Domestic change investigation and monitoring-oriented proof image rapid import and quality detection method

Abstract

The invention relates to the technical field of image detection, in particular to a territorial change investigation and monitoring-oriented proof image quick import and quality detection method, which comprises the following steps of: importing proof images from a plurality of heterogeneous image sources in parallel, analyzing geographic coordinates, shooting time and image format attributes through a metadata extraction engine, and determining the quality of the proof images according to the geographic coordinates, the shooting time and the image format attributes; and a standardized template is adopted to complete metadata structure unification. And then constructing a multi-dimensional quality detection chain, automatically detecting the space consistency, the time effectiveness and the image integrity of the proof image, generating a multi-dimensional quality score based on a weighted scoring mechanism, and outputting a high-quality, qualified and unqualified image quality grade set according to a preset threshold value. And for unqualified images, defect types are automatically identified, quality improvement rules are matched to generate improvement suggestions, re-acquisition scheduling is executed, and tasks are assigned to acquisition terminals. According to the invention, full-process automation of evidence-proof images from import, detection to improvement is realized, and the reliability of land change investigation evidence-proof data is improved.

Description

Technical Field

[0001] This invention relates to the field of image detection technology, and in particular to a method for rapid import and quality detection of evidentiary images for land change survey and monitoring. Background Technology

[0002] In land use change surveys and monitoring, evidence images are a crucial data source for determining the authenticity and legality of land use changes. With the application of various data collection methods, such as drone patrols, mobile terminal acquisition, and field surveys using handheld devices, evidence image sources have become increasingly diverse and heterogeneous. Images generated by different acquisition terminals exhibit significant differences in metadata formats, geographic coordinate representations, and recording times, leading to low image import efficiency, inconsistent formats, and data that cannot be directly used for subsequent comparison and analysis.

[0003] In existing technologies, the import and quality inspection of evidence images still primarily rely on manual methods. For example, image sources are manually identified and organized by operators, and metadata information needs to be manually compared and recorded. Image quality judgment depends on human experience assessment, lacking quantifiable quality standards. Furthermore, existing technologies lack multi-dimensional image quality inspection capabilities; for instance, they cannot automatically identify whether the image shooting location is accurate, whether the shooting time meets investigation requirements, or whether the image data is complete and undamaged. This results in a large number of low-quality images being imported into the system, increasing subsequent manual review and re-acquisition tasks. In addition, existing solutions cannot form an automated quality improvement mechanism. Faced with substandard images, they can only rely on manual backtracking, lacking a closed-loop process for problem identification, quality improvement suggestions, re-acquisition scheduling, and feedback monitoring. Ultimately, this leads to low acquisition efficiency, high fieldwork costs, and unstable image quality. Summary of the Invention

[0004] This invention provides a method for rapid import and quality inspection of evidence images for land change survey monitoring. It automates the entire process from image import and inspection to improvement, thereby enhancing the quality and reliability of land change survey evidence data.

[0005] A method for rapid import and quality inspection of evidence images for land change survey and monitoring includes the following steps:

[0006] S1: Import evidence images from multiple heterogeneous image sources in parallel, extract the original metadata through the metadata extraction engine, and use a standardized template to unify the format of the original metadata to generate a standardized metadata data package;

[0007] S2: Based on the standardized metadata, a multi-dimensional quality detection chain is constructed, which includes spatial consistency detection, temporal validity detection and image integrity detection. Spatial consistency detection uses the geographic coordinate information in the standardized metadata to perform spatial matching with the benchmark geographic database.

[0008] S3: The evidence images are detected through the multi-dimensional quality detection chain to generate a multi-dimensional quality score, and the evidence images are dynamically graded based on a preset quality threshold to output a set of image quality levels.

[0009] S4: Based on the set of image quality levels, automatically generate quality improvement suggestions, mark and re-acquire unqualified images, and push qualified images to the land change survey evidence system.

[0010] Optionally, S1 includes:

[0011] S11: Initialize the metadata extraction engine, configure the connection parameters of the metadata extraction engine to simultaneously access multiple heterogeneous image sources, and establish parallel data channels with the multiple heterogeneous image sources.

[0012] S12: Import the evidence images from the multiple heterogeneous image sources in parallel through the metadata extraction engine, and simultaneously extract the original metadata of the evidence images, wherein the original metadata includes geographic coordinates, shooting time and image format attributes;

[0013] S13: Input the raw metadata into a predefined standardized template, and use the standardized template to perform format unification processing on the raw metadata to generate a standardized metadata data package with consistent structure.

[0014] Optionally, the initialization of the metadata extraction engine includes: configuring the metadata extraction engine to support multiple access protocols such as HTTP, FTP, and direct database connection, and setting independent connection timeout parameters and retry mechanisms for each heterogeneous image source.

