Method and system for automatic processing of accessory results for multi-source remote sensing images

Abstract

This invention relates to the field of image processing technology, specifically to an automated processing method and system for ancillary results of multi-source remote sensing imagery. The method includes: parsing and extracting core parameters such as side viewpoint, temporal phase, and data source type for each remote sensing image from the original metadata file set, generating a set of structured metadata objects; generating a set of standard scene numbers based on the data source type, track number, and image acquisition time information in the structured metadata object set; and batch-producing the full-scene image metadata files, projection information files, and coordinate information files for each remote sensing image based on the completed full-scene image set, standard scene number set, and structured metadata object set, thus obtaining an initial set of ancillary results. This invention, by parsing and extracting core parameters and generating a set of structured metadata objects, provides a unified, accurate, and machine-readable data source for all subsequent automated steps, eliminating production errors caused by misunderstandings of metadata at the source.

Description

Technical Field

[0001] This invention relates to the field of image processing technology, specifically to an automated processing method and system for ancillary results of multi-source remote sensing images. Background Technology

[0002] Currently, traditional methods rely on manual parsing of unstructured descriptive information in raw metadata files. Due to significant differences in metadata formats and description methods across different data sources, manual parsing is prone to misunderstandings, leading to inaccurate extraction of key parameters from the metadata and subsequent production errors. Furthermore, traditional methods do not automatically generate unique standard scene numbers that are strongly associated with images according to preset rules. Images from different data sources are often identified according to their own naming conventions, causing naming confusion. This makes it difficult to quickly and accurately identify and locate specific images in cross-source image management, increasing the difficulty of batch management and cross-retrieval, and reducing work efficiency.

[0003] Furthermore, for coordinate information files, different calculation methods or sources may lead to inconsistencies in the number of decimal places. Traditional methods lack a unified decimal place processing mechanism, making it impossible to guarantee the consistency of coordinate data format and precision. This can cause compatibility issues or ambiguities in subsequent data processing and applications, affecting the accurate registration and analysis of images. In addition, when processing mosaicked shingled images, traditional methods struggle to perform spatial analysis and attribute assignment based on key information in the mosaic surface vector file and metadata, such as side viewpoint, temporal phase, and data source, to generate shingled block vector files in batches. This prevents each shingled block from effectively inheriting the key attributes of the original image, failing to meet the needs of shingled image management, statistics, and querying, and reducing the practicality and manageability of shingled images. Summary of the Invention

[0004] To achieve the above objectives, the present invention provides the following technical solution: an automated processing method for attachment results of multi-source remote sensing imagery, comprising: Obtain a set of multi-source remote sensing images and its corresponding set of raw metadata files; the set of raw metadata files includes unstructured descriptive information of key parameters of side view, temporal phase and data source type of each remote sensing image; parse and extract the core parameters of side view, temporal phase and data source type of each remote sensing image from the set of raw metadata files to generate a set of structured metadata objects; Based on the data source type, track number, and image acquisition time information in the structured metadata object set, a standard scene number set is generated; based on the completed whole scene image set, standard scene number set, and structured metadata object set, whole scene image metadata files, projection information files, and coordinate information files for each remote sensing image are produced in batches to obtain the initial set of attachment results; The coordinate information files in the initial set of attached results are processed to unify the decimal places, generating a set of standardized coordinate information files; the standardized coordinate information files and remote sensing images are quality checked to obtain a complete set of whole-scene image attached results; the files in the complete set of whole-scene image attached results are automatically organized into the corresponding whole-scene image results folders according to the standard scene number and image type classification rules, generating a structured results directory; Based on the mosaic image set in the multi-source remote sensing image set, as well as the mosaic surface vector file and the structured metadata object set, the mosaic block vector file is generated in batches through spatial analysis and spatial assignment, and the projection information file is generated in batches according to the projection information of the image set, forming a set of image attachment results. The files in the set of attached images are automatically organized into the corresponding image result folders according to the image naming rules, generating a structured result directory. The structured result directory is then checked for completeness. If any files are missing, they are added or generated until the structured result directory is complete and error-free. The final structured result directory is then used as the final automated result of the multi-source remote sensing image processing.

[0005] Preferably, each object in the structured metadata object set corresponds to a core parameter of a remote sensing image; Based on the data source type, track number, and image acquisition time information in the structured metadata object set, a standard scene number set is generated, including: Extract data source type, track number, and image acquisition time information from a collection of structured metadata objects; According to the preset naming rules, the extracted information is concatenated into a standard scene number; the naming rules include the combination of data source identifier, track number, image acquisition time and band identifier; By associating standard scene numbers with their corresponding remote sensing images, a set of standard scene numbers is obtained.

[0006] Preferably, based on the completed set of whole-view images, standard view number set, and structured metadata object set, batch-produce whole-view image metadata files, projection information files, and coordinate information files for each remote sensing image to obtain an initial set of attachment results, including: Name the whole-scene image metadata file for each remote sensing image based on the standard scene number in the standard scene number set; Extract projection information from a collection of structured metadata objects or a collection of whole-scene images to generate a projection information file; Based on the geographic extent and resolution of the remote sensing image, calculate and generate coordinate information files; By associating the whole-view image metadata file, projection information file, and coordinate information file with the corresponding whole-view remote sensing image, an initial set of attachment results is obtained.

[0007] Preferably, the coordinate information files in the initial set of attachments are processed to standardize the decimal places, generating a standardized set of coordinate information files, including: Iterate through the coordinate information files in the initial set of attachment results; The number of decimal places in each coordinate information file is counted, and rounding is performed according to the preset decimal place rules; The processed coordinate information files are updated into the initial set of attachment results to generate a standardized set of coordinate information files.

[0008] Preferably, based on the mosaic image set in the multi-source remote sensing image set, as well as the mosaic surface vector file and the structured metadata object set, mosaic block vector files are generated in batches through spatial analysis and spatial assignment. Furthermore, projection information files are generated in batches based on the projection information of the image set, forming a set of image set ancillary results, including: Determine whether mosaicked imagery exists in a multi-source remote sensing image set; If mosaicked imagery exists, load the mosaic surface vector file; Based on the side view, temporal and data source type information in the structured metadata object set, spatial analysis and attribute assignment are performed on the mosaic surface vector file, and mosaic block vector files are generated in batches according to the tiling module; Add the mosaic block vector files to the complete set of attachments to complete the update.

