An unmanned aerial vehicle image and multi-modal map region clipping method and system based on posture information

Abstract

The application discloses a kind of unmanned aerial vehicle image and multi-modal map area cutting method and system based on attitude information, first obtain the unmanned aerial vehicle image containing EXIF metadata, extract GPS information and attitude parameter, calculate image ground real projection center point longitude and latitude in combination with sensor parameter graph;Again, according to pixel size, sensor physical size etc., calculate ground resolution and geographical coverage, correct the projection angle to get four real longitude and latitude, then four corner coordinates are converted to map projection coordinate system, generate cutting window in combination with base map geographic reference information, high-precision cutting is carried out in infrared, visible light map base map respectively, corresponding result is carried out after spatial alignment and consistency verification, and relevant data is saved in JSON format structure.The method realizes accurate area matching and cutting, improves the spatial accuracy and processing efficiency of multi-source remote sensing data fusion and target identification, and is suitable for remote sensing surveying and mapping, resource investigation and emergency monitoring and other scenes.

Description

Technical Field

[0001] This invention belongs to the field of image geographic information processing technology, and in particular relates to a method and system for cropping UAV images and multimodal map regions based on attitude information. Background Technology

[0002] With the development of UAV remote sensing technology, low-altitude high-resolution imagery has been widely used in fields such as forest and grassland inspection, urban planning, and emergency response. In order to improve the georeferenced accuracy and multimodal information matching capability of image data, it is usually necessary to accurately map the image areas captured by UAVs to a known map coordinate system and perform region-level cropping and fusion of multiple modal images, such as infrared and visible light.

[0003] However, due to factors such as tilted shooting, changes in heading, and uneven flight altitude during drone flight, the geographical areas covered by the images exhibit irregular projection shapes in the map coordinate system. Traditional methods relying on static cropping based on the image center point are insufficient to meet the accuracy requirements of practical applications. Furthermore, existing cropping methods often ignore the coupling relationship between image attitude parameters and sensor parameters, failing to accurately reconstruct the true coverage area of ​​the image. This results in spatial offset between the image and the map, affecting subsequent positioning and registration tasks.

[0004] Therefore, there is an urgent need for a cropping method that can comprehensively consider image pose information, sensor characteristics, and map coordinate system transformation relationships to achieve high-precision region matching and cropping between UAV images and multimodal map data, providing stable and reliable basic data support for subsequent geographic analysis and applications. Summary of the Invention

[0005] To address the above technical problems, this invention provides a method and system for cropping UAV images and multimodal map regions based on attitude information.

[0006] The technical solution adopted by this invention to solve its technical problem is:

[0007] A method for cropping regions from UAV images and multimodal maps based on attitude information, the method comprising the following steps:

[0008] S100: Acquire drone images with EXIF ​​metadata, read GPS information from the images, extract attitude parameters of the shooting location, and calculate the latitude and longitude coordinates of the true projection center point of the image on the ground in combination with sensor parameters;

[0009] S200: Based on the image's pixel size, the physical size of the image sensor, the camera's focal length, and the flight altitude, the ground resolution of the image is calculated, thereby determining the actual size range covered by the image in geographic space. At the same time, the geographic projection angle of the image is corrected by combining attitude information to obtain the true latitude and longitude coordinates of the four corner points of the image.

[0010] S300: Convert the latitude and longitude coordinates of the four corner points of the image according to the map projection coordinate system used, and combine the geographic reference information of the map base map to calculate the pixel coordinate boundary of the area covered by the image in the map image, and generate a clipping window in the map coordinate system.

[0011] S400: In the infrared map base map and the visible light map base map respectively, the corresponding area of ​​the image is cropped with high precision according to the generated cropping window to obtain the infrared image cropping result and the visible light image cropping result that are consistent with the position of the UAV image.