[0015] Optionally, S2 includes:

[0016] S21: Based on the standardized metadata, configure spatial consistency detection parameters, temporal validity detection parameters, and image integrity detection parameters respectively to form detection chain configuration parameters;

[0017] S22: According to the detection chain configuration parameters, spatial consistency detection, temporal validity detection and image integrity detection are performed in parallel. Spatial consistency detection is achieved by comparing the geographic coordinates in the standardized metadata data package with the spatial features of the benchmark geographic database. Temporal validity detection is achieved by verifying the relationship between the timestamps in the standardized metadata data package and the survey time range. Image integrity detection is achieved by analyzing the image format features and data integrity flags in the standardized metadata data package.

[0018] S23: The spatial consistency detection results, temporal validity detection results, and image integrity detection results are fused to construct a multi-dimensional quality detection chain containing three detection dimensions.

[0019] Optionally, the configuration of spatial consistency detection parameters includes: setting spatial matching tolerance thresholds and connection parameters of the benchmark geodatabase; the configuration of time validity detection parameters includes: defining the survey time range and allowed time deviation values; the configuration of image integrity detection parameters includes: setting a list of supported image formats and integrity verification rules.

[0020] Optionally, the spatial consistency detection specifically includes: extracting geographic coordinates from the standardized metadata data, calculating the Euclidean distance between them and the corresponding reference coordinates in the benchmark geographic database, and determining spatial consistency when the Euclidean distance is less than a preset tolerance threshold. The temporal validity detection specifically includes: extracting timestamp information from the standardized metadata data, verifying whether it falls within a predefined investigation time range, and checking whether the timestamps of adjacent evidence images conform to the logical temporal order. The image integrity detection specifically includes: verifying whether the image format belongs to the supported format list based on the image format features in the standardized metadata data, and simultaneously determining whether the image data is complete and undamaged through checksum verification.

[0021] Optionally, S3 includes:

[0022] S31: Initialize the runtime environment of the multi-dimensional quality inspection chain, configure the calculation parameters and weight coefficients of the multi-dimensional quality score, and generate a configured multi-dimensional quality inspection chain instance;

[0023] S32: Run the configured multi-dimensional quality inspection chain instance to perform a comprehensive inspection of the evidence image, and calculate a multi-dimensional quality score based on the inspection results of three dimensions: spatial consistency, temporal validity and image integrity.

[0024] S33: Compare and analyze the multidimensional quality score with the preset quality threshold, classify the evidence images according to the comparison results, and generate a set of image quality levels including three levels: excellent, qualified, and unqualified.

[0025] Optionally, the preset quality thresholds include a high-quality threshold, a qualified threshold, and a non-qualified threshold, wherein the high-quality threshold is set to 85 points, the qualified threshold is set to 60 points, and the non-qualified threshold is set to below 60 points.

[0026] Optionally, S4 includes:

[0027] S41: Parse the set of image quality levels, identify unqualified images in the set of image quality levels and generate a list of unqualified images, and at the same time extract qualified images from the set of image quality levels to form a list of qualified images.

[0028] S42: Analyze the quality defect types based on the list of non-conforming images, match the predefined quality improvement rule base according to the defect type, and generate personalized quality improvement suggestions for each non-conforming image;

[0029] S43: Automatically execute re-acquisition scheduling tasks based on quality improvement suggestions, assign the list of unqualified images and quality improvement suggestions to the corresponding acquisition terminals, and push the list of qualified images to the land change survey evidence system to complete data entry.

[0030] Optionally, the quality improvement rule base contains improvement scheme templates for different types of quality defects, and the matching of the predefined quality improvement rule base includes: selecting the corresponding improvement scheme template from the quality improvement rule base according to the type of quality defect, and generating quality improvement suggestions containing specific improvement measures.

[0031] The beneficial effects of this invention are:

[0032] This invention employs a parallel import mechanism from multiple heterogeneous image sources, combined with a metadata extraction engine to automatically parse image EXIF ​​information and file header information. It also utilizes a standardized JSON template to unify metadata format, coordinates, time, and image format, thus resolving the problem of slow import due to complex image sources and inconsistent formats in existing land surveys. Furthermore, the system generates a unique identifier for each standardized metadata data package and establishes a metadata index table, enabling full traceability of information such as image source, acquisition terminal, and geographical location, significantly improving the reliability and standardization of evidentiary image data management.

[0033] This invention achieves automated end-to-end screening of evidentiary image quality by employing spatial consistency detection, temporal validity detection, and image integrity detection based on the CRC32 checksum algorithm. The system verifies spatial matching relationships based on Euclidean distance, checks the timeliness of image acquisition through temporal range and logical temporal order, and ensures the image data is undamaged through CRC32 verification. Furthermore, it uses the analytic hierarchy process (AHP) for weight calculation, constructing a weighted quality scoring system to automatically score and classify images, achieving quantitative quality assessment. Compared to traditional methods relying on manual sampling and subjective judgment, this solution enables unattended quality detection, quantifiable results, and repeatable score verification, significantly improving detection efficiency and accuracy.