[0009] Preferably, based on the completed set of whole-view images, standard view number set, and structured metadata object set, batch production of whole-view image metadata files, projection information files, and coordinate information files for each remote sensing image is performed to obtain an initial set of attachment results, which also includes: Projection information is extracted from the mosaic image set to generate a projection information file, and then decimal place unification processing is performed. Based on the geographic extent and resolution of the remote sensing image, calculate and generate coordinate information files; By associating the mosaic tiled image projection information file, coordinate information file, and corresponding mosaic tiled remote sensing image, an initial set of attachment results is obtained.

[0010] Preferably, the files in the image frame attachment set are automatically organized into the corresponding image frame result folders according to the image frame naming rules, generating a structured result directory, including: Load the preset directory structure specification; the directory structure specification includes the naming rules and hierarchical relationship of folders; According to the cataloging requirements for the whole-scene image results and the mosaic-frame image results, the files in the complete set of attachment results are classified into the corresponding folders; Following the directory structure specifications, folders are automatically created and files are moved to generate a structured output directory.

[0011] Preferably, the structured output catalog is checked for completeness. If any files are missing, they are generated and supplemented until the structured output catalog is complete and error-free. The final structured output catalog is then used as the final automated processing result of the multi-source remote sensing imagery, including: Traverse the structured output directory and check if the files in each folder are complete; If a missing file is found, it will be regenerated according to its type. Repeat the integrity verification process until the structured output catalog is complete and error-free. The final structured output catalog is used as an appendix to the multi-source remote sensing imagery in the automated processing results.

[0012] Preferably, the method further includes: The vector annotation results of manual geographic location accuracy and quality checks are batch-produced into quality check sheets and planar accuracy check statistical tables according to national or local standards, resulting in a set of inspection record results. Accordingly, the vector annotation results of manual geographic location accuracy and quality checks are batch-produced into quality check sheets and planar accuracy check statistical tables according to national or local standards, resulting in a set of inspection record results, including: Load the quality inspection vector file; the quality inspection vector file includes quality problem notes for remote sensing images, and the accuracy check vector includes image coordinates and measured coordinates; Based on the preset sizing modules and quality inspection rules, the quality problem notes are associated according to the sizing modules, and the number and distribution of various quality problems are statistically analyzed. According to the accuracy check rules, the standard deviation and limit deviation are preset according to different resolutions, the residuals are calculated according to the image coordinates and the measured coordinates, and the mean error and score are statistically calculated. Fill the quality inspection results and accuracy inspection results into the quality inspection form and the plane accuracy inspection statistics form according to the sectional module, respectively; Add the quality checklist and accuracy checklist to the standardized coordinate information file set to obtain a complete set of attachment results.

[0013] An automated processing system for attachment results of multi-source remote sensing imagery, applicable to the aforementioned automated processing method for attachment results of multi-source remote sensing imagery, including: The data acquisition unit is used to acquire a set of multi-source remote sensing images and its corresponding set of raw metadata files. The set of raw metadata files includes unstructured descriptive information of key parameters such as side view, temporal phase, and data source type of each remote sensing image. The unit parses and extracts the core parameters of side view, temporal phase, and data source type of each remote sensing image from the set of raw metadata files to generate a set of structured metadata objects. The file production unit is used to generate a standard scene number set based on the data source type, track number and image acquisition time information in the structured metadata object set; based on the completed whole scene image set, standard scene number set and structured metadata object set, it batch produces whole scene image metadata files, projection information files and coordinate information files for each remote sensing image, and obtains the initial attachment result set. The quality inspection unit is used to standardize the decimal places of the coordinate information files in the initial set of attached results, generating a set of standardized coordinate information files; to perform quality inspection on the set of standardized coordinate information files and remote sensing images, resulting in a complete set of attached results for whole-scene images; and to automatically organize the files in the complete set of attached results for whole-scene images into the corresponding whole-scene image result folders according to the standard scene number and image type classification rules, generating a structured result catalog. The slab image unit is used to generate slab image sets in a multi-source remote sensing image set, as well as slab surface vector files and structured metadata object sets, in batches through spatial analysis and spatial assignment, and to generate slab block vector files in batches according to the projection information of the slab image set, forming a set of slab image attachment results. The results verification unit is used to automatically organize the files in the set of attached results of segmented images into the corresponding segmented image results folders according to the segmented image naming rules, and generate a structured results directory; it performs a completeness verification on the structured results directory, and if there are missing files, it supplements and generates the missing files until the structured results directory is complete and error-free, and uses the final structured results directory as the final automated result of the multi-source remote sensing image.

[0014] Compared with the prior art, the beneficial effects of the present invention are: This invention provides a unified, accurate, and machine-readable data source for all subsequent automated steps by parsing and extracting core parameters and generating a set of structured metadata objects, thus eliminating production errors caused by misunderstandings of metadata from the source. Furthermore, by using structured parameters, it automatically generates unique standard scene numbers according to preset rules and strongly associates them with images, achieving standardized and unified identification of cross-source images, eliminating confusion caused by different naming habits, and providing a key for batch management and cross-retrieval. This invention generates metadata files, projection information files, and coordinate information files in batches based on standard scene numbers and structured parameters. The metadata files are automatically named according to the standard scene numbers without manual intervention and are automatically extracted from images or metadata to ensure strict matching with images. Through decimal place unification processing, the problem of inconsistent coordinate decimal places caused by different calculation methods or sources is eliminated, ensuring the format standardization and accuracy of coordinate data and avoiding compatibility problems or ambiguities caused by accuracy differences. This invention performs spatial analysis and attribute assignment based on the mosaic surface vector file and information such as side view, temporal phase, and data source in the metadata, and generates a batch of tiling block vector files. This ensures that each tiling block inherits the key attributes of the original image, meeting the needs of tiling management, statistics, and querying. Simultaneously, corresponding projection information files are produced in batches to maintain the integrity of the tiling results. This invention automatically categorizes and assigns massive amounts of files to corresponding result folders according to a preset directory structure, achieving standardized and regulated archiving of result packages. Anyone can quickly locate files using a fixed path, greatly improving the usability and professionalism of the deliverables. Attached Figure Description

[0015] Figure 1 This is a schematic flowchart of the overall method in one embodiment of the present invention; Figure 2 This is a schematic diagram of the overall system architecture in one embodiment of the present invention.

