Multi-unmanned aerial vehicle image splicing method and system based on control information

Abstract

The invention relates to the field of image processing, and provides a multi-unmanned aerial vehicle image splicing method and system based on control information. The method comprises the steps of collecting original image data of multiple unmanned aerial vehicles and flight information of the multiple unmanned aerial vehicles; correcting the corresponding original image data through the flight information of the unmanned aerial vehicle to obtain corrected image data; performing feature extraction on the corrected image data to obtain a matching point pair set; and constructing a self-adaptive homography matrix according to the matching point pair set, and splicing multiple pieces of corrected image data based on the self-adaptive homography matrix to obtain a splicing result. According to the invention, the precision and efficiency of image splicing can be improved.

Description

Technical Field

[0001] This invention relates to the field of image processing technology, and in particular to a method and system for stitching multi-UAV images based on control information. Background Technology

[0002] With the rapid development of UAV technology, multi-UAV collaborative operations have been widely applied in fields such as aerial photography, geographic surveying, and environmental monitoring. Existing multi-UAV image stitching technologies are mainly based on traditional image registration methods, achieving image stitching through steps such as extracting image feature points, calculating feature descriptors, and performing feature matching. Typical methods include feature point extraction and matching based on the SIFT (Scale-Invariant Feature Transform) algorithm, fast feature detection based on the SURF (Speeded Up Robust Features) algorithm, and real-time feature matching based on the ORB (Oriented FAST and Rotated BRIEF) algorithm. These methods perform well when processing image sequences from a single UAV, achieving relatively accurate image stitching results.

[0003] However, existing technologies have significant shortcomings when processing images captured collaboratively by multiple drones. First, traditional feature matching methods rely solely on image content for registration, ignoring the rich control information generated during drone flight (such as GPS positioning, IMU attitude, and flight altitude), resulting in low matching efficiency and a high risk of mismatches. Second, existing methods lack global consistency constraints when processing multi-drone images, failing to fully utilize the relative positional and temporal relationships between drones, which easily leads to accumulated errors and geometric distortions. Finally, existing technologies have limited ability to handle scale differences and perspective changes caused by different flight altitudes and shooting angles in multi-drone systems, affecting the accuracy and visual quality of the final stitching result. Summary of the Invention

[0004] This invention provides a method and system for stitching multi-UAV images based on control information, in order to overcome the shortcomings of the prior art.

[0005] This invention provides a method for stitching multi-UAV images based on control information, comprising: S1: Collect raw image data and flight information of multiple drones; S2: Correct the corresponding raw image data using the drone's flight information to obtain corrected image data; S3: Perform feature extraction on the corrected image data to obtain a set of matching point pairs; S4: Construct an adaptive homography matrix based on the set of matching point pairs, and stitch together multiple corrected image data based on the adaptive homography matrix to obtain the stitching result.

[0006] According to the multi-UAV image stitching method based on control information provided by the present invention, step S1 further includes: S11: Collect flight information of the UAV through the flight control system to obtain a flight information set; S12: Acquire raw image data from multiple drones to obtain a raw image set; S13: Associate and store the flight information set with the original image set to obtain an information database.

[0007] According to the multi-UAV image stitching method based on control information provided by the present invention, the flight information in step S11 includes: yaw angle, pitch angle, roll angle, and GPS positioning data; the original image data in step S12 includes: image data and corresponding image acquisition timestamp.

[0008] According to the multi-UAV image stitching method based on control information provided by the present invention, the expression of the corrected image data in step S2 is: ; in, The original image data before correction. For the heading angle, This is the corrected image data.

[0009] According to the multi-UAV image stitching method based on control information provided by the present invention, step S3 further includes: S31: Perform Gaussian smoothing on the corrected image data to obtain smoothed image data; S32: Select the pixel corresponding to the minimum gray value from the smoothed image data as the feature point; S33: Calculate the Euclidean distance between the feature points of the current image and all pixels in the neighboring images of the current image, and output the pixel corresponding to the minimum value of the Euclidean distance as the matching point; S34: Output the matching points of the current image and the matching points on the neighboring images of the current image as matching point pairs, and obtain a set of matching point pairs.

