An embedded information system satellite positioning mode control method

Abstract

The application provides an embedded information system satellite positioning mode control method, relates to the technical field of image processing, and is used for monitoring and management of position determination in the flight process of a UAV. The embedded information system satellite positioning mode control method specifically comprises the following steps: acquiring position information of a UAV device to be positioned; obtaining an overhead flight panoramic image by using a camera of the UAV device to be positioned; obtaining a high-resolution image corresponding to the position information acquired by a satellite system; and sending the high-resolution image to an embedded information system in the UAV device to be positioned, so that the embedded information system determines a positioning mode by using a recognition result of the high-resolution image. Through the above steps, the application can evaluate and correct the recognition accuracy of a positioning system, and ensure the safety of UAV flight in an unknown area.

Description

Technical Field

[0001] This invention relates to the field of image recognition, and more specifically to a satellite positioning mode control method for embedded information systems, used for monitoring and management of the position determination during the flight of unmanned aerial vehicles (UAVs). Background Technology

[0002] Drones, as a multifunctional aerial platform, have been widely applied in various fields of production and daily life. Supported by modern technology, drones can perform diverse tasks in different environments, such as damage assessment after natural disasters, environmental monitoring, and wildlife protection. Their rapid deployment capability makes them an ideal choice for preliminary situation assessments in emergency situations. In agriculture, drones use high-definition cameras to collect data on soil, crop growth, and pests and diseases, helping farmers make more accurate decisions. Drones also play a crucial role in urban planning and construction. They are used to monitor construction progress, conduct land surveying, and perform urban spatial planning. Through aerial photography, drones offer a faster, safer, and lower-cost data collection method than traditional ground surveying. Furthermore, in the power and oil industries, drones are used to inspect power lines and pipelines, which not only improves inspection efficiency but also significantly reduces the risks to personnel working in hazardous environments.

[0003] Despite the expanding applications and technological advancements of drones, positioning accuracy remains a critical issue during mission execution. Traditional drone positioning methods primarily rely on their onboard Global Positioning System (GPS). This system determines the drone's location by receiving signals from satellites in geostationary orbit. However, in complex or unknown environments, such as densely populated urban areas or mountainous forests, GPS signals can be blocked or interfered with, leading to inaccurate positioning. Furthermore, GPS provides the drone's absolute geographical location; it is ineffective in providing relative location information within the environment, such as the positions of nearby buildings, ground obstacles, or other aircraft. This limits the application of drones in unknown or complex environments, as they cannot effectively identify and avoid potential obstacles, which is crucial for ensuring drone flight safety and mission success.

[0004] To address the shortcomings of existing technologies, this invention proposes a satellite positioning mode control method for embedded information systems. This method integrates satellite positioning and image recognition technologies to provide more accurate location information and can also identify and process potential risks near the UAV's location in real time. This allows for more effective risk avoidance and improved flight safety. Summary of the Invention

[0005] This invention provides a satellite positioning mode control method for an embedded information system, which specifically includes the following steps:

[0006] S1: Obtain the location information of the drone device to be located;

[0007] S2: Obtain a panoramic image of the drone from above using the camera of the drone device to be positioned;

[0008] S3: Use the satellite system to capture a high-resolution image corresponding to the location information obtained in step S1;

[0009] S4: The high-resolution image is sent to the embedded information system in the drone device to be located, and the embedded information system uses the high-resolution image recognition result to determine the positioning mode.

[0010] The present invention also provides an embedded information system satellite positioning mode control device according to the embodiments of this specification, the device comprising:

[0011] Location positioning module: The location positioning module acquires the location information of the drone device to be located;

[0012] Panoramic Image Generation Module: The panoramic image generation module uses the camera of the UAV device to be positioned to capture a top-down panoramic image of the flight.

[0013] Satellite image acquisition module: The satellite image acquisition module utilizes high-resolution images corresponding to the location information acquired during satellite system photography;

[0014] Mode control module: The mode control module sends the high-resolution image to the embedded information system in the UAV device to be located, and the embedded information system uses the high-resolution image recognition result to determine the positioning mode.