[0012] S500: Performs spatial alignment processing on the cropping results of infrared and visible light images and verifies the consistency between images. At the same time, it saves the original GPS information of the image, the coordinates of the true center point, the latitude and longitude coordinates of the four corner points of the cropped area, and the coordinate system transformation parameters in a structured format as a JSON file.

[0013] Preferably, S100 includes:

[0014] S110: Acquire drone images with EXIF ​​metadata, parse the images using an image processing library, extract GPS information fields including latitude and longitude, reference direction and flight altitude, and calculate the pseudo center point coordinates of the image based on latitude and longitude and reference direction;

[0015] S120: Extracts tilt angle, heading angle, image size, sensor parameters, and focal length from image metadata;

[0016] S130: Based on image size, sensor parameters, focal length, and flight altitude, combined with tilt angle, calculate the latitude and longitude coordinates of the true projection center point of the image on the ground.

[0017] Preferably, in S110, the coordinates of the pseudo-center point of the image are calculated based on latitude, longitude, and reference direction, specifically expressed by the formula:

[0018] ;

[0019] ;

[0020] In the formula, The latitude represents the coordinates of the pseudo-center point of the image. The longitude representing the coordinates of the pseudo-center point of the image. , , These represent the degrees, minutes, and seconds of latitude, respectively. , , These represent the degrees, minutes, and seconds of longitude, respectively. If the latitude reference direction is "S", then... If the longitude direction is "W", then .

[0021] Preferably, the latitude and longitude coordinates of the actual projection center point of the image on the ground in S130 are specifically expressed by the formula:

[0022] ;

[0023] ;

[0024] In the formula, Represents the latitude coordinates of the center point of the image's true projection onto the ground; Represents the longitude coordinates of the center point of the image's true projection onto the ground; Indicates the flight altitude at the time the image was taken; Indicates the camera's focal length; These represent the width and height of the camera sensor, respectively. These represent the image's pixel width and height, respectively. This represents the ground length corresponding to each degree of latitude. It represents the ground length corresponding to each degree of longitude.

[0025] Preferably, S200 includes:

[0026] S210: Based on the image's pixel size, sensor physical size, focal length, and flight altitude, combined with the tilt angle, back-project the camera's imaging plane to calculate the actual coverage width and height of the image on the ground;

[0027] S220: Based on the latitude and longitude of the true center point and the actual coverage width and height of the image on the ground, and further combined with the heading angle and tilt angle, the coordinate rotation matrix is ​​used to perform attitude correction on the image coverage area to obtain the latitude and longitude coordinates of the four corner points after rotation correction.

[0028] Preferably, S210 is specifically expressed by the formula:

[0029] ;

[0030] ;

[0031] In the formula, These represent the actual width and height of the image coverage on the ground, respectively. Indicates the flight altitude at the time the image was taken; Indicates the camera's focal length; These represent the width and height of the camera sensor, respectively. These represent the image's pixel width and height, respectively. Indicates the camera tilt angle.

[0032] Preferably, S220 is specifically expressed by the formula:

[0033] ;

[0034] ;

[0035] ;

[0036] ;

[0037] ;

[0038] In the formula, Indicates the heading angle. and These represent the image offsets in the latitude and longitude directions, respectively. A, B, C, and D represent the ground projection coordinates of the top left, top right, bottom left, and bottom right corners of the image, respectively. These represent the latitude and longitude of point A in the upper left corner of the image, respectively. These represent the latitude and longitude of point B in the upper right corner of the image, respectively. These represent the latitude and longitude of point C in the lower left corner of the image, respectively. These represent the latitude and longitude of point D in the lower right corner of the image, respectively.

[0039] Preferably, S300 is represented as:

[0040] ;

[0041] ;

[0042] in, This represents a geographic projection transformation function. The input is WGS84 coordinates, and the output is planar coordinates in the map projection coordinate system. Represents the inverse of the affine transformation matrix of a GeoTIFF image; The pixel coordinates representing the boundary; This indicates the cropped area of ​​the image within the map image; This represents a polygonal geometric region formed by connecting these four vertices in a clockwise or counterclockwise order.