[0010] According to the multi-UAV image stitching method based on control information provided by the present invention, step S4 further includes: S41: Based on the coordinate values ​​of multiple matching point pairs in the set of matching point pairs, establish an adaptive homography matrix; S42: Perform homography transformation on multiple corrected image data using the adaptive homography matrix to obtain multiple transformed image data; S43: Calculate the temporal movement parameters of image sequences of multiple UAVs by transforming the image center point position of the image data; S44: Sequentially stitch together multiple transformed image data according to the time-shifting parameters to obtain a stitched image sequence, and output the stitching result.

[0011] According to the multi-UAV image stitching method based on control information provided by the present invention, the expression for the temporal movement parameter in step S43 is as follows: ; in, To transform the image data index value, For the first Temporal shift parameters of Zhang transform image data, For the first The x-coordinate of the center point of the transformed image data The width of the reference image.

[0012] Based on the time-series shift parameters, step S44 specifically includes: In the image sequence to which the transformed image data belongs, when the temporal shift parameter of the current image is 1, the next frame of the image to be stitched is used for stitching. In the image sequence to which the transformed image data belongs, when the temporal shift parameter of the current image is -1, the previous frame of the image to be stitched is used for stitching. In the image sequence to which the transformed image data belongs, when the temporal shift parameter of the current image is 0, the current image sequence is output to obtain the stitched image sequence, and the stitching result is output.

[0013] A second aspect of the present invention provides a multi-UAV image stitching system based on control information, comprising: Acquisition module: Used to acquire raw image data and flight information of multiple drones; Correction module: Used to correct the corresponding raw image data based on the UAV's flight information to obtain corrected image data; Extraction module: used to extract features from the corrected image data to obtain a set of matching point pairs; The stitching module is used to construct an adaptive homography matrix based on the set of matching points, and stitch multiple corrected image data together based on the adaptive homography matrix to obtain the stitching result.

[0014] A third aspect of the present invention provides a multi-UAV image stitching device based on control information, comprising: A memory and at least one processor, wherein the memory stores instructions; At least one of the processors invokes the instructions in the memory to cause a multi-UAV image stitching device based on control information to perform a multi-UAV image stitching method based on control information as described in any of the preceding claims.

[0015] A fourth aspect of the present invention provides a computer-readable storage medium storing instructions that, when executed by a processor, implement a multi-UAV image stitching method based on control information as described in any of the preceding claims.

[0016] This invention provides a multi-UAV image stitching method, system, device, and storage medium based on control information. By collecting flight information from multiple UAVs and associating it with the original image data to construct an information database, it provides rich prior constraint information for subsequent image processing, significantly improving the accuracy and stability of image stitching. This avoids the mismatch and accumulated error problems that easily occur when traditional methods rely solely on image content for matching, while also greatly reducing the computational complexity of feature matching and improving processing efficiency. Subsequently, this invention corrects the original image data using flight information, effectively eliminating the impact of UAV flight attitude changes on the geometric accuracy of the image, laying a good foundation for subsequent feature extraction and matching, resulting in stitching results with higher geometric consistency and spatial accuracy. Secondly, the feature extraction method based on corrected image data, through Gaussian smoothing preprocessing and gray-level extreme value detection, obtains more stable and reliable feature points, reduces noise interference, and improves the repeatability and discriminability of feature points, thereby enhancing the robustness of the subsequent matching process. Combining Euclidean distance calculation to determine matching point pairs can quickly and accurately identify the correspondence between adjacent images, avoiding the need for traditional complex feature descriptor calculations. The time overhead of computation is significantly reduced, improving matching efficiency, making it particularly suitable for real-time processing of large-scale multi-UAV image data. Finally, an adaptive homography matrix is ​​constructed and homography transformation is performed, accurately describing the geometric transformation relationships between images at different viewpoints and altitudes. This effectively addresses scale differences and viewpoint variations in multi-UAV systems, ensuring the geometric accuracy of the stitching results. By calculating the image center point position of the transformed image data to determine the temporal movement parameters, the stitching order of multiple images can be rationally arranged, fully utilizing the temporal information of UAV flight and avoiding geometric distortions and visual discontinuities that may result from random stitching, making the final stitching result more natural and smooth. Furthermore, the sequential stitching strategy based on temporal movement parameters in this invention can intelligently stitch images according to their actual spatial and temporal relationships, ensuring stitching integrity, optimizing the stitching path, reducing unnecessary redundant calculations, and improving overall processing efficiency. Simultaneously, through reasonable inter-frame stitching logic, error propagation during the stitching process can be effectively controlled, ensuring the stability and reliability of large-scale multi-UAV image stitching. This provides a high-quality image stitching solution for practical applications such as aerial photography, geographic surveying, and environmental monitoring. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0018] Figure 1 A flowchart illustrating a multi-UAV image stitching method based on control information provided by the present invention; Figure 2 This is a schematic diagram of the structure of a multi-UAV image stitching system based on control information provided by the present invention.