[0016] In the diagram: 1. Data acquisition unit; 2. Document production unit; 3. Quality inspection unit; 4. Segmented image unit; 5. Result verification unit. Detailed Implementation

[0017] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0018] Example 1, please refer to Figure 1 This invention provides a technical solution: an automated processing method for supplementary results of multi-source remote sensing imagery, comprising: S1. Obtain a set of multi-source remote sensing images and its corresponding set of raw metadata files; wherein, the set of raw metadata files includes unstructured descriptive information of key parameters of side view, temporal phase and data source type of each remote sensing image; parse and extract the core parameters of side view, temporal phase and data source type of each remote sensing image from the set of raw metadata files to generate a set of structured metadata objects. S2. Generate a set of standard scene numbers based on the data source type, track number, and image acquisition time information in the set of structured metadata objects; based on the completed set of whole scene images, the set of standard scene numbers, and the set of structured metadata objects, batch produce the whole scene image metadata files, projection information files, and coordinate information files for each remote sensing image to obtain the initial set of attachment results. S3. Perform decimal unification processing on the coordinate information files in the initial set of attachment results to generate a standardized set of coordinate information files; perform quality checks on the standardized set of coordinate information files and remote sensing images to obtain a complete set of whole-scene image attachment results; automatically organize the files in the complete set of whole-scene image attachment results into the corresponding whole-scene image result folders according to the standard scene number and image type classification rules to generate a structured result directory. S4. Based on the mosaic image set in the multi-source remote sensing image set, as well as the mosaic surface vector file and the structured metadata object set, generate mosaic block vector files in batches through spatial analysis and spatial assignment, and generate projection information files in batches according to the projection information of the image set, forming a set of image attachment results. S5. Automatically organize the files in the set of attached images of multi-frame images into the corresponding multi-frame image result folders according to the naming rules of multi-frame images, and generate a structured result directory; perform integrity verification on the structured result directory, and if there are missing files, supplement and generate the missing files until the structured result directory is complete and error-free, and use the final structured result directory as the final result of automated processing of multi-source remote sensing images.

[0019] It should be noted that a multi-source remote sensing image set refers to remote sensing image data from different data sources (such as different satellites, different sensors, etc.); the raw metadata file set corresponds to these remote sensing images and contains unstructured descriptive information of some key parameters of each remote sensing image, such as side view (the angle between the satellite and the vertical direction when it takes the picture), phase (the time when the image was acquired), data source type (such as which satellite or sensor it comes from), etc.; this information may be recorded arbitrarily in text form in the raw metadata file without a fixed format and structure; From the original metadata file set, core parameters such as side view, time phase, and data source type of each remote sensing image are extracted using specific methods; then, these extracted parameters are organized according to a certain structure to form a set of structured metadata objects; in this way, each structured metadata object corresponds to a remote sensing image and clearly records the key parameters of that image. Regarding the generation of a collection of structured metadata objects, the specific details include: Construct an externally configurable XML parsing configuration file. The XML parsing configuration file should at least include the data source type identifier, node level, side view node names, unit conversion coefficients, and standard field mapping relationships. The data source type identifier is used to uniquely distinguish different satellites or sensors, such as mainstream domestic satellites JL1 series, SV1, SVN3, GF1, GF2, ZY1E, zy303, TRIPLESAT, CB04A, etc. Data source type identifier, data source node, side view node, time phase node, node layer number, secondary node name, angle unit, time format; JL1*,SatelliteID,90-SatelliteElevation,CenterTime,3,ProductInfo,degree,YYYY-MM-DD HH:MM:SS; DP*,SatelliteID,90-SatelliteElevation,CenterTime,3,ProductInfo,degree,YYYY-MM-DD HH:MM:SS; SVN3,SatelliteID,90-SatelliteElevation,CenterTime,3,ProductInfo,degree,YYYY-MM-DD HH:MM:SS; GF1*,SatelliteID,SatelliteZenith,CenterTime,2,\,degree, YYYY-MM-DD HH:MM:SS; GF2,SatelliteID,SatelliteZenith,CenterTime,2,\,degree, YYYY-MM-DD HH:MM:SS; GF6,SatelliteID,SatelliteZenith,CenterTime,2,\,degree, YYYY-MM-DD HH:MM:SS; SV1,SatelliteID,SatelliteZenith,CenterTime,2,\,degree, YYYY-MM-DD HH:MM:SS; CB04A,SatelliteID,SatelliteZenith,CenterTime,2,\ ,degree, YYYY-MM-DD HH:MM:SS; ZY1E,SatelliteID,SatelliteZenith,CenterTime,2,\ ,degree, YYYY-MM-DD HH:MM:SS; ZY1F,SatelliteID,SatelliteZenith,CenterTime,2,\ ,degree, YYYY-MM-DD HH:MM:SS; TRIPLESAT,Satellite_Name,Sat_Elevation,Begin_Time,3,MetaData,degree,YYYY-MM-DD HH:MM:SS.SSSSSS; zy303,SatelliteID,90-SatAltitude,AcquisitionTime,3,ProductInfo ,degree,YYYY-MM-DD HH:MM:SS; During parsing, the original metadata file set is first traversed, and the data source type corresponding to each metadata file is automatically identified based on the filename, file header information, or preset rules. Then, based on the identified data source type, the corresponding parsing rules are loaded from the XML parsing configuration file. If the metadata file is in XML format (such as the MSS.xml metadata file of GF1), then the XML parser extracts the original values ​​of fields such as side view, data source type, and time phase based on the predefined node level and node name in the XML parsing configuration file. For example, for GF1, the XML parsing configuration file can define the node of the side view as level 2, the node name as SatelliteZenith, the node name of the time phase as CenterTime, and the node name of the data source as SatelliteID. If the metadata file is TXT or other unstructured text, then regular expression matching is used to extract it according to the predefined matching pattern in the mapping table; The extracted raw values ​​are standardized according to the unit conversion coefficients defined in the mapping table. For example, the side view is uniformly converted to degrees, and the time phase is uniformly converted to YYYY-MM-DD HH:MM:SS format. The extracted and transformed core parameters, such as side view, time phase, data source type, track number, and image acquisition time, are encapsulated according to a preset data structure to generate a key-value structured metadata object; the structured metadata objects of all images together constitute a collection of structured metadata objects; Based on the data source type, orbit number (the satellite's identification number when it is in orbit) and image acquisition time information in the structured metadata object collection, standard scene numbers are generated according to certain rules. Based on the already produced set of whole-scene images (complete remote sensing image scenes), set of standard scene numbers, and set of structured metadata objects, specific tools or methods are used to batch produce metadata files (recording detailed attribute information of the images), projection information files (indicating the projection method used by the images, such as Mercator projection), and coordinate information files (recording the coordinate positions of each point on the images). These files are combined to form the initial set of attachment results. The coordinate information files in the initial set of attachments are processed to standardize the decimal places. This is to make the coordinate information more standardized and uniform, and to avoid errors or inconsistencies caused by different decimal places. After processing, a set of standardized coordinate information files is generated. Then, the standardized coordinate information file set and remote sensing images are subjected to quality checks to check whether the images are clear and whether the files are complete and undamaged. After the checks, a complete set of full-scene image attachments is obtained. The files in the complete panoramic image attachment set are automatically organized into the corresponding panoramic image result folders according to the standard view number and image type classification rules. For example, different types of files belonging to the same standard view number are placed in one folder, which generates a structured result directory, making it convenient to find and manage the results. Based on the mosaic image set, mosaic surface vector file, and structured metadata object set in the multi-source remote sensing image set, mosaic block vector files are generated in batches through spatial analysis (such as determining the relationship between image location and mosaic surface vector) and spatial assignment (assigning corresponding attribute information to the image); then projection information files are created in batches according to the projection information of the image set. These files are combined to form the image set attachment results set. The files in the set of attached images are automatically organized into the corresponding image result folders according to the image naming rules, generating a structured result directory. Then, the structured result directory is checked for completeness to see if any files are missing. If any missing files are found, they are added and generated until the structured result directory is complete and error-free. This final structured result directory is the final automated processing result of the multi-source remote sensing image.