[0015] An electronic device includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the aforementioned satellite positioning mode control method for an embedded information system.

[0016] A computer-readable storage medium storing a computer program that, when executed by a processor, implements the aforementioned satellite positioning mode control method for an embedded information system.

[0017] Compared with existing technologies, this invention proposes a satellite positioning mode control method for embedded information systems. In the process of UAV positioning, this invention first uses the positioning system configured on the UAV itself to locate its position. At the same time, based on the position information, a high-resolution image taken by a satellite with the position as the geometric center is acquired. That is, the position information is identified as the geometric center point of the high-resolution image taken by the satellite. By judging the displacement deviation between the UAV position in the actual high-resolution image taken by the satellite and the center point, the accuracy of the current positioning system is judged, and if it is inaccurate, other positioning systems are switched.

[0018] In the process of high-resolution image recognition, this invention utilizes the panoramic view taken by the UAV itself for guidance. The panoramic view taken by the UAV should be contained within the high-resolution image taken by the satellite. This invention improves the accuracy of UAV recognition by finding similar regions in the panoramic image and the high-resolution image. At the same time, the high-resolution image is obtained based on the position information emitted by the UAV. Under the condition that the position information is unbiased, the UAV should be located at the center of the high-resolution image. That is, the probability of the UAV appearing near the center of the high-resolution image is higher than the probability of appearing at the edge of the image. Therefore, this invention establishes a Gaussian distribution function with the geometric center of the high-resolution image as the midpoint to guide the recognition network to perform key target detection in the central region. Attached Figure Description

[0019] To more clearly illustrate the technical solutions of the embodiments of this application, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 This is a flowchart of the satellite positioning mode control process of the present invention; Detailed Implementation

[0021] The embodiments of this application will now be described in detail with reference to the accompanying drawings.

[0022] The following specific examples illustrate the implementation of this application. Those skilled in the art can easily understand other advantages and effects of this application from the content disclosed in this specification. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. This application can also be implemented or applied through other different specific embodiments, and the details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of this application. It should be noted that, in the absence of conflict, the following embodiments and features in the embodiments can be combined with each other. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0023] It should be noted that various aspects of embodiments within the scope of the appended claims are described below. It will be apparent that the aspects described herein can be embodied in a wide variety of forms, and any particular structure and / or function described herein is merely illustrative. Based on this application, those skilled in the art will understand that one aspect described herein can be implemented independently of any other aspect, and two or more of these aspects can be combined in various ways. For example, any number and aspects set forth herein can be used to implement the device and / or practice the method. Additionally, this device and / or method can be implemented using structures and / or functionalities other than one or more of the aspects set forth herein.

[0024] Additionally, specific details are provided in the following description to facilitate a thorough understanding of the examples. However, those skilled in the art will understand that practice can be carried out without these specific details.

[0025] This specification presents an embodiment of a satellite positioning mode control method for an embedded information system, which specifically includes the following steps:

[0026] S1: Obtain the location information of the drone device to be located;

[0027] S2: Obtain a panoramic image of the drone from above using the camera of the drone device to be positioned;

[0028] S3: Use the satellite system to capture a high-resolution image corresponding to the location information obtained in step S1;

[0029] S4: The high-resolution image is sent to the embedded information system in the drone device to be located, and the embedded information system uses the high-resolution image recognition result to determine the positioning mode.

[0030] Specifically, in one embodiment, the UAV device to be located should be equipped with multiple positioning systems, such as the US GPS, Russia's GLONASS, China's BeiDou, and the EU's Galileo system. Preferably, the UAV device to be located is equipped with a Global Navigation Satellite System (GNSS), which can receive signals from multiple satellite systems around the world. This multi-system reception capability ensures that positioning information can be obtained anywhere in the world. In particular, when the signal of one system is weak or unavailable, other systems can supplement the positioning information, improving the overall positioning reliability.