[0043] Preferably, in S500, the cropping results of the infrared image and the visible light image are spatially aligned. A correspondence is established through corner coordinates or image features, and the infrared image is mapped to the visible light image coordinate system using affine transformation to achieve geographical consistency between the two images.

[0044] A UAV image and multimodal map region cropping system based on attitude information includes an image acquisition unit and a computer system connected thereto. The computer system is equipped with a processing unit and a multimodal map cropping model connected in sequence, wherein:

[0045] The image acquisition unit is used to acquire UAV images with pose information and map base maps with geographic information, and send the images to the computer system;

[0046] The processing unit is used to extract pose information from UAV images and geographic information from the base map, and then send them to the multimodal map cropping model.

[0047] The multimodal map cropping model receives pose information and geographic information, and uses a pose-based UAV image and multimodal map region cropping method to crop the multimodal map, resulting in cropped image pairs.

[0048] The aforementioned method and system for cropping UAV images and multimodal maps based on attitude information first acquires UAV images with EXIF ​​metadata and extracts their GPS position and attitude parameters. Combining sensor parameters and imaging size, the system calculates the true projection center coordinates and coverage area of ​​the image on the ground. Then, based on flight attitude information, the image projection angle is corrected, and the true latitude and longitude coordinates of the four corners of the image are obtained and transformed into the map projection coordinate system to generate a cropping window. In the multimodal base map, high-precision cropping of infrared and visible light images is performed according to this cropping window. Furthermore, spatial alignment processing is performed on the cropping results through corner matching or image features to ensure consistency of different modal images in geographic space. Finally, the original GPS information, projection center coordinates, corner coordinates, and coordinate transformation parameters are structured and saved as a JSON file. This method achieves accurate region matching and cropping between UAV images and multimodal maps, which helps improve the spatial accuracy and processing efficiency of multi-source remote sensing data fusion and target recognition. Attached Figure Description

[0049] Figure 1 This is a flowchart of a method for cropping UAV images and multimodal map regions based on attitude information in one embodiment of the present invention;

[0050] Figure 2This is a schematic diagram of the positioning and display of the ground real projection center on a multimodal map in one embodiment of the present invention, wherein (a) is the projection center display in the visible light map, and (b) is the projection center display in the infrared map;

[0051] Figure 3 This is a schematic diagram of the infrared and visible light clipping results before and after spatial alignment in one embodiment of the present invention. Detailed Implementation

[0052] To enable those skilled in the art to better understand the technical solution of the present invention, the present invention will be further described in detail below with reference to the accompanying drawings.

[0053] In one embodiment, such as Figure 1 As shown, a method for cropping UAV images and multimodal map regions based on attitude information includes the following steps:

[0054] S100: Acquire UAV images with EXIF ​​(Exchangeable Image File Format) metadata, read GPS (Global Positioning System) information from the images, extract attitude parameters of the shooting location, and calculate the latitude and longitude coordinates of the true projection center point of the image on the ground in combination with sensor parameters; among which, attitude parameters include pseudo-center latitude and longitude, heading angle, tilt angle, flight altitude, etc.

[0055] In one embodiment, S100 includes:

[0056] S110: Acquire drone images with EXIF ​​metadata, parse the images using an image processing library, extract GPS information fields including latitude and longitude, reference direction and flight altitude, and calculate the pseudo center point coordinates of the image based on latitude and longitude and reference direction;

[0057] S120: Extracts tilt angle, heading angle, image size, sensor parameters, and focal length from image metadata; among which, heading angle... Indicates the angle between the drone and true north; tilt angle Indicates the angle between the camera's optical axis and the vertical direction;

[0058] S130: Based on image size, sensor parameters, focal length, and flight altitude, combined with tilt angle, calculate the latitude and longitude coordinates of the true projection center point of the image on the ground.