[0020] In one alternative embodiment, each object in the structured metadata object set corresponds to a core parameter of a remote sensing image; Based on the data source type, track number, and image acquisition time information in the structured metadata object set, a standard scene number set is generated, including: Extract data source type, track number, and image acquisition time information from a collection of structured metadata objects; According to the preset naming rules, the extracted information is concatenated into a standard scene number; the naming rules include the combination of data source identifier, track number, image acquisition time and band identifier; By associating standard scene numbers with their corresponding remote sensing images, a set of standard scene numbers is obtained.

[0021] It should be noted that the data source type indicates which satellite or sensor the image comes from; the orbit number is the satellite's number when it is in orbit, and the satellite's orbital position is different at different times; the image acquisition time is the specific date and time when the satellite took the image; by searching and filtering, this key information is extracted from the structured metadata object; The preset naming rules stipulate that the data source identifier, track number, image acquisition time, and band identifier should be combined in a certain order and manner; like building blocks, the extracted information is pieced together according to the rules to form a unique standard scene number, which facilitates subsequent identification and management. The generated standard view number is paired with the corresponding remote sensing image, just like labeling each image with a standard view number, establishing a one-to-one correspondence between them; in this way, each image has its own standard view number; collecting the standard view numbers corresponding to all images forms a standard view number set, which can help to quickly find each image and understand its basic information.

[0022] In an optional embodiment, based on the completed set of whole-view images, the set of standard view numbers, and the set of structured metadata objects, whole-view image metadata files, projection information files, and coordinate information files for each remote sensing image are produced in batches to obtain an initial set of attachment results, including: Name the whole-scene image metadata file for each remote sensing image based on the standard scene number in the standard scene number set; Extract projection information from a collection of structured metadata objects or a collection of whole-scene images to generate a projection information file; Based on the geographic extent and resolution of the remote sensing image, calculate and generate coordinate information files; By associating the whole-view image metadata file, projection information file, and coordinate information file with the corresponding whole-view remote sensing image, an initial set of attachment results is obtained.

[0023] It should be noted that the whole-view image metadata file is used to record in detail the various attributes of each remote sensing image, such as the image capture time, sensor type, data format, etc.; according to each standard view number in the standard view number set, the metadata file of the corresponding remote sensing image is given the same name as the standard view number. The purpose of this is to establish a clear correspondence between the metadata file and the remote sensing image, so that the corresponding metadata file can be quickly found through the standard view number, which facilitates the subsequent management and query of image information. Projection information is key information describing how remote sensing imagery is displayed on a map. It determines the correspondence between points on the imagery and their actual geographic coordinates. The structured metadata object collection stores various metadata for each remote sensing image, including projection information. The images themselves in the whole-view imagery collection may also carry some projection-related information. The projection information is extracted from these two sources and then organized into a dedicated projection information file. Geographic extent refers to the actual geographic area covered by the remote sensing image; resolution indicates the actual ground distance represented by each pixel on the image; using these two pieces of information, geographic extent and resolution, the actual geographic coordinates corresponding to each pixel on the image can be calculated; organizing these calculated coordinate information into a coordinate information file is very important for subsequent geographic analysis and processing of the image; Previously, metadata files, projection information files, and coordinate information files for the remote sensing images have been generated. Now, these files need to be paired with the corresponding remote sensing images; it's like equipping each image with an information package containing the image's metadata, projection information, and coordinate information. Collecting all the images and their corresponding information packages together forms the initial set of attachment results. This set provides complete basic information for subsequent processing and application of the remote sensing images.

[0024] In an optional embodiment, the coordinate information files in the initial set of attachment results are processed to unify the decimal places, generating a standardized set of coordinate information files, including: Iterate through the coordinate information files in the initial set of attachment results; The number of decimal places in each coordinate information file is counted, and rounding is performed according to the preset decimal place rules; The processed coordinate information files are updated into the initial set of attachment results to generate a standardized set of coordinate information files.

[0025] It should be noted that we need to find every coordinate information file in the initial set of attachments, in preparation for processing them. Each coordinate information file records a lot of coordinate data, which usually has a decimal part; counting the number of decimal places means looking at how many digits are after the decimal point in each coordinate data; the preset decimal place rules define a standard format for these coordinate data, for example, it is stipulated that only 4 digits can be retained after the decimal point in the coordinate data; Then, according to this rule, the coordinate data is rounded. Rounding means that if the fifth digit after the decimal point is greater than or equal to 5, the fourth digit is added by 1; if it is less than 5, the fifth digit and subsequent digits are discarded. This is done to make the coordinate data format more uniform and standardized, which is convenient for subsequent use and analysis. The rounded coordinate information file is placed back into its corresponding position in the original initial attachment results set, replacing the original coordinate information file. After this operation, all coordinate information files in the initial attachment results set become standardized files that conform to the preset decimal place rules. These standardized coordinate information files are then extracted to form a standardized coordinate information file set, in which the files all have a uniform and standardized coordinate data format.