[0031] In addition, the UAV to be located should be equipped with an inertial measurement unit (IMU), which is a combination of accelerometers, gyroscopes, and sometimes magnetometers, used to provide the UAV's real-time motion status, including speed, direction, and attitude. The IMU employs microelectromechanical systems (MEMS) technology; IMU data is crucial when GNSS signals are unavailable or unstable, and can provide position and navigation information either auxiliaryly or independently.

[0032] The drone to be located should be equipped with a communication module, which is responsible for data exchange with the ground control station or internet cloud service. This data exchange includes flight data, mission control commands, real-time video streams, and location information. The communication module includes one or more technologies such as Wi-Fi, Bluetooth, LTE, and 5G. For missions requiring long distances or high reliability, LTE and 5G provide stable connections across wide areas, support high data rate transmission, and are well-suited for transmitting high-definition video or large amounts of sensor data. Simultaneously, these modules should be configured with security protocols to ensure the security and privacy of data transmission.

[0033] The process of obtaining a top-down panoramic image of flight using the camera of the drone device to be positioned specifically includes:

[0034] S2-1: When the location information of the drone device to be located is obtained, the camera of the drone device is activated and video recording begins, and the start recording time is recorded.

[0035] S2-2: Record a video of a preset duration, and extract keyframes and preprocess the video;

[0036] S2-3: Stitch together the preprocessed keyframes to obtain a bird's-eye view panoramic image of the flight.

[0037] The keyframe extraction from the video includes:

[0038] S2-2-1: Initialize the keyframe list L, set the first frame F1 of the video as the first keyframe K1, and add it to the keyframe list L;

[0039] S2-2-2: For the i-th frame F i Calculate F i With keyframe K j The keyframe K contains the motion component MV, local description difference LD, global description difference FD, and similarity S. j The keyframes are the latest ones added to the keyframe list L, where 2≤i≤N, 1≤j<i, and N is the total number of video frames;

[0040] S2-2-3: Calculate the overall difference D. If the overall difference D is greater than the preset difference threshold δ... D And the similarity S is greater than the preset overlap threshold δ S When, then the i-th frame F i As a potential keyframe, update i = i+1, and repeat steps S2-2-2 and S2-2-3 until the number of iterations reaches the preset time window τ or i > N;

[0041] S2-2-4: Within a preset time window τ, select the frame with the largest overall difference D among the potential keyframes as the keyframe K. j+1 Add to L, update j = j + 1, and repeat steps S2-2-2, S2-2-3 and S2-2-4 until all video frames are traversed, and output the keyframe list L.

[0042] The mobile component MV is defined as:

[0043]

[0044] Where, N f The number of feature points, and For the i-th frame F i and the j-th keyframe K j The coordinates of the nth matching feature point;

[0045] The local descriptive difference (LD) is defined as follows:

[0046]

[0047] Where, N p Here, Dim represents the number of key points, and desc represents the local description dimension. i,n [d] and desc j，n [d] represents the i-th frame F i and the j-th keyframe K j The d-th dimension of the nth matching keypoint;

[0048] In one implementation, the local description may employ the SIFT algorithm, where the SIFT descriptor is a 128-dimensional vector that describes the image gradient information around the keypoint.

[0049] The global description difference FD is defined as follows:

[0050]

[0051] Where, N B The number of buckets in the histogram, hist i [b] and hist j [b] represents the i-th frame F i and the j-th keyframe K j The b-th bucket in the histogram.

[0052] This invention first extracts key frames from the video captured by the drone when acquiring panoramic images. During the video image capture process, this invention does not require controlling the flight speed or direction of the drone. In the process of selecting key frames, this invention calculates the inter-frame difference using the motion component, local description difference, and global description difference to ensure the computational efficiency of key frame extraction. At the same time, it calculates the inter-frame similarity to ensure the smoothness of the subsequent panoramic image construction.