[0019] Figure descriptions: 100, Acquisition module; 200, Calibration module; 300, Extraction module; 400, Stitching module. Detailed Implementation

[0020] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, embodiments of this invention, and should not be construed as limiting the invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention. In the description of this invention, it should be understood that the terminology used is for descriptive purposes only and should not be construed as indicating or implying relative importance.

[0021] The embodiments of the present invention are described below with reference to the figures.

[0022] like Figure 1 As shown, the present invention provides a method for stitching multi-UAV images based on control information, including: S1: Collect raw image data and flight information of multiple drones.

[0023] The flight information in step S1 includes: yaw angle, pitch angle, roll angle, and GPS positioning data. The raw image data in step S1 includes: image data and corresponding image acquisition timestamps.

[0024] Furthermore, flight information includes parameters such as yaw angle, pitch angle, roll angle, and GPS positioning data. Yaw angle refers to the angle between the direction of the UAV's nose and due north, pitch angle refers to the angle between the UAV's longitudinal axis and the horizontal plane, roll angle refers to the angle of rotation of the UAV's body around its longitudinal axis, and GPS positioning data includes the UAV's latitude and longitude coordinates and altitude information.

[0025] Step S1 further includes: S11: Collect flight information of the UAV through the flight control system to obtain a flight information set; S12: Obtain raw image data of multiple UAVs to obtain a raw image set.

[0026] Specifically, in step S11, the UAV's flight control system monitors the UAV's attitude changes in real time through its built-in IMU (Inertial Measurement Unit). The IMU includes a three-axis gyroscope and a three-axis accelerometer. The gyroscope measures the angular velocity change and integrates it to obtain the attitude angle. The accelerometer measures the linear acceleration and calculates the attitude by combining it with the gravity vector. At the same time, the GPS receiver receives satellite signals and calculates the UAV's spatial position coordinates through a triangulation algorithm. The flight control system collects and stores the attitude data and position data at a preset frequency as a flight information set. In step S12, the digital camera on the UAV collects images at a preset shooting interval. The camera's image sensor converts the light signal into digital image data. Each time a shot is taken, the camera synchronously records the current system timestamp, forming a correspondence between image data and timestamps. The image data from multiple UAVs are aggregated to form an original image set.

[0027] S13: Associate and store the flight information set with the original image set to obtain an information database.

[0028] Specifically, in step S13, the present invention accurately associates the flight information set with the original image set through timestamps. Since the flight control system and the camera system are based on a unified time reference, each image is matched one-to-one with the flight information at the time of its capture by timestamp matching, and a complete information database is established. The control information obtained in the information database will serve as the basis for image matching and geometric correction in subsequent steps to ensure the accuracy and reliability of the stitching results.

[0029] S2: Correct the corresponding original image data using the drone's flight information to obtain corrected image data.

[0030] The expression for the corrected image data in step S2 is: ; in, The original image data before correction. For the heading angle, This is the corrected image data.

[0031] Furthermore, the yaw angle is obtained by fusing data from the magnetic compass and gyroscope of the flight control system. The rotation transformation matrix established in this invention is a two-dimensional rotation matrix containing a linear combination of cosine and sinine. This invention uses matrix multiplication to rotate the coordinates of each pixel in the original image, aligning the image direction with standard north. During the correction process, for each pixel (x, y) in the image, it is first converted to a coordinate system with the image center as the origin, then the rotation transformation matrix is ​​applied for coordinate transformation, and finally the transformed coordinates are remapped back to the pixel coordinate system, completing the redistribution of pixel values.