[0026] In an optional embodiment, based on the mosaic image set in the multi-source remote sensing image set, as well as the mosaic surface vector file and the structured metadata object set, mosaic block vector files are generated in batches through spatial analysis and spatial assignment. Furthermore, projection information files are generated in batches according to the projection information of the image set, forming a set of image attachment results, including: Determine whether mosaicked imagery exists in a multi-source remote sensing image set; If mosaicked imagery exists, load the mosaic surface vector file; Based on the side view, temporal and data source type information in the structured metadata object set, spatial analysis and attribute assignment are performed on the mosaic surface vector file, and mosaic block vector files are generated in batches according to the tiling module; Add the mosaic block vector files to the complete set of attachments to complete the update.

[0027] It should be noted that mosaic imagery is a special form of image organization. It usually involves dividing a large area into multiple smaller images and then stitching or mosaicking these images together. The task is to carefully examine this large warehouse to see if there are any images that have undergone this mosaic imagery process. Once it is determined that mosaic sub-images exist in the multi-source remote sensing image set, the corresponding mosaic surface vector file needs to be loaded. The mosaic surface vector file can be understood as a map template describing the boundaries and structure of the mosaic sub-images. It records the shape, position and other information of each mosaic sub-image in a vector manner. Loading this file is equivalent to obtaining this map template, so that further processing and analysis can be carried out. Spatial analysis of the mosaic facet vector file is performed using side-view, temporal, and data source type information from a collection of structured metadata objects, as detailed below: The system iterates through each sectional module, calculating its geometric center point or surface boundary. It then performs a spatial overlay analysis between the spatial extent of the sectional module and the coverage area of ​​the original remote sensing image. Specifically, it uses a "spatial connection" operation to find the original full-view image that intersects with the geometric center point of the sectional module or has the largest overlap area with the surface features of the sectional module. After completing the spatial matching, the system automatically establishes a one-to-one correspondence between the sectional module and the original image. Based on these analyses, each tiling module in the mosaic face vector file is assigned corresponding attribute values. Then, mosaic block vector files are generated in batches according to these tiling modules. The attribute assignments are as follows: Based on the spatial correspondence established in the previous step, the structured metadata object corresponding to the original whole-scene image is found from the set of structured metadata objects. Then, the key fields (such as side view, time phase, data source type, track number, image acquisition time, etc.) in the structured metadata object are directly copied or filled into the attribute table of the current frame segmentation module through mapping rules to form new attribute fields. For example, the attribute table of the frame segmentation module will add columns such as "Original Image ID", "Side View", "Time Phase", and "Data Source", whose values ​​are inherited from the original image. Add the generated mosaic vector file to the complete set of attached results; this updates the complete set of attached results, making it contain more comprehensive and detailed information related to remote sensing images, which will facilitate further in-depth research and application of remote sensing images. For example: Suppose that after image mosaicking, an original mosaic surface file covering two images is generated; Spatially connect the original mosaic surface file with the standard map sheet vector to identify that the mosaic surface spans two map sheets, H49G001001 and H49G001002; continue to spatially connect it with the county template to identify the specific location; Construct an attribute asset table with the content {scene number:"GF1_PMS2_..._L1A14220013001", side view:1.76, time phase:20260111, data source:"GF1"}, and associate it with the attribute table of the mosaic facet through the original scene number field; after success, automatically fill in the 1.76° side view and other information corresponding to the scene number into the attributes of all mosaic facet fragments belonging to this image; The area calculation interface is called to calculate the area of ​​the mosaic block within a certain region as 152.37 square kilometers; at the same time, the system date 20260513 is obtained and filled into the production completion time field; Based on the map sheet number attribute, the total mosaic surface is divided into two independent files, H49G001001 mosaic block.shp and H49G001002 mosaic block.shp, and stored in the results directory of the corresponding map sheet respectively.

[0028] In an optional embodiment, based on the completed set of whole-view images, the set of standard view numbers, and the set of structured metadata objects, whole-view image metadata files, projection information files, and coordinate information files for each remote sensing image are produced in batches to obtain an initial set of attachment results, which also includes: Projection information is extracted from the mosaic image set to generate a projection information file, and then decimal place unification processing is performed. Based on the geographic extent and resolution of the remote sensing image, calculate and generate coordinate information files; By associating the mosaic tiled image projection information file, coordinate information file, and corresponding mosaic tiled remote sensing image, an initial set of attachment results is obtained.

[0029] It should be noted that projection information is important data describing the actual positional relationship of points on these images on the Earth's surface, such as which projection method is used to map the three-dimensional Earth's surface onto the two-dimensional image plane; picking out this projection information from the mosaic image set is like finding the key positioning information from a pile of complex drawings. Then, these extracted projection information are organized into a special projection information file, just like organizing the positioning instructions into a separate document. Since images from different sources may differ in the decimal precision of the projection information, in order to unify and standardize subsequent processing, the decimal places in the projection information file need to be standardized, for example, all are retained to 6 decimal places. The geographic extent of remote sensing imagery clearly defines the area of ​​the Earth's surface covered by the imagery; resolution indicates the actual ground distance represented by each pixel on the imagery; based on these two important pieces of information, just like locating the imagery in the large coordinate system of the Earth, the coordinate information of each key point (such as the four corner points) on the imagery can be calculated, and then these coordinate information can be organized to generate a coordinate information file, which records the accurate location information of the imagery in geographic space. Linking the projection information file, coordinate information file, and corresponding mosaic remote sensing imagery is like labeling each image with its projection method and location in geospatial space. In this way, each mosaic remote sensing image has complete supporting information. These images and their supporting information are combined to form the initial set of appendix results, providing basic and comprehensive data for subsequent image analysis and applications.

[0030] In an optional embodiment, the files in the set of attached images are automatically organized into the corresponding image result folders according to the image naming rules, generating a structured result directory, including: Load the preset directory structure specification; the directory structure specification includes the naming rules and hierarchical relationship of folders; According to the cataloging requirements for the whole-scene image results and the mosaic-frame image results, the files in the complete set of attachment results are classified into the corresponding folders; Following the directory structure specifications, folders are automatically created and files are moved to generate a structured output directory.