[0053] The process of stitching together the preprocessed keyframes to obtain a bird's-eye view panoramic image specifically includes:

[0054] S2-3-1: For the preceding and following keyframes K... j and K j+1 Feature maps f are obtained by performing feature extraction separately. j and f j+1 ;

[0055] S2-3-2: For feature map f j and f j+1 Perform feature matching to obtain a related feature description map;

[0056] S2-3-3: Utilize the analytical network to output the affine transformation parameter matrix based on the associated feature description map;

[0057] S2-3-4: Perform image registration based on the affine transformation parameter matrix and fuse them to obtain the i-th panoramic image;

[0058] S2-3-5: Repeat steps S2-3-1 to S2-3-4 until all keyframes are traversed, and stitch together to output a top-down panoramic flight image.

[0059] The keyframes K before and after j and K j+1 Feature maps f are obtained by performing feature extraction separately. j and f j+1 The feature extraction network consists of four residual modules, each designed as follows:

[0060] O t =ReLU(BN(W 3*3 *ReLU(BN(W 1*1 *I t +b 1*1 ))+b 3*3 )+I t );

[0061] Among them, I t and O t W represents the input and output of the t-th residual module. 1*1 and W 3*3 These are the 1x1 and 3x3 convolutional kernel weights, respectively, b 1*1 and b 3*3 These are the corresponding bias terms. BN and ReLU are the batch normalization and activation functions, respectively. The stride and padding are both 1 in the convolution calculation. The input to the first residual module is any keyframe K in the keyframe list L. j The output of the fourth residual module is the feature map f. j ;

[0062] The method for calculating the associated feature description map is as follows:

[0063]

[0064] Among them, f j and f j+1 The feature map size is w*h*N c The associated feature description map Cr is sized as w*h*(w*h), where w and h represent the length and width, respectively, and (w*h) is the number of channels in the Cr. N c Representing the feature map f j and f j+1 The number of channels, x, y, z represent the length, width, and channel index of the feature map, respectively;

[0065] The parsing network is configured as follows:

[0066] f mid =ReLU(W2*(MaxPool(ReLU(W1*Crl+b1)))+b2);

[0067] f affine =FC 2 (Flatten(f mid ));

[0068] Where W1 and W2 are the parameters of the first and second convolutional kernels, respectively, b1 and b2 are the bias terms of the first and second layers, respectively, and f mid f represents the hidden features in the middle of the parsing network. affineLet F_t represent the affine transformation parameter matrix, and Flatten denote the flattening operation. 2 This represents two fully connected mapping operations.

[0069] In the process of panoramic image stitching, this invention first extracts the feature points of the keyframes to be stitched. Unlike traditional feature point extraction methods, this application constructs a residual module to extract the feature points of the keyframes. The panoramic image of this invention is stitched from keyframe images, and there are significant differences between the keyframe images. Therefore, the feature point extraction method based on neural networks can extract the deep semantic features between keyframes, and obtain the affine transformation parameter matrix after the keypoint extraction is completed, thereby realizing the keyframe image stitching.

[0070] In one implementation, the high-resolution image corresponding to the location information acquired in step S1 using the satellite system needs to be transmitted to the ground control center via the communication equipment configured for the unmanned aerial vehicle (UAV). The location information includes the UAV's longitude, latitude, and altitude. After receiving the location information, the control center verifies its accuracy and completeness and performs format conversion to adapt to the satellite control system. The control center calculates the camera's shooting adjustment amount by analyzing the UAV's location information and satellite status information. This adjustment includes the satellite lens's pitch and yaw angles to ensure the satellite camera is aligned with the UAV's location. Once the ground control center has completed processing, it transmits the control information to the satellite control system via the communication equipment for lens angle adjustment. Once the satellite has correctly adjusted the lens angle and resolution, it enters the shooting program. In this stage, the satellite operating system controls the camera to capture high-resolution images according to control parameters and transmits the high-resolution image data to the ground control center. After receiving the image data, the ground control center performs preliminary image processing and image enhancement.

