmitter of the drone;

[0032] Step 102: extract target device information from the image data to obtain color component values ​​and imaging depth of field values ​​of the target device in the image data;

[0033] Step 103: using an intelligent visual positioning model to parse and output the positioning information of the target device based on the multiple configuration parameters, the AOA positioning algorithm associated parameters, the color component values, and the imaging depth of field values; the intelligent visual positioning model uses the historical: known positioning information of the target device, multiple configuration parameters of the drone that performs the positioning of the target device, the AOA positioning algorithm associated parameters, and the image data collected by the shooting device to train the deep neural network.

[0034] from Figure 1It can be seen from the process shown that in the embodiment of the present invention, multiple configuration parameters of the drone that performs target device positioning, AOA positioning algorithm associated parameters, and image data collected by the shooting device are obtained; the drone is equipped with a shooting device, the shooting device is electrically connected to the drone, and a pre-trained intelligent visual positioning model is configured in the drone; the configuration parameters include the positioning data of the drone and the camera parameters of the shooting device; the AOA positioning algorithm associated parameters include the parameters of the antenna arranged at the specified position of the target device and the amplitude of the wireless signal emitted by the built-in wireless signal transmitter of the drone; the target device information is extracted from the image data to obtain the color component value and the imaging depth of field value of the target device in the image data; the intelligent visual positioning model is used to parse and output the positioning information of the target device based on the multiple configuration parameters, the AOA positioning algorithm associated parameters, the color component value and the imaging depth of field value; the intelligent visual positioning model uses the historical: known positioning information of the target device, multiple configuration parameters of the drone that performs target device positioning, AOA positioning algorithm associated parameters, and image data collected by the shooting device to train the deep neural network. Compared with the technical solution of applying the AOA positioning algorithm in the prior art, the embodiment of the present invention combines the visual image data, the parameters of the antenna deployed at the specified position of the target device, and the AOA positioning algorithm associated parameters of the amplitude of the wireless signal transmitted by the built-in wireless signal transmitter of the drone, and adopts a deep neural network for positioning analysis, thereby avoiding the complex processing process of the existing AOA positioning algorithm, simplifying the indoor drone positioning processing process, reducing the indoor drone positioning processing time, and improving the high-precision positioning of indoor target devices.

[0035] In specific implementation, when performing indoor target device positioning by drone, the drone acts as a wireless signal transmitter, the target device acts as a wireless signal receiver, and the drone and the target device are in the same indoor environment.

[0036] In order to realize the combination of AOA algorithm and visual analysis algorithm, targeted screening is carried out on multiple associated parameters of AOA positioning algorithm and various visual data of indoor shooting pictures of UAV. Specifically, the associated parameters of AOA positioning algorithm include the parameters of antenna arranged at the specified position of target device and the amplitude of wireless signal transmitted by built-in wireless signal transmitter of UAV. Multiple configuration parameters of UAV include positioning data of UAV and camera parameters of shooting equipment.

[0037] In one embodiment, the camera parameters of the camera device include at least one or any combination of the following:

[0038] Shooting distance, horizontal resolution, vertical resolution, dynamic range width;

[0039] Among them, the dynamic range width is the difference between the maximum grayscale value of the pixel that can be captured by the shooting device and the minimum grayscale value of the pixel that can be captured by the shooting device. The horizontal resolution and the vertical resolution can be the ability of the camera to distinguish details when capturing an image, and the shooting distance is the imaging distance.

[0040] In one embodiment, the parameters of the antenna deployed at the designated position of the target device include: the number of antennas in the antenna array and the distance between adjacent antennas.

[0041] Furthermore, the multiple antennas in the antenna array have the same size and the antenna array is arranged as a rectangular array.

[0042] For example, if the target device is a product of the same series in a shopping mall or a mechanical device in a factory, antennas are deployed at designated locations of the target device, such as the middle area. The antennas have the same size and the array layout is a rectangular array.

[0043] Figure 2 is a schematic diagram of the structure of a drone and a photographing device in an embodiment of the present invention, Figure 2 In the present invention, the drone is equipped with shooting equipment, and a customized structure can be used to carry the shooting equipment, so as to provide a carrying platform for different drones and cameras, effectively ensuring the validity and stability of the positioning results of the intelligent visual positioning model.

[0044] The picture taken by the shooting device should include the target device. In one embodiment, the imaging area of ​​the target device in the picture taken indoors by the drone only includes the target device itself.