[0032] S3: Perform feature extraction on the corrected image data to obtain a set of matching point pairs.

[0033] Step S3 further includes: S31: Perform Gaussian smoothing on the corrected image data to obtain smoothed image data.

[0034] In step S31, the corrected image data is first convolved using a Gaussian smoothing algorithm to reduce image noise and blur details. Each pixel value of the obtained smoothed image data is the result of the original pixel values ​​being weighted by Gaussian weights.

[0035] S32: Select the pixel corresponding to the minimum gray value from the smoothed image data as the feature point.

[0036] Furthermore, the minimum grayscale value refers to the grayscale value of a pixel being less than the grayscale values ​​of all other pixels in its local neighborhood. During data processing, the smoothed image data in step S31 first needs to be converted to a grayscale image. If the original image is a color image, the pixel values ​​of the RGB channels are converted to a single grayscale value through weighted averaging. Subsequently, each pixel in the grayscale image is traversed. In this embodiment, any pixel is first selected as the current pixel, and the grayscale values ​​of its eight neighboring pixels within a 3×3 neighborhood are checked. If the grayscale value of the current pixel is less than the grayscale values ​​of all neighboring pixels in the neighborhood, then the pixel is marked as a feature point.

[0037] S33: Calculate the Euclidean distance between the feature points of the current image and all pixels in the neighboring images of the current image, and output the pixel corresponding to the minimum Euclidean distance as the matching point.

[0038] Furthermore, this invention uses Euclidean distance to measure the distance between two points in multidimensional space. During the distance calculation process, the feature points of the current image are calculated one by one with all pixels in the adjacent images to form a distance matrix. Each element in the matrix represents the Euclidean distance between a feature point and a pixel. The number of rows in the distance matrix is ​​equal to the number of feature points in the current image, and the number of columns is equal to the total number of pixels in the adjacent images. The pixel corresponding to the minimum value of the Euclidean distance is determined by traversing each row of the distance matrix to find the minimum value position. The pixel corresponding to this position is the matching point of the current feature point.

[0039] S34: Output the matching points of the current image and the matching points on the neighboring images of the current image as matching point pairs, and obtain a set of matching point pairs.

[0040] Each matching point pair obtained in step S34 contains two elements: a feature point in the current image and a corresponding matching point in the adjacent image. The data structure of the matching point pair contains four fields: coordinates of the feature point in the current image, gray value of the feature point in the current image, coordinates of the matching point in the adjacent image, and gray value of the matching point in the adjacent image. The matching point pair set is constructed by traversing all feature points in the current image and finding the corresponding matching point for each feature point. Each matching point pair in the set records the correspondence between the two images.

[0041] S4: Construct an adaptive homography matrix based on the set of matching point pairs, and stitch together multiple corrected image data based on the adaptive homography matrix to obtain the stitching result.

[0042] Step S4 further includes: S41: Based on the coordinate values ​​of multiple matching point pairs in the set of matching point pairs, establish an adaptive homography matrix.

[0043] Furthermore, in step S41, the present invention first extracts the source image coordinates and target image coordinates of each matching point pair in the matching point pair set, and then uses the least squares method to solve for the nine parameters of the homography matrix H, wherein the mathematical expression of the homography matrix is: ; Subsequently, a system of linear equations was established. To solve for the matrix parameters, where The matrix consists of the coordinates of the matching point pairs. The matrix consists of target coordinate values, and the solution process uses the singular value decomposition (SVD) method to ensure the numerical stability of the matrix.

[0044] S42: Perform homography transformation on multiple corrected image data using the adaptive homography matrix to obtain multiple transformed image data.

[0045] Furthermore, in step S42, the present invention transforms the coordinates of each pixel in the multiple corrected image data using an adaptive homography matrix. Specifically, for any pixel in the source image, the transformed homogeneous coordinates are obtained through matrix multiplication and adaptive homography matrix operation, and then the actual coordinates are obtained through normalization.