[0071] The method of adjusting the attitude of remote sensing satellites using ground location information is existing technology and will not be elaborated here. In one implementation, the process of using a satellite system to photograph UAV equipment also includes adjusting the resolution of the satellite system's camera and fine-tuning the satellite's orbit. Both resolution adjustment and satellite orbit adjustments are existing technologies and will not be elaborated here either. References: "An On-Orbit Autonomous Focusing Method for Agile Remote Sensing Satellites Based on Lunar Imaging," Yin Yanhe, Proceedings of the 8th National Conference on High-Resolution Earth Observation; "A Single-Orbit Scheduling Method for Agile Imaging Satellites Oriented to Regional Targets," Song Jinyun, Digital Technology and Applications.

[0072] In one embodiment, the embedded information system uses the high-resolution image recognition results to determine the positioning mode, and employs an image fusion network to identify the location of the UAV.

[0073] The image fusion network includes a high-resolution image feature extraction module, a multi-scale fusion module, an information guidance module, and a localization module;

[0074] The high-resolution image feature extraction module includes four enhanced residual skip structures, wherein the first enhanced residual skip structure is defined as:

[0075]

[0076] Among them, I H Represents a high-resolution image. This represents the output feature map of the first enhanced residual skip structure;

[0077] The second to fourth enhanced residual jump structures are defined as follows:

[0078]

[0079] in, The output feature map of the m-th enhanced residual jump structure is represented, where 2≤m≤4;

[0080] The multi-scale fusion module includes three parallel branches for feature enhancement, and the input of the three parallel branches is the output of the fourth enhanced residual skip structure. The first and second layers of the three parallel branches are defined as follows:

[0081]

[0082]

[0083] Where 1≤p≤3, and Conv represents the first and second layer output feature maps of the p-th parallel branch, respectively. p,1 and Conv p,2 These represent the convolution calculations of the first and second layers of the p-th parallel branch, respectively, where the kernel sizes of the first to third parallel branches are 3*3, 5*5, and 7*7, respectively.

[0084] The outputs of the three parallel branches are accumulated and fused pixel by pixel to obtain the output O of the multi-scale fusion module. M ;

[0085] The input to the information guidance module is the output O of the multi-scale fusion module. M The information guidance module is defined as follows:

[0086]

[0087]

[0088]

[0089]

[0090] Among them, O inf The output feature map F of the information guidance module represents the output feature map. s F loc and F img These represent location satellite image guidance, location information guidance, and panoramic information guidance, respectively, σ, add, and These represent activation functions, pixel-wise addition, and pixel-wise multiplication, respectively. These represent 1x1 convolution operations used for query, key, value, and output channel adjustment, respectively. C dim This represents the number of channels after dimensionality reduction. `expand` and `reshape` represent channel expansion and size adjustment, respectively. c and y c O M The center point coordinates σ x and σ y I represents the standard deviation of the Gaussian distribution along the horizontal and vertical axes, respectively. par To create a panoramic image of the flight from above, max axis＝1 This indicates that the maximum value should be retrieved from each row.

[0091] The output of the information guidance module O inf The data is input into a positioning module composed of a fully connected network to obtain the location results of the UAV device in the satellite image;

[0092] If the distance between the target UAV device in the satellite image and the center of the image is less than a preset threshold, the current positioning system is used to continue locating the UAV device; otherwise, another positioning system is used to continue locating the UAV device.

[0093] In the process of drone positioning, this invention first uses the positioning system configured on the drone itself to locate its position. At the same time, it acquires a high-resolution satellite image with the location as the geometric center based on the location information. That is, the location information is identified as the geometric center point of the high-resolution satellite image. By judging the displacement deviation between the drone's position in the actual high-resolution satellite image and the center point, it determines whether the current positioning system is accurate, and if it is inaccurate, it switches to another positioning system.