[0059] In one embodiment, the pseudo-center point coordinates of the image are calculated in S110 based on latitude, longitude, and reference direction, specifically expressed by the formula:

[0060] ;

[0061] ;

[0062] In the formula, The latitude represents the coordinates of the pseudo-center point of the image. The longitude representing the coordinates of the pseudo-center point of the image. , , These represent the degrees, minutes, and seconds of latitude, respectively. , , These represent the degrees, minutes, and seconds of longitude, respectively. If the latitude reference direction is "S", then... If the longitude direction is "W", then .

[0063] In one embodiment, the latitude and longitude coordinates of the actual projection center point of the image on the ground in S130 are specifically expressed by the formula:

[0064] ;

[0065] ;

[0066] In the formula, Represents the latitude coordinates of the center point of the image's true projection onto the ground; Represents the longitude coordinates of the center point of the image's true projection onto the ground; Indicates the flight altitude at the time the image was taken; Indicates the camera's focal length; These represent the width and height of the camera sensor, respectively. These represent the image's pixel width and height, respectively. This represents the ground length corresponding to each degree of latitude. This represents the length of the Earth's surface corresponding to each degree of longitude. Furthermore, in this embodiment, the Earth model is approximated as a sphere. The empirical value is 111,320 meters. The calculation formula is: The value 111320 is calculated by dividing the Earth's equatorial circumference of 40075 km by 360°.

[0067] See Figure 2 , Figure 2 This is a schematic diagram illustrating the positioning and display of the ground's true projection center on a multimodal map according to an embodiment of the present invention, where (a) shows the projection center display in the visible light map, and (b) shows the projection center display in the infrared map. Based on the actual imaging model of the camera in the air, combined with the physical dimensions of the image sensor (including the sensor's width)... With height Image resolution (pixel width) With pixel height ), camera's optical focal length Camera altitude during drone flight and the tilt angle between the camera's optical axis and the vertical direction. Calculate the spatial offset of the center point of the image's true projection on the ground relative to the original GPS point.

[0068] S200: Based on the image's pixel size, the physical size of the image sensor, the camera's focal length, and the flight altitude, the ground resolution of the image is calculated, thereby determining the actual size range covered by the image in geographic space. At the same time, the geographic projection angle of the image is corrected by combining attitude information to obtain the true latitude and longitude coordinates of the four corner points of the image.

[0069] In one embodiment, S200 includes:

[0070] S210: Based on the image's pixel size, sensor physical size, focal length, and flight altitude, combined with the tilt angle, back-project the camera's imaging plane to calculate the actual coverage width and height of the image on the ground;

[0071] S220: Based on the latitude and longitude of the true center point and the actual coverage width and height of the image on the ground, and further combined with the heading angle and tilt angle, the coordinate rotation matrix is ​​used to perform attitude correction on the image coverage area to obtain the latitude and longitude coordinates of the four corner points after rotation correction.

[0072] In one embodiment, S210 is specifically expressed by the formula:

[0073] ;

[0074] ;

[0075] In the formula, These represent the actual width and height of the image coverage on the ground, respectively. Indicates the flight altitude at the time the image was taken; Indicates the camera's focal length; These represent the width and height of the camera sensor, respectively. These represent the image's pixel width and height, respectively. Indicates the camera tilt angle.

[0076] Specifically, this process is based on the principle of perspective projection. By back-projecting the camera's imaging size from the sensor plane onto the ground, the projected width and height of the image in geographic space are obtained. Considering the camera's tilt angle, to avoid estimation errors, the vertical component of the image boundary needs to be cosine corrected, thereby more accurately reconstructing the boundary of the image's projected area on the ground. This estimation provides the dimensional basis and scale constraints for subsequent geographic projection transformation and image cropping window generation.