[0046] S43: Calculate the temporal movement parameters of image sequences of multiple UAVs by transforming the image center point position of the image data.

[0047] The expression for the timing shift parameter in step S43 is: ; in, To transform the image data index value, For the first Temporal shift parameters of Zhang transform image data, For the first The x-coordinate of the center point of the transformed image data The width of the reference image.

[0048] Furthermore, the time-series shift parameter calculated by this invention reflects the direction and magnitude of the UAV image sequence's movement over time. The above expression means that when the x-coordinate of the center point of the transformed image data is greater than half the width of the reference image, the time-series shift parameter is 1, indicating movement to the right; when the x-coordinate of the center point is less than half the width of the reference image, the time-series shift parameter is -1, indicating movement to the left; and when the x-coordinate of the center point is equal to half the width of the reference image, the time-series shift parameter is 0, indicating no movement.

[0049] S44: Sequentially stitch together multiple transformed image data according to the time-shifting parameters to obtain a stitched image sequence, and output the stitching result.

[0050] Specifically, step S44 includes: In the image sequence to which the transformed image data belongs, when the temporal shift parameter of the current image is 1, the next frame of the image to be stitched is used for stitching; when the temporal shift parameter of the current image is -1, the previous frame of the image to be stitched is used for stitching; when the temporal shift parameter of the current image is 0, the current image sequence is output to obtain the stitched image sequence, and the stitching result is output.

[0051] In step S44, the present invention determines the stitching strategy by judging the temporal shift parameter value of the current image. When the temporal shift parameter is 1, the algorithm selects the next frame of the image to be stitched for stitching operation, which is specifically implemented by the pixel fusion algorithm of the overlapping area of ​​the image. The fusion algorithm adopts the weighted average method, and the weight coefficient is calculated according to the distance from the pixel to the boundary of the overlapping area. When the temporal shift parameter is -1, the algorithm selects the previous frame of the image to be stitched for stitching operation, and the continuity of the image is achieved by the reverse pixel fusion processing. When the temporal shift parameter is 0, the algorithm directly outputs the current image sequence as the stitching result. In the overall processing, the geometric consistency of the image is guaranteed by the accurate calculation of the homography matrix, and the correctness of the stitching order is ensured by the calculation of the temporal shift parameter.

[0052] like Figure 2 As shown, the present invention also provides a multi-UAV image stitching system based on control information, comprising: Acquisition module 100: Used to acquire raw image data and flight information of multiple drones; Correction module 200: Used to correct the corresponding original image data based on the flight information of the UAV to obtain corrected image data; Extraction module 300: used to extract features from the corrected image data to obtain a set of matching point pairs; Stitching module 400: Used to construct an adaptive homography matrix based on the set of matching point pairs, and stitch multiple corrected image data based on the adaptive homography matrix to obtain a stitching result.

[0053] A third aspect of the present invention provides a multi-UAV image stitching device based on control information, comprising: A memory and at least one processor, wherein the memory stores instructions; At least one of the processors invokes the instructions in the memory to cause a multi-UAV image stitching device based on control information to perform a multi-UAV image stitching method based on control information as described in any of the preceding claims.

[0054] A fourth aspect of the present invention provides a computer-readable storage medium storing instructions that, when executed by a processor, implement a multi-UAV image stitching method based on control information as described in any of the preceding claims.

[0055] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0056] This invention presents a multi-UAV image stitching method based on control information, which significantly improves the accuracy and efficiency of image stitching. By incorporating UAV flight information for image correction, this invention effectively eliminates geometric distortions caused by attitude changes in traditional stitching methods, resulting in more accurate and reliable stitching results. The construction of an adaptive homography matrix fully utilizes the geometric constraints of multiple matching point pairs, enhancing the robustness of the transformation and reducing stitching errors caused by mismatched feature points. The computational mechanism of time-series shift parameters intelligently determines the stitching order of images, avoiding image overlap and stitching discontinuities in traditional methods, ensuring the continuity and consistency of multiple images. The overall method maintains good stitching performance even in complex flight environments, making it particularly suitable for aerial image processing over large geographical areas, providing high-quality panoramic image data for UAV remote sensing applications.