[0094] In the process of high-resolution image recognition, this invention utilizes the panoramic view taken by the UAV itself for guidance. The panoramic view taken by the UAV should be contained within the high-resolution image taken by the satellite. This invention improves the accuracy of UAV recognition by finding similar regions in the panoramic image and the high-resolution image. At the same time, the high-resolution image is obtained based on the position information emitted by the UAV. Under the condition that the position information is unbiased, the UAV should be located at the center of the high-resolution image. That is, the probability of the UAV appearing near the center of the high-resolution image is higher than the probability of appearing at the edge of the image. Therefore, this invention establishes a Gaussian distribution function with the geometric center of the high-resolution image as the midpoint to guide the recognition network to perform key target detection in the central region.

[0095] This specification also proposes an embedded information system satellite positioning mode control device, which includes:

[0096] Location positioning module: The location positioning module acquires the location information of the drone device to be located;

[0097] Panoramic Image Generation Module: The panoramic image generation module uses the camera of the UAV device to be positioned to capture a top-down panoramic image of the flight.

[0098] Satellite image acquisition module: The satellite image acquisition module utilizes high-resolution images corresponding to the location information acquired during satellite system photography;

[0099] Mode control module: The mode control module sends the high-resolution image to the embedded information system in the UAV device to be located, and the embedded information system uses the high-resolution image recognition result to determine the positioning mode.

[0100] An electronic device includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the aforementioned satellite positioning mode control method for an embedded information system.

[0101] A computer-readable storage medium storing a computer program that, when executed by a processor, implements the aforementioned satellite positioning mode control method for an embedded information system.

[0102] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. 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), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0103] In this specification, the same or similar parts between the various embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, the descriptions of the embodiments described later are relatively simple, and relevant parts can be referred to the descriptions of the foregoing embodiments.

[0104] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. An embedded information system satellite positioning mode control method, characterized by, The method comprises the following steps: S1: obtaining position information of a to-be-positioned unmanned aerial vehicle device; S2: obtaining an overhead flight panoramic image by using a camera of the to-be-positioned unmanned aerial vehicle device; S3: obtaining a high-resolution image corresponding to the position information obtained in step S1 by using a satellite system; S4: sending the high-resolution image to an embedded information system in the to-be-positioned unmanned aerial vehicle device, and determining a positioning mode by using a high-resolution image recognition result of the embedded information system; In the step of determining the positioning mode by using the high-resolution image recognition result of the embedded information system, an image fusion network is used for unmanned aerial vehicle position recognition; the image fusion network comprises a high-resolution image feature extraction module, a multi-scale fusion module, an information guiding module, and a positioning module; The input of the information guiding module is the output O of the multi-scale fusion module M The information guiding module is defined as: ； ； wherein, O inf represents the output feature map of the information guidance module, F s , F loc and F img represent position satellite image guidance, position information guidance and panoramic information guidance respectively, σ, add and respectively represent an activation function, pixel-wise addition and pixel-wise multiplication, respectively represent 1*1 convolution operations for query, key, value and output channel adjustment, C dim represents the number of channels after dimension reduction, expand and reshape respectively represent channel expansion and size adjustment, x c and y c represent the center point coordinates of O M , σ x and σ y respectively represent the standard deviation of the Gaussian distribution along the horizontal and vertical axes, I par is a top-view flight panoramic image, max axis＝1 represents taking the maximum value by row; the output O inf of the information guidance module is input into a positioning module composed of a fully connected network to obtain the result of the unmanned device to be positioned in the satellite image; When a result of the to-be-positioned unmanned aerial vehicle device in the satellite image is less than a preset threshold from a center position of the image, a current position positioning system is used to continue to position the unmanned aerial vehicle device, otherwise, another positioning system is used to continue to position the unmanned aerial vehicle device.

2. The method of claim 1, wherein the method further comprises: The step of obtaining the overhead flight panoramic image by using the camera of the to-be-positioned unmanned aerial vehicle device comprises the following steps: S2-1: starting the camera of the unmanned aerial vehicle device and beginning to record a video when the position information of the to-be-positioned unmanned aerial vehicle device is obtained, and recording a starting recording time; S2-2: recording a preset time length of video, and performing key frame extraction and preprocessing on the video; S2-3: splicing the preprocessed key frames to obtain the overhead flight panoramic image.