[0045] After the image data is obtained, the target device information is extracted from the image data to obtain the color component value and imaging depth value of the target device in the image data.

[0046] In one embodiment, extracting target device information from image data to obtain color component values ​​and imaging depth values ​​of the target device in the image data may include:

[0047] The target device information is extracted from the image data to obtain the color component value and imaging depth value of each pixel point in the imaging area of ​​the target device.

[0048] Finally, the data obtained previously is used to analyze and output the positioning information of the target device using the intelligent visual positioning model.

[0049] Figure 3 FIG. 1 is a specific example of a method for visually locating an indoor target of a UAV according to an embodiment of the present invention. Figure 3 As shown, the intelligent visual positioning model is trained as follows:

[0050] Step 301: construct a sample by using historical: known positioning information of a target device, multiple configuration parameters of a drone that performs positioning of the target device, AOA positioning algorithm associated parameters, and image data collected by a shooting device;

[0051] Step 302: Allocate the samples into a training set, a test set, and a validation set according to a specified ratio;

[0052] Step 303: train the deep neural network using the training set, test the deep neural network using the test set, and verify the deep neural network using the verification set, and obtain the trained deep neural network as the output of the intelligent visual positioning model.

[0053] During implementation, in each training process performed on the deep neural network, the positioning information of a certain indoor device that has been measured is used as the single output content of the deep neural network, and multiple configuration data of the drone, the number of antennas and the distance between adjacent antennas of the antenna array of a certain indoor device, the amplitude of the wireless signal emitted by the built-in wireless signal transmitter of the drone, and the color component value and imaging depth of field value of each pixel point in the imaging area of ​​a certain indoor device are set as multiple input contents of the deep neural network to complete this training process, thereby ensuring the training effect of each training process performed on the deep neural network.

[0054] In one embodiment, the number of times the deep neural network is trained is positively correlated with the total number of pixels in the image captured by the drone indoors. That is, the greater the total number of pixels in the entire image, the more times the deep neural network is trained.

[0055] In one embodiment, the data input into the intelligent visual positioning model, that is, the data acquired in advance, may also include a plurality of visualization parameters that are specifically screened and related to the target device, thereby achieving an organic fusion of the AOA positioning algorithm and the visual analysis algorithm.

[0056] In summary, the embodiments of the present invention provide a method for visual positioning of indoor targets of unmanned aerial vehicles, which can combine the AOA algorithm with the visual analysis algorithm to complete the high-precision positioning of indoor target devices. Specifically, by building an intelligent visual positioning model with a customized structure, the positioning information of the indoor target device is intelligently analyzed based on multiple associated parameters of the AOA algorithm and various visual data of the indoor shooting pictures of the unmanned aerial vehicle. In this way, the positioning accuracy of the indoor target device is improved while skipping the complex processing process of the standard AOA algorithm, from multiple antennas receiving signals separately to measuring phase differences, to calculating the angle at which the signals reach, and finally calculating the positioning information of the indoor target device.

[0057] The embodiment of the present invention also provides a UAV indoor target visual positioning device, as described in the following embodiment. Since the principle of solving the problem by the device is similar to that of the UAV indoor target visual positioning method, the implementation of the device can refer to the implementation of the UAV indoor target visual positioning method, and the repeated parts will not be repeated.

[0058] Figure 4 FIG. 1 is a schematic diagram of a UAV indoor target visual positioning device according to an embodiment of the present invention. Figure 4 As shown, the device comprises:

[0059] The data acquisition module 401 is used to acquire multiple configuration parameters of a drone that performs target device positioning, AOA positioning algorithm associated parameters, and image data collected by a shooting device; the drone is equipped with a shooting device, the shooting device is electrically connected to the drone, and a pre-trained intelligent visual positioning model is configured in the drone; the configuration parameters include the positioning data of the drone and the camera parameters of the shooting device; the AOA positioning algorithm associated parameters include the parameters of the antenna deployed at the designated position of the target device and the amplitude of the wireless signal transmitted by the built-in wireless signal transmitter of the drone;

[0060] The image processing module 402 is used to extract the target device information from the image data to obtain the color component value and imaging depth value of the target device in the image data;

[0061] The positioning analysis module 403 is used to use an intelligent visual positioning model to parse and output the positioning information of the target device based on the multiple configuration parameters, the AOA positioning algorithm associated parameters, the color component values ​​and the imaging depth of field values; the intelligent visual positioning model uses the historical: known positioning information of the target device, multiple configuration parameters of the drone that performs the target device positioning, the AOA positioning algorithm associated parameters, and the image data collected by the shooting device to train the deep neural network.