[0077] In one embodiment, S220 is specifically expressed by the formula:

[0078] ;

[0079] ;

[0080] ;

[0081] ;

[0082] ;

[0083] In the formula, Indicates the heading angle. and These represent the image offsets in the latitude and longitude directions, respectively. A, B, C, and D represent the ground projection coordinates of the top left, top right, bottom left, and bottom right corners of the image, respectively. These represent the latitude and longitude of point A in the upper left corner of the image, respectively. These represent the latitude and longitude of point B in the upper right corner of the image, respectively. These represent the latitude and longitude of point C in the lower left corner of the image, respectively. These represent the latitude and longitude of point D in the lower right corner of the image, respectively.

[0084] Specifically, the obtained image contains the latitude and longitude of the true ground center point. This is used as the reference position of the image in the geographic coordinate system, and further combined with the actual coverage width of the image. With height Based on the combination of the center point and the offset, the latitude and longitude coordinates of the ground projection at the four corners of the image are calculated.

[0085] S300: Convert the latitude and longitude coordinates of the four corner points of the image according to the map projection coordinate system used, and combine the geographic reference information of the map base map to calculate the pixel coordinate boundary of the area covered by the image in the map image, and generate a clipping window in the map coordinate system.

[0086] Specifically, because infrared maps and visible light maps use different geographic coordinate reference systems, their spatial projection methods differ. Therefore, before performing the cropping operation, it is necessary to read the coordinate system information of each map separately and construct a corresponding coordinate converter based on the map projection system used (such as WGS84, CGCS2000, Web Mercator, etc.). By accurately converting the geographic latitude and longitude coordinates of the image corner points into the projected coordinates of the corresponding map image, and then combining this with the affine transformation parameters of the map image, the pixel boundaries of the cropping window can be precisely located.

[0087] In one embodiment, S300 is represented as:

[0088] ;

[0089] ;

[0090] in, This represents a geographic projection transformation function. The input is WGS84 coordinates, and the output is planar coordinates in the map projection coordinate system. Represents the inverse of the affine transformation matrix of a GeoTIFF image; The pixel coordinates representing the boundary; This indicates the cropped area of ​​the image within the map image; This represents a polygonal geometric region formed by connecting these four vertices in a clockwise or counterclockwise order.

[0091] S400: Based on the generated cropping window, the corresponding area of ​​the image is cropped with high precision in both the infrared and visible light base maps to obtain infrared and visible light image cropping results that are consistent with the position of the UAV image.

[0092] Specifically, based on the geographic coordinates of the four corner points of the image determined in S300, and combined with the projection coordinate system information of the map base map, corresponding polygonal clipping windows can be constructed on both the infrared image base map and the visible light image base map. These clipping windows are consistent with the projection area of ​​the UAV image in geographic space, accurately corresponding to the actual geographic coverage of the area captured by the UAV.

[0093] In one embodiment, S500 includes: performing spatial alignment processing on the cropping results of the infrared image and the visible light image, verifying the consistency between the images, and simultaneously saving the original GPS information of the image, the coordinates of the true center point, the latitude and longitude coordinates of the four corner points of the cropping area, and the coordinate system transformation parameters in a structured format as a JSON file.

[0094] Specifically, see Figure 3 , Figure 3This is a schematic diagram of the infrared and visible light cropping results before and after spatial alignment in one embodiment of the present invention. In S500, spatial alignment processing is performed on the multimodal cropping results, that is, spatial alignment processing is performed on the cropped infrared image and the visible light image. A correspondence is established through corner coordinates or image features, and affine transformation is used to map the infrared image to the visible light image coordinate system to achieve geographical consistency between the two images.