[0034] This invention automatically identifies substandard images by analyzing a set of image quality levels, automatically matches them to a quality improvement rule base based on defect type, and generates personalized quality improvement suggestions. Simultaneously, it combines regional clustering analysis to reduce the number of repeated field data collections. The system automatically executes re-collection scheduling tasks, pushes the tasks and collection parameters to the collection terminal, and monitors the execution progress in real time based on a task status tracker, achieving a complete closed-loop management from "problem identification → suggestion provision → re-collection → result feedback." This mechanism effectively avoids the problems of repeated trips and blind re-photographing by field personnel, significantly reducing the cost of land survey fieldwork and improving the success rate of re-collection and overall evidence collection efficiency. Attached Figure Description

[0035] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only for this invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0036] Figure 1 This is a schematic diagram of the method flow according to an embodiment of the present invention;

[0037] Figure 2 This is a schematic diagram of the S3 process in an embodiment of the present invention. Detailed Implementation

[0038] The present invention will now be described in detail with reference to the accompanying drawings and specific embodiments. It should also be noted that, to make the embodiments more comprehensive, the following embodiments are the best and preferred embodiments, and those skilled in the art can use other alternative methods to implement some well-known technologies; moreover, the accompanying drawings are only for more specific description of the embodiments and are not intended to specifically limit the present invention.

[0039] It should be noted that the use of terms such as "an embodiment," "an embodiment," "an exemplary embodiment," and "some embodiments" in the specification indicates that the described embodiment may include a specific feature, structure, or characteristic, but not every embodiment necessarily includes that specific feature, structure, or characteristic. Furthermore, when a specific feature, structure, or characteristic is described in connection with an embodiment, implementing such a feature, structure, or characteristic in conjunction with other embodiments (whether explicitly described or not) should be within the knowledge of those skilled in the art.

[0040] Generally, terms can be understood at least partly from their use in context. For example, depending at least partly on the context, the term "one or more" as used herein can be used to describe any feature, structure, or characteristic in a singular sense, or a combination of features, structures, or characteristics in a plural sense. Additionally, the term "based on" can be understood not necessarily to convey an exclusive set of factors, but rather, alternatively, depending at least partly on the context, to allow for the presence of other factors that are not necessarily explicitly described.

[0041] like Figures 1-2 As shown, a method for rapid import and quality inspection of evidentiary images for land change survey and monitoring includes the following steps:

[0042] S1: Import evidence images from multiple heterogeneous image sources in parallel, extract raw metadata using a metadata extraction engine, and unify the format of the raw metadata using a standardized template to generate a standardized metadata data package, specifically:

[0043] S11: Deploy the metadata extraction engine on the computer server and load its runtime configuration file. This configuration file specifies the access protocol types supported by the metadata extraction engine: HTTP, FTP, and direct database connection. The interface configuration module sets connection parameters for each protocol type. For HTTP, the connection parameters are the target URL, access token, and request header format; for FTP, the connection parameters are the host address, port number, username, and password; and for the direct database connection protocol, the connection parameters are the database driver type, connection string, and authentication method.

[0044] Before establishing a connection, a connection timeout parameter and a retry mechanism are set independently for each heterogeneous image source. The connection timeout parameter defines the maximum waiting time for a response, and the retry mechanism defines the number of retries and the retry interval after a failure.

[0045] Once the connection is established, an independent parallel data channel identifier is assigned to each heterogeneous image source to identify the data transmission link corresponding to different image sources, so as to realize parallel data access and subsequent synchronization management.

[0046] S12: Import evidence images from multiple heterogeneous image sources in parallel using the metadata extraction engine, and extract the original metadata at the same time;

[0047] A set of data acquisition threads, the same number as the number of heterogeneous image sources, is established using a multi-threaded scheduling module, with each data acquisition thread corresponding to one heterogeneous image source.

[0048] After the data acquisition thread starts, it initiates image download requests to various heterogeneous image sources through a preset communication protocol and monitors the data transmission status of each thread in real time. While downloading the evidence images, each thread calls the metadata parsing module built into the metadata extraction engine to extract the original metadata from the image files.

[0049] The process of extracting raw metadata includes: parsing the geographic coordinate field and shooting time field from the EXIF ​​information of the evidence image; reading the image format attribute field from the file header information; and combining the parsed geographic coordinate field, shooting time field, and image format attribute field according to the key-value pair structure to form a raw metadata object containing geographic coordinates, shooting time, and image format attributes.