[0031] It should be noted that, assuming there is a pre-designed file storage map, this map specifies in detail which folders should be placed in different types of files, and how these folders are nested. This is the preset directory structure specification. It includes folder naming rules, such as folder names should be based on the content, source, or purpose of the files, and cannot be randomly chosen. It also specifies the hierarchical relationship, showing the containment relationship between folders. Loading this specification is like putting this map into your brain or computer's processing program so that you can organize the documents according to this rule later. The catalog organization requirement is to find the corresponding folders for the different types and uses of whole-view imagery and mosaic imagery; to carefully analyze the characteristics of each file, determine whether it belongs to whole-view imagery or mosaic imagery, and then classify and put them into the corresponding folders according to the previously loaded catalog structure specifications. After classifying the files, you need to create folders according to the directory structure specifications. If a folder at a certain level does not exist, the computer will automatically create it. Then, move the classified files to their corresponding folders. Through this process, the originally messy files are organized into a well-structured directory, with each file having its own fixed location for easy retrieval and use later.

[0032] In an optional embodiment, the structured output catalog is checked for completeness. If any files are missing, they are generated and supplemented until the structured output catalog is complete and error-free. The final structured output catalog is then used as the final automated processing result of the multi-source remote sensing imagery, including: Traverse the structured output directory and check if the files in each folder are complete; If a missing file is found, it will be regenerated according to its type. Repeat the integrity verification process until the structured output catalog is complete and error-free. The final structured output catalog is used as an appendix to the multi-source remote sensing imagery in the automated processing results.

[0033] It should be noted that the structured output directory is traversed; for example, if a certain folder is specifically used to store the metadata files of the whole scene image, the administrator needs to check whether all the metadata files that should exist are in this folder, whether the number is sufficient, and whether the file names are correct. If a missing file is found in a folder during the inspection, you need to find a way to recreate it based on the type of the missing file. Different types of files have different generation methods. For example, if a metadata file is missing, it can be regenerated using a specialized metadata generation tool or by rewriting it according to certain rules based on the relevant information of the image. If an image data file is missing, you can check if there is a backup or extract and process it from the original data source. New problems may have arisen during the regeneration process, or there may be missing files in other folders; therefore, the entire structured output directory needs to be checked again and again, which is the process of duplication integrity verification. Once the structured output catalog has been checked and supplemented multiple times and is complete and error-free, it can serve as the final output for processing multi-source remote sensing imagery. This output includes all the attachment files related to the multi-source remote sensing imagery. These files are organized together according to a certain structure to facilitate subsequent use, analysis, and research. In other words, it becomes the automated processing result of the attachment outputs for multi-source remote sensing imagery.

[0034] In an optional embodiment, the method further includes: The vector annotation results of manual geographic location accuracy and quality checks are batch-produced into quality check sheets and planar accuracy check statistical tables according to national or local standards, resulting in a set of inspection record results. Accordingly, the vector annotation results of manual geographic location accuracy and quality checks are batch-produced into quality check sheets and planar accuracy check statistical tables according to national or local standards, resulting in a set of inspection record results, including: Load the quality inspection vector file; the quality inspection vector file includes quality problem notes for remote sensing images, and the accuracy check vector includes image coordinates and measured coordinates; Based on the preset sizing modules and quality inspection rules, the quality problem notes are associated according to the sizing modules, and the number and distribution of various quality problems are statistically analyzed. According to the accuracy check rules, the standard deviation and limit deviation are preset according to different resolutions, the residuals are calculated according to the image coordinates and the measured coordinates, and the mean error and score are statistically calculated. Fill the quality inspection results and accuracy inspection results into the quality inspection form and the plane accuracy inspection statistics form according to the sectional module, respectively; Add the quality checklist and accuracy checklist to the standardized coordinate information file set to obtain a complete set of attachment results.

[0035] It should be noted that the national or local standards on which this application is based are: CH / T 1027-2012 Quality Inspection Procedure for Digital Orthophoto Maps, GB / T 18316-2008 Quality Inspection and Acceptance of Digital Surveying and Mapping Results, and GB / T24356-2023 Quality Inspection and Acceptance of Surveying and Mapping Results. The quality inspection vector file contains two important pieces of information. One is the quality problem notes of the remote sensing image, which details what problems exist in the image, such as blurry areas or cloud cover affecting observation. Another type is the accuracy check vector, which records image coordinates and measured coordinates. Image coordinates are the position information of a point on the image in a specific coordinate system, while measured coordinates are the accurate position information of that point obtained through field measurements. By comparing these two, the accuracy of the image can be checked. Load these files for subsequent analysis and processing. The pre-defined quality inspection rules specify which quality problems belong to which category and how to judge them. Based on these rules, the quality problem notes are linked to the corresponding segmentation modules, for example, which segmentation module a certain ambiguity problem appears in. Then, the number of various quality problems in each tiling module is counted. For example, if a certain tiling module has 3 blur problems and 2 cloud occlusion problems, it can also know the distribution of these quality problems in the entire image, which tiling modules have more problems and which have fewer problems. Remote sensing images of different resolutions have different accuracy requirements, so it is necessary to preset the standard deviation and limit deviation according to the resolution. The standard deviation can be understood as the degree to which the data deviates from the average value, while the limit deviation is the maximum allowable error range. The residual is the difference between the image coordinates and the measured coordinates. By calculating the residual, we can know how much the position of a point on the image deviates from its actual position. The standard error is an indicator that reflects the degree of dispersion of a set of errors. Statistical analysis of the standard error can provide an understanding of the overall accuracy. The score is an evaluation score given according to certain rules based on the accuracy. For example, the higher the accuracy, the higher the score. The specific thresholds for the preset standard deviation and limit deviation are set according to the image spatial resolution classification, and the classification rules are as follows: The maximum horizontal position error for flat and hilly areas is set to 10 pixels, and the median error is preset to 5 pixels. For difficult areas such as mountains and high mountains, the median error is preset to 1.5 times the median error. If the image resolution is 0.5 meters, the maximum horizontal position error for flat and hilly areas is 5 meters, and the median error is 2.5 meters. For mountains and high mountains, the median error is 7.5 meters, and the median error is 3.75 meters. The above thresholds can be manually adjusted via configuration files; Residual statistics: For each check point in the accuracy check vector file, automatically calculate the absolute value of the difference between its image coordinates and measured coordinates in the X and Y directions of the plane; record these two differences for each check point and mark whether they are within the preset tolerance range; Mean error calculation: Collect the differences of all checkpoints in the X direction, sum the squares and divide by the total number of checkpoints, then take the square root of the result to obtain the mean error in the X direction; similarly, obtain the mean error in the Y direction; the final mean error in the plane position is the square root of the sum of the squares of the mean errors in the X and Y directions. Score Calculation: The mathematical accuracy score is assessed using a testing method, and is calculated according to the following formula: When m0 ≥ m ≥ 0.3m0, S = 60 + 40 / (0.7 * m0) * (m0 - m); when m ≤ 0.3m0, S = 100; where: S is the score of the primary quality indicator or inspection item involving the mean square error, m0 is the allowable value of the mean square error, and m is the detected value of the mean square error; when the score S ≥ 90, the quality grade is Excellent; 75 ≤ S ≤ 90, the quality grade is Good; 60 ≤ S ≤ 75, the quality grade is Acceptable; S ≤ 60, the quality grade is Unacceptable. The quality checklist is a table that records quality issues, while the planar precision inspection statistics table records precision inspection results. The quality inspection results, such as the number and distribution of quality issues for each section module, as well as the precision inspection results, such as the mean square error and score, are entered into the corresponding tables. This way, the quality and precision status of each section module is immediately clear. The standardized coordinate information file set already contained some files related to the coordinates of remote sensing images. Now, by adding the quality checklist and the planar accuracy check statistics table, which record the quality and accuracy, this file set contains all the important information such as the coordinates, quality, and accuracy of the remote sensing images, becoming a complete set of appendix deliverables.