[0095] The aforementioned method and system for cropping UAV images and multimodal maps based on attitude information extracts spatial information from UAV images, including pseudo-center latitude and longitude, heading angle, tilt angle, and flight altitude, and calculates the true projection center position of the image on the ground by combining sensor parameters. Subsequently, based on image resolution, sensor size, and ground resolution, the actual size of the image coverage area is determined, and the geographic projection distortion of the image is corrected by combining attitude information to obtain the true latitude and longitude coordinates of the four corner points of the image. Through map projection coordinate system transformation and base map geographic reference data, the pixel position of the image on the map is calculated, and region cropping is performed on infrared and visible light images to obtain accurately aligned multimodal image cropping results. The cropping results and related geographic information are stored in a structured format, providing a data foundation for subsequent geographic registration, multi-source fusion, and positioning analysis. This invention offers high accuracy and strong adaptability, suitable for various scenarios such as remote sensing mapping, resource surveys, and emergency monitoring.

[0096] In one embodiment, a UAV image and multimodal map region cropping system based on attitude information is also provided, including an image acquisition unit and a computer system connected thereto. The computer system is equipped with a processing unit and a multimodal map cropping model connected in sequence, wherein:

[0097] The image acquisition unit is used to acquire UAV images with pose information and map base maps with geographic information, and send the images to the computer system;

[0098] The processing unit is used to extract pose information from UAV images and geographic information from the base map, and then send them to the multimodal map cropping model.

[0099] The multimodal map cropping model receives pose information and geographic information, and uses a pose-based UAV image and multimodal map region cropping method to crop the multimodal map, resulting in cropped image pairs.

[0100] Specific limitations regarding the attitude-information-based UAV image and multimodal map region cropping system can be found in the above description of the UAV image and multimodal map region cropping method, and will not be repeated here. Each module in the aforementioned attitude-information-based UAV image and multimodal map region cropping system can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the memory of a computer device as software, so that the processor can call and execute the corresponding operations of each module.

[0101] The above provides a detailed description of a method and system for cropping UAV images and multimodal maps provided by this invention. Specific examples have been used to illustrate the principles and implementation methods of this invention. The descriptions of the embodiments above are merely for the purpose of helping to understand the core ideas of this invention. It should be noted that those skilled in the art can make various improvements and modifications to this invention without departing from its principles, and these improvements and modifications also fall within the protection scope of the claims of this invention.

Claims

1. A method for cropping regions from UAV images and multimodal maps based on attitude information, characterized in that, The method includes the following steps: S100: Acquire drone images with EXIF ​​metadata, read GPS information from the images, extract attitude parameters of the shooting location, and calculate the latitude and longitude coordinates of the center point of the image's true projection on the ground by combining sensor parameters; S200: Based on the image's pixel size, the physical size of the image sensor, the camera's focal length, and the flight altitude, the ground resolution of the image is calculated, thereby determining the actual size range covered by the image in geographic space. Simultaneously, the geographic projection angle of the image is corrected using attitude information to obtain the true latitude and longitude coordinates of the four corner points of the image. S200 includes: S210: Based on the image's pixel size, sensor physical size, focal length, and flight altitude, combined with the tilt angle, back-project the camera's imaging plane to calculate the actual coverage width and height of the image on the ground; S220: Based on the latitude and longitude of the true center point and the actual coverage width and height of the image on the ground, and further combined with the heading angle and tilt angle, the coordinate rotation matrix is ​​used to correct the attitude of the image coverage area to obtain the latitude and longitude coordinates of the four corner points after rotation correction. S300: Convert the latitude and longitude coordinates of the four corner points of the image according to the map projection coordinate system used, and combine the geographic reference information of the map base map to calculate the pixel coordinate boundary of the area covered by the image in the map image, and generate a clipping window in the map coordinate system. S400: In the infrared map base map and the visible light map base map respectively, the corresponding area of ​​the image is cropped with high precision according to the generated cropping window to obtain the infrared image cropping result and the visible light image cropping result that are consistent with the position of the UAV image. S500: Spatial alignment processing is performed on the cropping results of infrared and visible light images, and the consistency between the images is verified. At the same time, the original GPS information of the image, the coordinates of the true center point, the latitude and longitude coordinates of the four corner points of the cropped area, and the coordinate system transformation parameters are saved in a structured format as a JSON file. In S500, spatial alignment processing is performed on the cropping results of infrared and visible light images. The correspondence is established through corner point coordinates or image features, and affine transformation is used to map the infrared image to the coordinate system of the visible light image to achieve geographical consistency between the two images.