[0050] Each raw metadata object is associated with a corresponding evidence image and stored in a temporary cache in memory as an object, providing basic data for subsequent standardization processing.

[0051] A metadata cache is established to temporarily store raw metadata extracted in parallel. This cache adopts a first-in, first-out (FIFO) queue structure to ensure the consistency of metadata processing order. Dynamic management is implemented based on the cache's maximum capacity threshold. When the number of raw metadata cached reaches a preset threshold, a batch processing mechanism is automatically triggered, transferring a batch of raw metadata from the cache to a standardized template module for unified format processing.

[0052] The cache management module continuously monitors cache usage during operation. When the cache's free capacity is insufficient, it automatically pauses new raw metadata write requests and resumes write operations after batch processing is completed and cache space is released. This design effectively prevents data loss, cache overflow, and thread blocking issues during parallel data acquisition, ensuring the continuity and stability of multi-source data import.

[0053] S13: Input the raw metadata into a predefined standardized template, and use the standardized template to process the raw metadata in a unified format to generate a standardized metadata data package with a consistent structure.

[0054] The standardized template module is invoked to input raw metadata output from the metadata cache into the template processing unit in batches.

[0055] The standardized template is defined based on JSON format and includes a fixed field structure and strict data type specifications. The fields include three parts: GeoCoordinate, CaptureTime, and ImageFormat.

[0056] In the format unification process, the geographic coordinate field in the original metadata is first converted to the WGS84 coordinate system. Coordinate transformation algorithms are used to convert non-WGS84 format coordinates (including GCJ-02 or BD-09) to WGS84 coordinates. Then, the shooting time field is converted to the ISO 8601 standard time format (YYYY-MM-DDTHH:MM:SSZ) to ensure consistent time expression across different devices and systems. Finally, the image format attributes are converted to uppercase string format to ensure consistency of attribute values ​​in subsequent database indexing and matching.

[0057] After format conversion, standardized metadata data packets with consistent structure are generated, and a unique identifier (UUID) is assigned to each standardized metadata data packet. A one-to-one mapping relationship is established between this unique identifier and the corresponding evidence image, and the mapping relationship is written into the metadata index table for subsequent retrieval, tracing, and quality inspection.

[0058] Through this standardized template processing, the metadata of all heterogeneous image sources has been completely unified in terms of semantic structure and field format, providing a data foundation for cross-system comparison and subsequent spatial consistency detection.

[0059] S14: Call the data validation module to perform integrity and standardization checks on the generated standardized metadata data package. The validation includes the following three items:

[0060] (1) Verify whether the geographic coordinates are within the valid range, that is, the latitude value is between -90° and 90°, and the longitude value is between -180° and 180°;

[0061] (2) Verify whether the shooting time conforms to the logical order rules, that is, the shooting time shall not be earlier than the set benchmark time (such as the start time of the land change survey task) and shall not be later than the current time;

[0062] (3) Verify whether the image format attribute belongs to the supported format type set, which includes JPG, PNG, TIFF and BMP types.

[0063] When a standardized metadata package is found to have missing fields, format errors, or abnormal value ranges, it is marked as unqualified and removed from the valid set. The anomaly type and error source are recorded for subsequent traceability. Validated standardized metadata packages are written to the standardized metadata repository, providing high-quality input data for subsequent multi-dimensional quality inspection chains.

[0064] S2: Based on standardized metadata, a multi-dimensional quality detection chain is constructed, including spatial consistency detection, temporal validity detection, and image integrity detection. Spatial consistency detection utilizes geographic coordinate information from the standardized metadata to perform spatial matching with a benchmark geographic database. Specifically:

[0065] S21: Obtain the verified standardized metadata data package and start the detection chain configuration module to configure the spatial consistency detection parameters, temporal validity detection parameters, and image integrity detection parameters respectively.

[0066] (1) Configure spatial consistency detection parameters: Read the GeoCoordinate field from the standardized metadata package for subsequent spatial consistency detection. Set the spatial matching tolerance threshold, which is used to determine whether the geographic coordinates in the evidence image are consistent with the reference coordinates of the benchmark geodatabase. At the same time, configure the connection parameters of the benchmark geodatabase, including the database access address, port number, access credentials and access permissions, for subsequent spatial coordinate comparison.

[0067] (2) Configure time validity detection parameters: Read the CaptureTime field from the standardized metadata data and define the survey time range (including start and end times). Set the allowable time deviation value to handle minor time differences caused by time synchronization errors of the acquisition equipment.

[0068] (3) Configure image integrity detection parameters: Generate a list of supported image formats based on the ImageFormat field in the standardized metadata package. This list includes JPG, PNG, TIFF, and BMP. The CRC32 checksum algorithm (Cyclic Redundancy Check 32-bit) is used as the sole image data integrity verification algorithm. The CRC32 checksum algorithm calculates a checksum value for the image file content based on a fixed generator polynomial, which serves as the basis for determining image data integrity. Simultaneously, the CRC32 checksum result field name and output format are stored in the detection chain configuration module.