[0036] Example 2, please refer to Figure 2 This invention provides a technical solution: an automated processing system for attachment results of multi-source remote sensing imagery, applicable to the aforementioned automated processing method for attachment results of multi-source remote sensing imagery, comprising: Data acquisition unit 1 is used to acquire a set of multi-source remote sensing images and its corresponding set of raw metadata files; wherein, the set of raw metadata files includes unstructured descriptive information of key parameters of side view, temporal phase and data source type of each remote sensing image; the core parameters of side view, temporal phase and data source type of each remote sensing image are parsed and extracted from the set of raw metadata files to generate a set of structured metadata objects. File production unit 2 is used to generate a standard scene number set based on the data source type, track number and image acquisition time information in the structured metadata object set; based on the completed whole scene image set, standard scene number set and structured metadata object set, it batch produces whole scene image metadata files, projection information files and coordinate information files for each remote sensing image to obtain the initial attachment result set; Quality inspection unit 3 is used to standardize the decimal places of the coordinate information files in the initial set of attached results, generating a set of standardized coordinate information files; to perform quality inspection on the set of standardized coordinate information files and remote sensing images, resulting in a complete set of attached results for whole-scene images; and to automatically organize the files in the complete set of attached results for whole-scene images into the corresponding whole-scene image result folders according to the standard scene number and image type classification rules, generating a structured result directory. The 4th slab image unit is used to generate mosaic block vector files in batches based on the mosaic slab image set in the multi-source remote sensing image set, as well as the mosaic surface vector file and the structured metadata object set, through spatial analysis and spatial assignment, and to generate projection information files in batches according to the projection information of the slab image set, forming a set of slab image attachment results. The result verification unit 5 is used to automatically organize the files in the set of attached results of the segmented image into the corresponding segmented image result folder according to the segmented image naming rules, and generate a structured result directory; to perform a completeness verification on the structured result directory, and if there are missing files, to supplement and generate the missing files until the structured result directory is complete and error-free, and to use the final structured result directory as the final result of the automated processing of multi-source remote sensing images.

[0037] The embodiments of the present invention have been described in detail above with reference to the accompanying drawings. However, the present invention is not limited thereto. Various changes can be made within the scope of knowledge possessed by those skilled in the art without departing from the spirit of the present invention.

Claims

1. An automated processing method for attachment results of multi-source remote sensing imagery, characterized in that, include: Obtain a set of multi-source remote sensing images and its corresponding set of raw metadata files; the set of raw metadata files includes unstructured descriptive information of key parameters of side view, temporal phase and data source type of each remote sensing image; parse and extract the core parameters of side view, temporal phase and data source type of each remote sensing image from the set of raw metadata files to generate a set of structured metadata objects; Based on the data source type, track number, and image acquisition time information in the structured metadata object set, a standard scene number set is generated; based on the completed whole scene image set, standard scene number set, and structured metadata object set, whole scene image metadata files, projection information files, and coordinate information files for each remote sensing image are produced in batches to obtain the initial set of attachment results; The coordinate information files in the initial set of attached results are processed to unify the decimal places, generating a set of standardized coordinate information files; the standardized coordinate information files and remote sensing images are quality checked to obtain a complete set of whole-scene image attached results; the files in the complete set of whole-scene image attached results are automatically organized into the corresponding whole-scene image results folders according to the standard scene number and image type classification rules, generating a structured results directory; Based on the mosaic image set in the multi-source remote sensing image set, as well as the mosaic surface vector file and the structured metadata object set, the mosaic block vector file is generated in batches through spatial analysis and spatial assignment, and the projection information file is generated in batches according to the projection information of the image set, forming a set of image attachment results. The files in the set of attached images are automatically organized into the corresponding image result folders according to the image naming rules, generating a structured result directory. The structured result directory is then checked for completeness. If any files are missing, they are added or generated until the structured result directory is complete and error-free. The final structured result directory is then used as the final automated result of the multi-source remote sensing image processing.

2. The automated processing method for attachment results of multi-source remote sensing imagery according to claim 1, characterized in that, Each object in the structured metadata object collection corresponds to a core parameter of a remote sensing image; Based on the data source type, track number, and image acquisition time information in the structured metadata object set, a standard scene number set is generated, including: Extract data source type, track number, and image acquisition time information from a collection of structured metadata objects; According to the preset naming rules, the extracted information is concatenated into a standard scene number; the naming rules include the combination of data source identifier, track number, image acquisition time and band identifier; By associating standard scene numbers with their corresponding remote sensing images, a set of standard scene numbers is obtained.

3. The automated processing method for attachment results of multi-source remote sensing imagery according to claim 2, characterized in that, Based on the completed set of whole-view images, standard view number sets, and structured metadata objects, batch-produce the whole-view image metadata files, projection information files, and coordinate information files for each remote sensing image, resulting in an initial set of attachments, including: Name the whole-scene image metadata file for each remote sensing image based on the standard scene number in the standard scene number set; Extract projection information from a collection of structured metadata objects or a collection of whole-scene images to generate a projection information file; Based on the geographic extent and resolution of the remote sensing image, calculate and generate coordinate information files; By associating the whole-view image metadata file, projection information file, and coordinate information file with the corresponding whole-view remote sensing image, an initial set of attachment results is obtained.