2. The method according to claim 1, characterized in that, S100 includes: S110: Acquire drone images with EXIF ​​metadata, parse the images using an image processing library, extract GPS information fields including latitude and longitude, reference direction and flight altitude, and calculate the pseudo center point coordinates of the image based on latitude and longitude and reference direction; S120: Extracts tilt angle, heading angle, image size, sensor parameters, and focal length from image metadata; S130: Based on image size, sensor parameters, focal length, and flight altitude, combined with tilt angle, calculate the latitude and longitude coordinates of the true projection center point of the image on the ground.

3. The method according to claim 2, characterized in that, The pseudo-center point coordinates of the image are calculated based on latitude, longitude, and reference direction in S110, and are expressed by the following formula: In the formula, The latitude represents the coordinates of the pseudo-center point of the image. The longitude representing the coordinates of the pseudo-center point of the image. These represent the degrees, minutes, and seconds of latitude, respectively. These represent the degrees, minutes, and seconds of longitude, respectively. If the latitude reference direction is "S", then... If the longitude direction is "W", then .

4. The method according to claim 3, characterized in that, The latitude and longitude coordinates of the true projection center point of the image onto the ground in S130 are expressed by the following formula: In the formula, Represents the latitude coordinates of the center point of the image's true projection onto the ground; The longitude coordinates of the center point of the image's true projection onto the ground; The flight altitude at the time the image was taken; Camera focal length; The width and height of the camera sensor; Image pixel width and height; This represents the ground length corresponding to each degree of latitude. It represents the ground length corresponding to each degree of longitude.

5. The method according to claim 4, characterized in that, S210 is specifically expressed by the formula: In the formula, , The flight altitude at the time the image was taken; Camera focal length; The width and height of the camera sensor; Image pixel width and height, Indicates the camera tilt angle.

6. The method according to claim 5, characterized in that, S220 is specifically expressed by the formula: In the formula, Indicates the heading angle. and These represent the image offsets in the latitude and longitude directions, respectively. A, B, C, and D represent the ground projection coordinates of the top left, top right, bottom left, and bottom right corners of the image, respectively. These represent the latitude and longitude of point A in the upper left corner of the image, respectively. These represent the latitude and longitude of point B in the upper right corner of the image, respectively. These represent the latitude and longitude of point C in the lower left corner of the image, respectively. These represent the latitude and longitude of point D in the lower right corner of the image, respectively.

7. The method according to claim 6, characterized in that, S300 is represented as: in, This represents a geographic projection transformation function. The input is WGS84 coordinates, and the output is planar coordinates in the map projection coordinate system. Represents the inverse of the affine transformation matrix of a GeoTIFF image; The pixel coordinates representing the boundary; This indicates the cropped area of ​​the image within the map image; This represents a polygonal geometric region formed by connecting these four vertices in a clockwise or counterclockwise order.

8. A UAV image and multimodal map region cropping system based on attitude information, characterized in that, It includes an image acquisition unit and a computer system connected to it. The computer system is equipped with a processing unit and a multimodal map cropping model connected in sequence, wherein: The image acquisition unit is used to acquire UAV images with pose information and map base maps with geographic information, and send the images to the computer system; The processing unit is used to extract pose information from UAV images and geographic information from the base map, and then send them to the multimodal map cropping model. The multimodal map cropping model receives pose information and geographic information and performs multimodal map cropping using the method described in any one of claims 1 to 7 to obtain cropped image pairs.

Images