[0069] After configuration, detection chain configuration parameters are generated to control subsequent detection execution.

[0070] S22: Perform spatial consistency detection, temporal validity detection, and image integrity detection according to the detection chain configuration parameters.

[0071] The detection chain execution module is started, and three detection dimensions are executed in parallel for each standardized metadata data packet. The specific process is as follows:

[0072] (1) Spatial consistency check execution process: The GeoCoordinate field is extracted from the standardized metadata data as the input geographic coordinates, and the corresponding reference coordinates are read through the benchmark geographic database interface. Then, the distance between the input coordinates and the reference coordinates is calculated using the Euclidean distance calculation method:

[0073] ;

[0074] in, , This represents the coordinate values ​​of the evidence image (from the GeoCoordinate field). , Represents the reference coordinates of the baseline geographic database.

[0075] When distance When the spatial consistency detection status is less than the spatial matching tolerance threshold, the spatial consistency detection status is marked as "spatial consistent".

[0076] (2) Time validity check process: Extract the CaptureTime field from the standardized metadata and determine whether it falls within the survey time range defined in configuration S21. Simultaneously, read the CaptureTime of the previous image and check that the temporal order of the two images must meet the following requirements:

[0077] ;

[0078] When CaptureTime falls within the survey time range and meets the above time logic order, the time validity check status is marked as "time valid".

[0079] (3) Image integrity detection process: The ImageFormat field is read and matched against the list of supported image formats. When the image format is valid, the CRC32 checksum algorithm is used to read the CRC checksum value of the image file and calculate the CRC32 checksum result.

[0080] ;

[0081] The calculated CRC32 value is compared with the CRC check value of the image data storage record. If they match, the image data is determined to be undamaged, and the detection status is marked as "data integrity".

[0082] After the three detection dimensions are completed, the detection output is written to the detection result cache. The detection result cache adopts a distributed storage architecture, and each detection result is appended with a unique detection result identifier and a detection completion timestamp to maintain the sequential consistency and traceability of the detection results. The cache does not block the detection task during the writing process, ensuring the stability of high-concurrency processing.

[0083] S23: Fuse the detection results to construct a multi-dimensional quality detection chain. Activate the detection result fusion module to weight the spatial consistency detection results, temporal validity detection results, and image integrity detection results according to preset weights, generating a comprehensive quality score, represented as:

[0084] ;

[0085] in, , , These are the quantitative results corresponding to the three tests, , , These represent the weights for the corresponding detection dimensions. The overall quality score and the detection results from the three dimensions are encapsulated together to form a multi-dimensional quality detection chain for the evidence image. During the construction of the detection chain, a dynamic update mechanism is enabled. When anomalies persist in a certain detection dimension, the corresponding detection parameters are automatically adjusted to improve the accuracy of subsequent detections.

[0086] S24: Perform verification tests on the completed multi-dimensional quality inspection chain to ensure the reliability of the inspection chain. Execute the verification tests sequentially:

[0087] Verify the accuracy of spatial consistency detection using test images with precise geographic coordinates;

[0088] The accuracy of CaptureTime determination was verified using data from simulated shooting times at different times.

[0089] Verify the sensitivity of CRC32 checksum using an image containing corrupted CRC values.

[0090] Once verified, the detection chain is put into practice in the actual evidence image examination process.

[0091] S3: The evidence images are inspected through a multi-dimensional quality detection chain to generate multi-dimensional quality scores. Based on preset quality thresholds, the evidence images are dynamically graded, and a set of image quality levels is output, specifically:

[0092] S31: Initialize the runtime environment of the multi-dimensional quality inspection chain, configure the calculation parameters and weight coefficients of the multi-dimensional quality score, and generate a configured multi-dimensional quality inspection chain instance.

[0093] First, the multi-dimensional quality inspection chain runtime environment initialization module is started to prepare runtime resources for multi-dimensional quality score calculation. The initialization process includes: allocating a computing resource pool for quality score calculation, loading the multi-dimensional quality inspection chain execution engine, and establishing a data reading connection with the detection result cache area.

[0094] The Analytic Hierarchy Process (AHP) was used to calculate the weight coefficients for the three detection dimensions. Spatial consistency, temporal validity, and image integrity were defined as three independent evaluation indicators for the first layer. By constructing pairwise comparison matrices and solving the matrices according to the consistency principle, the weight coefficients of each evaluation indicator were obtained.