4. The automated processing method for attachment results of multi-source remote sensing imagery according to claim 3, characterized in that, The coordinate information files in the initial set of attachments are processed to standardize their decimal places, generating a standardized set of coordinate information files, including: Iterate through the coordinate information files in the initial set of attachment results; The number of decimal places in each coordinate information file is counted, and rounding is performed according to the preset decimal place rules; The processed coordinate information files are updated into the initial set of attachment results to generate a standardized set of coordinate information files.

5. The automated processing method for attachment results of multi-source remote sensing imagery according to claim 4, characterized in that, Based on the mosaic image set in the multi-source remote sensing image collection, as well as the mosaic surface vector files and structured metadata object set, mosaic block vector files are generated in batches through spatial analysis and spatial assignment. Furthermore, projection information files are created in batches based on the projection information of the image set, forming a set of image-attached outputs, including: Determine whether mosaicked imagery exists in a multi-source remote sensing image set; If mosaicked imagery exists, load the mosaic surface vector file; Based on the side view, temporal and data source type information in the structured metadata object set, spatial analysis and attribute assignment are performed on the mosaic surface vector file, and mosaic block vector files are generated in batches according to the tiling module; Add the mosaic block vector files to the complete set of attachments to complete the update.

6. The automated processing method for attachment results of multi-source remote sensing imagery according to claim 5, characterized in that, Based on the completed set of whole-view images, standard view number sets, and structured metadata objects, batch-produce the whole-view image metadata files, projection information files, and coordinate information files for each remote sensing image, resulting in an initial set of attachments, which also includes: Projection information is extracted from the mosaic image set to generate a projection information file, and then decimal place unification processing is performed. Based on the geographic extent and resolution of the remote sensing image, calculate and generate coordinate information files; By associating the mosaic tiled image projection information file, coordinate information file, and corresponding mosaic tiled remote sensing image, an initial set of attachment results is obtained.

7. The automated processing method for attachment results of multi-source remote sensing imagery according to claim 6, characterized in that, The files in the image frame attachment set are automatically organized into the corresponding image frame result folders according to the image frame naming rules, generating a structured result directory, including: Load the preset directory structure specification; the directory structure specification includes the naming rules and hierarchical relationship of folders; According to the cataloging requirements for the whole-scene image results and the mosaic-frame image results, the files in the complete set of attachment results are classified into the corresponding folders; Following the directory structure specifications, folders are automatically created and files are moved to generate a structured output directory.

8. The automated processing method for attachment results of multi-source remote sensing imagery according to claim 7, characterized in that, The structured output catalog undergoes integrity verification. If any files are missing, they are generated and supplemented until the structured output catalog is complete and error-free. The final structured output catalog is then used as the final automated processing result of the multi-source remote sensing imagery, including: Traverse the structured output directory and check if the files in each folder are complete; If a missing file is found, it will be regenerated according to its type. Repeat the integrity verification process until the structured output catalog is complete and error-free. The final structured output catalog is used as an appendix to the multi-source remote sensing imagery in the automated processing results.

9. The automated processing method for attachment results of multi-source remote sensing imagery according to claim 8, characterized in that, The method further includes: The vector annotation results of manual geographic location accuracy and quality checks are batch-produced into quality check sheets and planar accuracy check statistical tables according to national or local standards, resulting in a set of inspection record results. Accordingly, the vector annotation results of manual geographic location accuracy and quality checks are batch-produced into quality check sheets and planar accuracy check statistical tables according to national or local standards, resulting in a set of inspection record results, including: Load the quality inspection vector file; the quality inspection vector file includes quality problem notes for remote sensing images, and the accuracy check vector includes image coordinates and measured coordinates; Based on the preset sizing modules and quality inspection rules, the quality problem notes are associated according to the sizing modules, and the number and distribution of various quality problems are statistically analyzed. According to the accuracy check rules, the standard deviation and limit deviation are preset according to different resolutions, the residuals are calculated according to the image coordinates and the measured coordinates, and the mean error and score are statistically calculated. Fill the quality inspection results and accuracy inspection results into the quality inspection form and the plane accuracy inspection statistics form according to the sectional module, respectively; Add the quality checklist and accuracy checklist to the standardized coordinate information file set to obtain a complete set of attachment results.

10. An automated processing system for attachment results of multi-source remote sensing imagery, applicable to the automated processing method for attachment results of multi-source remote sensing imagery as described in any one of claims 1-9, characterized in that, include: The data acquisition unit is used to acquire a set of multi-source remote sensing images and its corresponding set of raw metadata files. The set of raw metadata files includes unstructured descriptive information of key parameters such as side view, temporal phase, and data source type of each remote sensing image. The unit parses and extracts the core parameters of side view, temporal phase, and data source type of each remote sensing image from the set of raw metadata files to generate a set of structured metadata objects. The file production unit is used to generate a standard scene number set based on the data source type, track number and image acquisition time information in the structured metadata object set; based on the completed whole scene image set, standard scene number set and structured metadata object set, it batch produces whole scene image metadata files, projection information files and coordinate information files for each remote sensing image, and obtains the initial attachment result set. The quality inspection unit is used to standardize the decimal places of the coordinate information files in the initial set of attached results, generating a set of standardized coordinate information files; to perform quality inspection on the set of standardized coordinate information files and remote sensing images, resulting in a complete set of attached results for whole-scene images; and to automatically organize the files in the complete set of attached results for whole-scene images into the corresponding whole-scene image result folders according to the standard scene number and image type classification rules, generating a structured result catalog. The slab image unit is used to generate slab image sets in a multi-source remote sensing image set, as well as slab surface vector files and structured metadata object sets, in batches through spatial analysis and spatial assignment, and to generate slab block vector files in batches according to the projection information of the slab image set, forming a set of slab image attachment results. The results verification unit is used to automatically organize the files in the set of attached results of segmented images into the corresponding segmented image results folders according to the segmented image naming rules, and generate a structured results directory; it performs a completeness verification on the structured results directory, and if there are missing files, it supplements and generates the missing files until the structured results directory is complete and error-free, and uses the final structured results directory as the final automated result of the multi-source remote sensing image.

Images