[0062] In one embodiment, the camera parameters of the camera include one or any combination of the following:

[0063] Shooting distance, horizontal resolution, vertical resolution, dynamic range width;

[0064] The dynamic range width is the difference between the maximum grayscale value of a pixel that can be captured by the shooting device and the minimum grayscale value of a pixel that can be captured by the shooting device.

[0065] In one embodiment, the parameters of the antenna deployed at the designated position of the target device include: the number of antennas in the antenna array and the distance between adjacent antennas.

[0066] In one embodiment, the multiple antennas in the antenna array have the same size and the antenna array is arranged as a rectangular array.

[0067] In one embodiment, the image processing module 402 is specifically used for:

[0068] The target device information is extracted from the image data to obtain the color component value and imaging depth value of each pixel point in the imaging area of ​​the target device.

[0069] In one embodiment, the intelligent visual positioning model is trained as follows:

[0070] Using historical: known positioning information of the target device, multiple configuration parameters of the drone that performs the positioning of the target device, AOA positioning algorithm associated parameters, and image data collected by the shooting device, the sample is constructed;

[0071] Allocate samples into training set, test set and validation set according to specified proportions;

[0072] The deep neural network is trained using the training set, tested using the test set, and verified using the validation set to obtain a trained intelligent visual positioning model.

[0073] In one embodiment, the number of times the deep neural network is trained is positively correlated with the total number of pixels occupied by the image captured indoors by the drone.

[0074] Figure 5 Schematic diagram of a computer device in an embodiment of the present invention. Figure 5 As shown, an embodiment of the present invention further provides a computer device 500, including a processor 501, a memory 502, and a computer program 503 stored in the memory 502 and executable on the processor 501, wherein the processor 501 implements the above-mentioned drone indoor target visual positioning method when executing the computer program 503.

[0075] An embodiment of the present invention further provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the above-mentioned method for visual positioning of indoor targets of unmanned aerial vehicles is implemented.

[0076] An embodiment of the present invention further provides a computer program product, which includes a computer program. When the computer program is executed by a processor, the above-mentioned drone indoor target visual positioning method is implemented.

[0077] In an embodiment of the present invention, multiple configuration parameters of a drone that performs target device positioning, AOA positioning algorithm associated parameters, and image data collected by a shooting device are obtained; the drone is equipped with a shooting device, the shooting device is electrically connected to the drone, and a pre-trained intelligent visual positioning model is configured in the drone; the configuration parameters include the positioning data of the drone and the camera parameters of the shooting device; the AOA positioning algorithm associated parameters include the parameters of the antenna deployed at the specified position of the target device and the amplitude of the wireless signal emitted by the built-in wireless signal transmitter of the drone; the target device information is extracted from the image data to obtain the color component value and the imaging depth of field value of the target device in the image data; the intelligent visual positioning model is used to parse and output the positioning information of the target device based on the multiple configuration parameters, the AOA positioning algorithm associated parameters, the color component value and the imaging depth of field value; the intelligent visual positioning model uses the historical: known positioning information of the target device, multiple configuration parameters of the drone that performs target device positioning, AOA positioning algorithm associated parameters, and image data collected by the shooting device to train a deep neural network. Compared with the technical solution of applying the AOA positioning algorithm in the prior art, the embodiment of the present invention combines the visual image data, the parameters of the antenna deployed at the specified position of the target device, and the AOA positioning algorithm associated parameters of the amplitude of the wireless signal transmitted by the built-in wireless signal transmitter of the drone, and adopts a deep neural network for positioning analysis, thereby avoiding the complex processing process of the existing AOA positioning algorithm, simplifying the indoor drone positioning processing process, reducing the indoor drone positioning processing time, and improving the high-precision positioning of indoor target devices.