[0095] Subsequently, independent scoring scales and calculation rules were established for each dimension based on weighting coefficients. For example, the scoring scale for the spatial consistency dimension was based on the ratio of Euclidean distance to tolerance threshold; the scoring scale for the temporal validity dimension was based on the degree of deviation of the timestamp falling within the survey time range; and the scoring scale for the image integrity dimension was based on the CRC32 check algorithm to determine the integrity of image data and assign scores.

[0096] Once completed, a fully configured multi-dimensional quality inspection chain instance is generated, and this instance is used as input for subsequent score calculations.

[0097] A quality scoring cache is established to store the runtime parameters of the configured multi-dimensional quality inspection chain instance, including the weight coefficients of each dimension, scoring scales, scoring rules, and the weight matrix of the analytic hierarchy process (AHP). The quality scoring cache uses a distributed in-memory database to ensure high-speed access performance of the scoring parameters during the scoring phase.

[0098] Each set of operating parameters is assigned a unique version identifier to record the correspondence between parameter versions and scoring results, preventing scoring results from becoming untraceable due to parameter changes. Before each calculation of the multidimensional quality score, the corresponding operating parameters are read from the quality score cache based on the version identifier, ensuring consistency in the scoring process.

[0099] S32: Run the configured multi-dimensional quality inspection chain instance to perform a comprehensive inspection of the evidence image, and calculate a multi-dimensional quality score based on the inspection results of three dimensions: spatial consistency, temporal validity and image integrity.

[0100] Start a multi-dimensional quality inspection chain instance and read the three inspection results (spatial consistency inspection result, temporal validity inspection result, and image integrity inspection result) of the corresponding evidence image from the inspection result cache.

[0101] The three test results were quantified into scores:

[0102] Spatial consistency score: The score is calculated based on the ratio between the Euclidean distance value in the detection results and the tolerance threshold;

[0103] Time validity score: The score is calculated based on the relationship between CaptureTime and the survey time range;

[0104] Image integrity score: The score is calculated based on whether the CRC32 check result matches and whether the image format belongs to the supported list.

[0105] Then, a weighted summation is performed based on the weight coefficients stored in the quality score cache:

[0106] ;

[0107] in, For spatial consistency score, Score for time efficiency. To score image integrity, , , These are the corresponding weighting coefficients.

[0108] The final output is a multidimensional quality score ranging from 0 to 100.

[0109] S33: Compare and analyze the multidimensional quality score with the preset quality threshold, classify the evidence images according to the comparison results, and generate a set of image quality levels including three levels: excellent, qualified, and unqualified.

[0110] The multidimensional quality scores are compared and analyzed with preset quality thresholds. The preset quality thresholds include a high-quality threshold, a qualified threshold, and a non-qualified threshold, where the high-quality threshold is set to 85 points, the qualified threshold is set to 60 points, and the non-qualified threshold is set to below 60 points.

[0111] For each illustrative image, a complete quality score record is generated, which includes spatial consistency score, temporal validity score, image integrity score, and final multidimensional quality score. The image is then classified into the corresponding level category, ultimately forming a set of image quality levels.

[0112] S34: Establish a verification mechanism for the image quality level set, and verify the accuracy of the classification of each level in the image quality level set through sampling review.

[0113] After the image quality level set is formed, the image quality level set verification module is activated to verify the accuracy of the level classification through sampling review.

[0114] Evidence images are randomly selected independently from three levels—high-quality, qualified, and unqualified—in a fixed proportion and submitted sequentially to a manual review terminal. Reviewers with image discrimination experience manually review the spatial consistency score, temporal validity score, and image integrity score of each evidence image to ensure that the level classification is consistent with the actual situation.

[0115] For evidence images that are disputed after manual review, the manual review process is automatically initiated, and the review results are recorded in the classification record database as a basis for subsequent detection chain optimization.

[0116] S4: Based on the image quality level set, automatically generate quality improvement suggestions, mark and re-acquire unqualified images, and push qualified images to the land change survey evidence system. Specifically:

[0117] S41: Parse the image quality level set, identify unqualified images and generate an image list, start the image quality parsing module, and read each evidence image record in the image quality level set. The image quality level set stores the level identifier, multi-dimensional quality score, and detailed scores for each evidence image in the form of a data structure, including spatial consistency score, temporal validity score, and image integrity score.

[0118] The classification process based on the grade identifier is as follows:

[0119] Read the image's rating label;

[0120] Determine whether the image belongs to the "unacceptable", "acceptable", or "high-quality" level;

[0121] For images rated "unacceptable", their image identifiers and corresponding multidimensional quality score details are written into the unacceptable image list;

[0122] For images rated "qualified" or "excellent", their image identifiers and corresponding multidimensional quality score details are written into the qualified image list.

[0123] During the classification process, ensure that each evidence image appears in only one list, and generate unique list numbers for both lists for subsequent reference.

