A method for managing sky-ground multi-source heterogeneous data

Abstract

The method, apparatus, and storage medium for managing multi-source heterogeneous data from space and ground proposed in this application involve acquiring multi-source heterogeneous data, performing corresponding preprocessing operations on different source data to obtain preprocessed data, and then using a unified spatiotemporal benchmark, data quality verification, and data attribute extraction to obtain low-level feature data. Attribute information is extracted from the low-level feature data to obtain corresponding metadata. Based on the metadata, an automatic multi-source data association strategy is used to obtain the relationships between the multi-source heterogeneous data, and a knowledge graph is generated based on these relationships. This application can obtain low-level feature data in a unified format corresponding to multi-source heterogeneous data, and extract and generate corresponding knowledge graphs from the low-level feature data using an automatic multi-source data association strategy, thereby facilitating rapid retrieval and analysis of input target objects.

Description

Technical Field

[0001] This application relates to the field of big data processing technology, and in particular to a method, apparatus and storage medium for managing heterogeneous data from multiple sources in the sky and ground. Background Technology

[0002] With the continuous development of integrated space-air-ground data, people can obtain satellite remote sensing data and ground-collected data in real time. However, given the large volume and diverse data types of both satellite remote sensing and ground-collected data, how to efficiently process, analyze, and store this multi-source data to manage this heterogeneous data is a pressing issue that needs to be addressed. Summary of the Invention

[0003] This application provides a method, apparatus, and storage medium for managing heterogeneous data from multiple sources (sky, ground, and air), thereby proposing an efficient method for managing such data.

[0004] The first aspect of this application proposes a method for managing heterogeneous data from multiple sources (sky, ground, and air), including:

[0005] Acquire multi-source heterogeneous data;

[0006] Perform corresponding preprocessing operations on the different source data in the multi-source heterogeneous data to obtain the corresponding preprocessed data;

[0007] The preprocessed data is processed through a unified spatiotemporal benchmark, data quality verification, and data attribute extraction to obtain the underlying feature data corresponding to the multi-source heterogeneous data.

[0008] The metadata corresponding to the underlying feature data is obtained by extracting attribute information from the underlying feature data.

[0009] Based on the aforementioned metadata, an automatic multi-source data association strategy is used to obtain the association relationships between the multi-source heterogeneous data.

[0010] A knowledge graph is generated based on the relationships to enable rapid retrieval and analysis of the input target object.

[0011] A second aspect of this application provides a management device for heterogeneous data from multiple sources (sky, ground, and air), comprising:

[0012] The acquisition module is used to acquire heterogeneous data from multiple sources.

[0013] The preprocessing module is used to perform corresponding preprocessing operations on different source data in the multi-source heterogeneous data to obtain corresponding preprocessed data;

[0014] The processing module is used to process the preprocessed data through a unified spatiotemporal benchmark, data quality verification, and data attribute extraction to obtain the underlying feature data corresponding to the multi-source heterogeneous data.

[0015] The extraction module is used to extract attribute information from the corresponding underlying feature data to obtain the metadata corresponding to the underlying feature data;

[0016] The mining module is used to obtain the association relationships between the multi-source heterogeneous data based on the metadata using a multi-source data automatic association strategy.

[0017] The generation module is used to generate a knowledge graph based on the relationships, so as to quickly retrieve and analyze the input target objects.

[0018] The third aspect of this application provides a non-transitory computer-readable storage medium, wherein the non-transitory computer-readable storage medium stores a computer program; when the computer program is executed by a processor, it implements the method shown in the first aspect above.