[0078] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0079] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0080] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0081] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0082] The specific embodiments described above further illustrate the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above description is only a specific embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for visual positioning of indoor targets of unmanned aerial vehicles, characterized in that: include: Acquire multiple configuration parameters of a drone that performs target device positioning, parameters associated with an angle of arrival (AOA) positioning algorithm, and image data collected by a camera; the drone is equipped with a camera, the camera is electrically connected to the drone, and a pre-trained intelligent visual positioning model is configured in the drone; the configuration parameters include positioning data of the drone and camera parameters of the camera; the AOA positioning algorithm associated parameters include parameters of an antenna deployed at a designated position of the target device and an amplitude of a wireless signal transmitted by a built-in wireless signal transmitter of the drone; Extract target device information from the image data to obtain the color component value and imaging depth value of the target device in the image data; An intelligent visual positioning model is adopted to parse and output the positioning information of the target device based on the multiple configuration parameters, the AOA positioning algorithm associated parameters, the color component values ​​and the imaging depth of field values; the intelligent visual positioning model uses the historical: known positioning information of the target device, multiple configuration parameters of the drone that performs the target device positioning, the AOA positioning algorithm associated parameters, and the image data collected by the shooting device to train the deep neural network.

2. The method for indoor target visual positioning of a UAV according to claim 1, characterized in that: The camera parameters of the camera include one or any combination of the following: Shooting distance, horizontal resolution, vertical resolution, dynamic range width; The dynamic range width is the difference between the maximum grayscale value of a pixel that can be captured by the shooting device and the minimum grayscale value of a pixel that can be captured by the shooting device.

3. The method for indoor target visual positioning of a UAV according to claim 1, characterized in that: The parameters of the antenna deployed at the designated position of the target device include: the number of antennas in the antenna array and the distance between adjacent antennas.

4. The method for indoor target visual positioning of a UAV as claimed in claim 3, characterized in that: The multiple antennas in the antenna array have the same size and the antenna array is arranged in a rectangular array.

5. The method for indoor target visual positioning of a UAV according to claim 1, characterized in that: Extract target device information from the image data to obtain the color component value and imaging depth value of the target device in the image data, including: The target device information is extracted from the image data to obtain the color component value and imaging depth value of each pixel point in the imaging area of ​​the target device.

6. The method for indoor target visual positioning of a UAV as claimed in claim 1, characterized in that: The intelligent visual positioning model is trained as follows: Using historical: known positioning information of the target device, multiple configuration parameters of the drone that performs the positioning of the target device, AOA positioning algorithm associated parameters, and image data collected by the shooting device, the sample is constructed; Allocate samples into training set, test set and validation set according to specified proportions; The deep neural network is trained using the training set, tested using the test set, and verified using the validation set to obtain a trained intelligent visual positioning model.

7. The method for visual positioning of indoor targets of unmanned aerial vehicles according to claim 6, characterized in that: The number of times the deep neural network is trained is positively correlated with the total number of pixels occupied by the indoor image captured by the drone.

8. A UAV indoor target visual positioning device, characterized in that: include: A data acquisition module is used to acquire multiple configuration parameters of a drone that performs target device positioning, AOA positioning algorithm-related parameters, and image data collected by a shooting device; the drone is equipped with a shooting device, the shooting device is electrically connected to the drone, and a pre-trained intelligent visual positioning model is configured in the drone; the configuration parameters include the positioning data of the drone and the camera parameters of the shooting device; the AOA positioning algorithm-related parameters include the parameters of an antenna arranged at a designated position of the target device and the amplitude of a wireless signal transmitted by a built-in wireless signal transmitter of the drone; An image processing module is used to extract target device information from image data to obtain color component values ​​and imaging depth values ​​of the target device in the image data; The positioning analysis module is used to use an intelligent visual positioning model to parse and output the positioning information of the target device based on the multiple configuration parameters, the AOA positioning algorithm associated parameters, the color component values ​​and the imaging depth of field values; the intelligent visual positioning model uses the historical: known positioning information of the target device, multiple configuration parameters of the drone that performs the target device positioning, the AOA positioning algorithm associated parameters, and the image data collected by the shooting device to train the deep neural network.

9. A computer device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the method for visual positioning of indoor targets of unmanned aerial vehicles according to any one of claims 1 to 7 is implemented.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method for visual positioning of indoor targets of a drone according to any one of claims 1 to 7 is implemented.

11. A computer program product, characterized in that The computer program product includes a computer program, and when the computer program is executed by a processor, the method for visual positioning of indoor targets of unmanned aerial vehicles according to any one of claims 1 to 7 is implemented.

Images