[0124] A defect analysis cache is established to temporarily store a list of non-conforming images, awaiting quality defect type analysis and processing. The defect analysis cache employs a priority queue structure, and processing priorities are set based on the multi-dimensional quality scores of the non-conforming images; images with lower scores have higher priority.

[0125] Add priority tags and timestamps to each non-compliant image record to ensure that the order and priority of defect processing are traceable.

[0126] S42: Analyze the quality defect types based on the list of non-conforming images and generate personalized quality improvement suggestions. Start the quality defect analysis module, read non-conforming image records one by one from the defect analysis cache, obtain multi-dimensional quality score details for each non-conforming image, and identify the quality defect types using the following methods:

[0127] When the spatial consistency score is lower than the set threshold, it is judged as a spatial consistency defect;

[0128] When the time validity score is lower than the set threshold, it is judged as a time validity defect;

[0129] When the image integrity score is low, it is judged as an image integrity defect.

[0130] Based on the defect type of each non-conforming image, a corresponding improvement scheme template is matched from the quality improvement rule base, for example:

[0131]

[0132] Further regional clustering analysis is performed, and non-compliant images with adjacent distribution locations are clustered based on their geographical coordinates to generate regionalized batch quality improvement suggestions, thereby reducing the repetitive workload of the acquisition terminal and improving the efficiency of re-acquisition.

[0133] Finally, a quality improvement suggestion containing improvement measures and implementation parameters is generated for each image.

[0134] S43: Automatically execute re-acquisition scheduling tasks based on quality improvement suggestions, assign the list of unqualified images and quality improvement suggestions to the acquisition terminals, and push qualified images into the database. The process includes:

[0135] Establish a reacquisition task queue, convert each image record in the list of unqualified images into a reacquisition task, and add a priority label;

[0136] Set the data collection parameters according to the quality improvement suggestions, including data collection location, data collection time window, data collection terminal number, etc.

[0137] Task assignment: Send the re-collection task and corresponding quality improvement suggestions to the collection terminal through the task scheduling interface;

[0138] Push qualified images into the database: Perform data format conversion on the list of qualified images, encapsulate the list data into a data packet that conforms to the interface specification of the land change survey and evidence collection system, and complete the data entry operation through the data interface.

[0139] After the task is assigned, the acquisition terminal directly performs on-site image acquisition based on the quality improvement suggestions.

[0140] S44: Establish a feedback monitoring mechanism to monitor the execution status of reacquisition tasks and update the task status flags in the list of unqualified images.

[0141] A feedback monitoring mechanism is activated to obtain the task execution status of the acquisition terminal in real time through the task status tracker, including the task reception status, on-site acquisition progress, and acquisition completion status.

[0142] The task status identifier in the list of defective images is updated in real time. The task status includes:

[0143] Status updates are stored in the task management database for closed-loop task management and subsequent audit tracking.

[0144] This invention encompasses any substitutions, modifications, equivalent methods, and solutions made within the spirit and scope of this invention. To provide the public with a thorough understanding of this invention, specific details are described in detail in the following preferred embodiments; however, those skilled in the art will fully understand the invention even without these details. Furthermore, to avoid unnecessary misunderstanding of the essence of this invention, well-known methods, processes, procedures, components, and circuits are not described in detail.

[0145] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A method for rapid import and quality inspection of evidentiary images for land change survey and monitoring, characterized in that, Includes the following steps: S1: Import evidence images from multiple heterogeneous image sources in parallel, extract the original metadata through the metadata extraction engine, and use a standardized template to unify the format of the original metadata to generate a standardized metadata data package; S2: Based on the standardized metadata, a multi-dimensional quality detection chain is constructed, which includes spatial consistency detection, temporal validity detection and image integrity detection. Spatial consistency detection uses the geographic coordinate information in the standardized metadata to perform spatial matching with the benchmark geographic database. S3: The evidence images are detected through the multi-dimensional quality detection chain to generate a multi-dimensional quality score, and the evidence images are dynamically graded based on a preset quality threshold to output a set of image quality levels. S4: Based on the set of image quality levels, automatically generate quality improvement suggestions, mark and re-acquire unqualified images, and push qualified images to the land change survey evidence system.

2. The method for rapid import and quality inspection of evidentiary images for land change survey and monitoring according to claim 1, characterized in that, S1 includes: S11: Initialize the metadata extraction engine, configure the connection parameters of the metadata extraction engine to simultaneously access multiple heterogeneous image sources, and establish parallel data channels with the multiple heterogeneous image sources. S12: Import the evidence images from the multiple heterogeneous image sources in parallel through the metadata extraction engine, and simultaneously extract the original metadata of the evidence images, wherein the original metadata includes geographic coordinates, shooting time and image format attributes; S13: Input the raw metadata into a predefined standardized template, and use the standardized template to perform format unification processing on the raw metadata to generate a standardized metadata data package with consistent structure.