[0057] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for stitching multi-UAV images based on control information, characterized in that, include: S1: Collect raw image data and flight information of multiple drones; S2: Correct the corresponding raw image data using the drone's flight information to obtain corrected image data; S3: Perform feature extraction on the corrected image data to obtain a set of matching point pairs; S4: Construct an adaptive homography matrix based on the set of matching point pairs, and stitch together multiple corrected image data based on the adaptive homography matrix to obtain the stitching result.

2. The multi-UAV image stitching method based on control information according to claim 1, characterized in that, Step S1 further includes: S11: Collect flight information of the UAV through the flight control system to obtain a flight information set; S12: Acquire raw image data from multiple drones to obtain a raw image set; S13: Associate and store the flight information set with the original image set to obtain an information database.

3. The multi-UAV image stitching method based on control information according to claim 2, characterized in that, The flight information in step S11 includes: yaw angle, pitch angle, roll angle, and GPS positioning data; the raw image data in step S12 includes: image data and corresponding image acquisition timestamps.

4. The multi-UAV image stitching method based on control information according to claim 1, characterized in that, The expression for the corrected image data in step S2 is: ； in, The original image data before correction. For the heading angle, This is the corrected image data.

5. The multi-UAV image stitching method based on control information according to claim 1, characterized in that, Step S3 further includes: S31: Perform Gaussian smoothing on the corrected image data to obtain smoothed image data; S32: Select the pixel corresponding to the minimum gray value from the smoothed image data as the feature point; S33: Calculate the Euclidean distance between the feature points of the current image and all pixels in the neighboring images of the current image, and output the pixel corresponding to the minimum value of the Euclidean distance as the matching point; S34: Output the matching points of the current image and the matching points on the neighboring images of the current image as matching point pairs, and obtain a set of matching point pairs.

6. The multi-UAV image stitching method based on control information according to claim 1, characterized in that, Step S4 further includes: S41: Based on the coordinate values ​​of multiple matching point pairs in the set of matching point pairs, establish an adaptive homography matrix; S42: Perform homography transformation on multiple corrected image data using the adaptive homography matrix to obtain multiple transformed image data; S43: Calculate the temporal movement parameters of image sequences of multiple UAVs by transforming the image center point position of the image data; S44: Sequentially stitch together multiple transformed image data according to the time-shifting parameters to obtain a stitched image sequence and output the stitching result.

7. The multi-UAV image stitching method based on control information according to claim 6, characterized in that, The expression for the timing shift parameter in step S43 is: ； in, To transform the image data index value, For the first Temporal shift parameters of Zhang transform image data, For the first The x-coordinate of the center point of the transformed image data The width of the reference image; Based on the time-series shift parameters, step S44 specifically includes: In the image sequence to which the transformed image data belongs, when the temporal shift parameter of the current image is 1, the next frame of the image to be stitched is used for stitching. In the image sequence to which the transformed image data belongs, when the temporal shift parameter of the current image is -1, the previous frame of the image to be stitched is used for stitching. In the image sequence to which the transformed image data belongs, when the temporal shift parameter of the current image is 0, the current image sequence is output to obtain the stitched image sequence, and the stitching result is output.

8. A multi-UAV image stitching system based on control information, characterized in that, include: Acquisition module: Used to acquire raw image data and flight information of multiple drones; Correction module: Used to correct the corresponding raw image data based on the UAV's flight information to obtain corrected image data; Extraction module: used to extract features from the corrected image data to obtain a set of matching point pairs; The stitching module is used to construct an adaptive homography matrix based on the set of matching points, and stitch multiple corrected image data together based on the adaptive homography matrix to obtain the stitching result.

9. A multi-UAV image stitching device based on control information, characterized in that, include: A memory and at least one processor, wherein the memory stores instructions; At least one of the processors invokes the instructions in the memory to cause a multi-UAV image stitching device based on control information to perform a multi-UAV image stitching method based on control information as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores instructions that, when executed by a processor, implement a multi-UAV image stitching method based on control information as described in any one of claims 1 to 7.

Images