3. The method of claim 2, wherein the method further comprises: The step of extracting the key frames from the video comprises the following steps: S2-2-1: initializing a key frame list L, setting a first frame F1 of the video as a first key frame K1, and adding the first key frame K1 to the key frame list L; S2-2-2: For the i-th frame F i , calculate the motion vector MV, local description difference LD, global description difference FD, and similarity S between F i and the key frame K j , where the key frame K j is the key frame newly added to the key frame list L, 2 ≤ i ≤ N, 1 ≤ j < i, and N is the total number of video frames; S2-2-3: calculate the comprehensive difference D, if the comprehensive difference D is greater than the preset difference threshold δ D and the similarity S is greater than the preset overlap threshold δ S , then the i-th frame F i is taken as a potential key frame, i is updated to i+1, steps S2-2-2 and S2-2-3 are repeatedly executed until the iteration number reaches the preset time window τ. S2-2-4: In a preset time window τ frames, select the frame with the largest integrated difference D in the potential key frames as the key frame K j+1 Add L, update j = j + 1, and repeat steps S2-2-2, S2-2-3 and S2-2-4 until all video frames are traversed, and output the key frame list L.

4. The method of claim 3, wherein: The step of splicing the preprocessed key frames to obtain the overhead flight panoramic image comprises the following steps: S2-3-1: to the front and back key frame K j and K j+1 respectively feature extraction to get feature map f j and f j+1 ; S2-3-2: performing feature matching on the feature map f j and f j+1 performing feature matching to obtain a correlation feature description map; S2-3-3: outputting an affine transformation parameter matrix according to the associated feature description graph by using an analysis network; S2-3-4: performing image registration according to the affine transformation parameter matrix, and fusing to obtain a jth panoramic image; S2-3-5: repeating steps S2-3-1 to S2-3-4 until all the key frames are traversed, and splicing to output the overhead flight panoramic image.

5. The method of claim 4, wherein: The high-resolution image feature extraction module comprises four enhanced residual jump structures, wherein the first enhanced residual jump structure is defined as: ； wherein I H represents a high-resolution image, represents the output feature map of the first enhanced residual skip structure; The second to fourth enhanced residual jump structures are defined as: ； wherein, denotes the output feature map of the mth enhanced residual skip structure, 2≤m≤4, and ReLU is an activation function.

6. The method of claim 5, wherein: The multi-scale fusion module includes three parallel branches for feature enhancement, and the inputs of the three parallel branches are the outputs of the fourth enhanced residual jump structure The first layer and the second layer of the three parallel branches are defined as follows, respectively: ； wherein 1≤p≤3, and denote the first and second layer output feature maps of the pth parallel branch, respectively, Conv p,1 and Conv p,2 denote the convolution calculation of the first and second layers of the pth parallel branch, respectively, wherein the convolution kernel sizes of the first to third parallel branches are 3*3, 5*5 and 7*7, respectively; BN is batch normalization. The outputs of the three parallel branches are pixel by pixel accumulated and fused to obtain the output O of the multi-scale fusion module M .

7. An embedded information system satellite positioning mode control apparatus, characterized by, The device is used to execute the embedded information system satellite positioning mode control method according to any one of claims 1-6, and the system comprises: A position positioning module: the position positioning module obtains position information of a to-be-positioned unmanned aerial vehicle device; A panoramic image generation module: the panoramic image generation module obtains an overhead flight panoramic image by using a camera of the to-be-positioned unmanned aerial vehicle device; A satellite image acquisition module: the satellite image acquisition module obtains a high-resolution image corresponding to the position information obtained in the step of using a satellite system to shoot; The mode control module sends the high-resolution image to an embedded information system in the unmanned aerial vehicle device to be positioned, and the embedded information system determines a positioning mode according to the high-resolution image recognition result. 8.An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the method of claim 1-6 when executing the computer program. 9.A computer readable storage medium storing a computer program, wherein the computer program is executable on a processor to implement the method of claim 1-6.

Images