3. The method for rapid import and quality inspection of evidentiary images for land change survey and monitoring according to claim 2, characterized in that, The initialization of the metadata extraction engine includes: configuring the metadata extraction engine to support multiple access protocols such as HTTP, FTP, and direct database connection, and setting independent connection timeout parameters and retry mechanisms for each heterogeneous image source.

4. The method for rapid import and quality inspection of evidentiary images for land change survey and monitoring according to claim 2, characterized in that, S2 includes: S21: Based on the standardized metadata, configure spatial consistency detection parameters, temporal validity detection parameters, and image integrity detection parameters respectively to form detection chain configuration parameters; S22: According to the detection chain configuration parameters, spatial consistency detection, temporal validity detection and image integrity detection are performed in parallel. Spatial consistency detection is achieved by comparing the geographic coordinates in the standardized metadata data package with the spatial features of the benchmark geographic database. Temporal validity detection is achieved by verifying the relationship between the timestamps in the standardized metadata data package and the survey time range. Image integrity detection is achieved by analyzing the image format features and data integrity flags in the standardized metadata data package. S23: The spatial consistency detection results, temporal validity detection results, and image integrity detection results are fused to construct a multi-dimensional quality detection chain containing three detection dimensions.

5. A method for rapid import and quality inspection of evidentiary images for land change survey and monitoring according to claim 4, characterized in that, The configuration of spatial consistency detection parameters includes: setting spatial matching tolerance threshold and connection parameters of the benchmark geographic database; the configuration of time validity detection parameters includes: defining the survey time range and the allowed time deviation value; and the configuration of image integrity detection parameters includes: setting the list of supported image formats and integrity verification rules.

6. The method for rapid import and quality inspection of evidentiary images for land change survey and monitoring according to claim 4, characterized in that, The spatial consistency detection specifically includes: extracting geographic coordinates from the standardized metadata data, calculating the Euclidean distance between them and the corresponding reference coordinates in the benchmark geographic database, and determining spatial consistency when the Euclidean distance is less than a preset tolerance threshold. The temporal validity detection specifically includes: extracting timestamp information from the standardized metadata data, verifying whether it falls within a predefined investigation time range, and checking whether the timestamps of adjacent evidence images conform to the logical temporal order. The image integrity detection specifically includes: verifying whether the image format belongs to the supported format list based on the image format characteristics in the standardized metadata data, and simultaneously determining whether the image data is complete and undamaged through checksum verification.

7. A method for rapid import and quality inspection of evidentiary images for land change survey and monitoring according to claim 4, characterized in that, S3 includes: S31: Initialize the runtime environment of the multi-dimensional quality inspection chain, configure the calculation parameters and weight coefficients of the multi-dimensional quality score, and generate a configured multi-dimensional quality inspection chain instance; S32: Run the configured multi-dimensional quality inspection chain instance to perform a comprehensive inspection of the evidence image, and calculate a multi-dimensional quality score based on the inspection results of three dimensions: spatial consistency, temporal validity and image integrity. S33: Compare and analyze the multidimensional quality score with the preset quality threshold, classify the evidence images according to the comparison results, and generate a set of image quality levels including three levels: excellent, qualified, and unqualified.

8. A method for rapid import and quality inspection of evidentiary images for land change survey and monitoring according to claim 7, characterized in that, The preset quality thresholds include a high-quality threshold, a qualified threshold, and a non-qualified threshold, wherein the high-quality threshold is set to 85 points, the qualified threshold is set to 60 points, and the non-qualified threshold is set to below 60 points.

9. A method for rapid import and quality inspection of evidentiary images for land change survey and monitoring according to claim 7, characterized in that, S4 includes: S41: Parse the set of image quality levels, identify unqualified images in the set of image quality levels and generate a list of unqualified images, and at the same time extract qualified images from the set of image quality levels to form a list of qualified images. S42: Analyze the quality defect types based on the list of non-conforming images, match the predefined quality improvement rule base according to the defect type, and generate personalized quality improvement suggestions for each non-conforming image; S43: Automatically execute re-acquisition scheduling tasks based on quality improvement suggestions, assign the list of unqualified images and quality improvement suggestions to the corresponding acquisition terminals, and push the list of qualified images to the land change survey evidence system to complete data entry.

10. A method for rapid import and quality inspection of evidentiary images for land change survey and monitoring according to claim 9, characterized in that, The quality improvement rule base contains improvement scheme templates for different types of quality defects. The matching of the predefined quality improvement rule base includes: selecting the corresponding improvement scheme template from the quality improvement rule base according to the type of quality defect, and generating quality improvement suggestions containing specific improvement measures.

Images