[0019] The technical solutions provided by the embodiments of this application bring at least the following beneficial effects:

[0020] The management method, apparatus, and storage medium for multi-source heterogeneous data from the sky and ground proposed in this application involve acquiring multi-source heterogeneous data, performing corresponding preprocessing operations on different source data to obtain preprocessed data, and then using a unified spatiotemporal benchmark, data quality verification, and data attribute extraction to obtain underlying feature data corresponding to the multi-source heterogeneous data. Attribute information is extracted from the underlying feature data to obtain corresponding metadata. Based on the metadata, an automatic association strategy for multi-source data is used to obtain the relationships between the multi-source heterogeneous data. A knowledge graph is generated based on these relationships for rapid retrieval and analysis of input target objects. Therefore, this application obtains unified-format underlying feature data by using a unified spatiotemporal benchmark, data quality verification, and data attribute extraction on the preprocessed data, enabling analysis using the unified-format underlying feature data corresponding to multi-source heterogeneous data. Simultaneously, the knowledge graph generated based on the relationships effectively integrates the multi-source heterogeneous data to facilitate rapid retrieval and analysis of input target objects.

[0021] Additional aspects and advantages of this application will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of this application. Attached Figure Description

[0022] The above and / or additional aspects and advantages of this application will become apparent and readily understood from the following description of the embodiments taken in conjunction with the accompanying drawings, wherein:

[0023] Figure 1 This is a flowchart illustrating a method for managing heterogeneous data from multiple sources (sky and ground) according to an embodiment of this application.

[0024] Figure 2 This is a schematic diagram of a device for managing heterogeneous data from multiple sources (sky and ground) according to an embodiment of this application. Detailed Implementation

[0025] The embodiments of this application are described in detail below. Examples of these embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and intended to explain this application, and should not be construed as limiting this application.

[0026] The following describes a method and apparatus for managing heterogeneous data from multiple sources (sky and ground) according to embodiments of this application, with reference to the accompanying drawings.

[0027] Example 1

[0028] Figure 1 This is a flowchart illustrating a method for managing heterogeneous data from multiple sources (sky and ground) according to an embodiment of this application. Figure 1 As shown, it may include:

[0029] Step 101: Obtain multi-source heterogeneous data.

[0030] In embodiments of the present invention, multi-source heterogeneous data may include satellite remote sensing data, aerial remote sensing data, ground video data, network information data, geographic information data, and business information data.

[0031] In embodiments of the present invention, satellite remote sensing data can be acquired from multiple satellites. The satellite payloads cover panchromatic, multispectral, hyperspectral, infrared, and SAR.

[0032] Furthermore, in embodiments of the present invention, aerial remote sensing data can be acquired through drone flight. Aerial remote sensing data may include visible light, infrared, and SAR type aerial remote sensing image data, visible light and infrared type video data, and telemetry data such as target tracks, flight tracks, and text messages.

[0033] Furthermore, in embodiments of the present invention, the ground video data mainly includes video surveillance data. Video surveillance data may include vehicle identification and passage data at checkpoints, pedestrian identification data at turnstiles, and facial recognition data from key locations.

[0034] In embodiments of the present invention, network information data may include text, video, audio, images, and geographic information obtained from the network.

[0035] In embodiments of the present invention, geographic information data may include basic support data and urban traffic data. Urban traffic data may include data such as surveillance videos, traffic control signals, traffic signs, parking lot locations, road condition monitoring, and traffic routes.

[0036] In embodiments of the present invention, business information data may include information on personnel, vehicles, organizations, cases, etc. Personnel information may include basic personnel information, household members, temporary residents, and registration of key personnel; vehicle information may include vehicle registration information, highway checkpoints (entry / exit), vehicles belonging to key personnel, and vehicles with traffic violations; organizations may include organization registration; and cases may include case handling measure execution tables, case handling measures, and basic case information.

[0037] Furthermore, in embodiments of the present invention, the format of the aforementioned data may include text data, audio data, image data, and video data.

[0038] Step 102: Perform corresponding preprocessing operations on the different source data in the multi-source heterogeneous data to obtain the corresponding preprocessed data.

[0039] In embodiments of the present invention, performing corresponding preprocessing operations on different source data in multi-source heterogeneous data to obtain corresponding preprocessed data may include performing preprocessing operations on satellite remote sensing data, aerial remote sensing data, ground video data, network information data, geographic information data, and business information data to obtain corresponding preprocessed data. Furthermore, different data correspond to different preprocessing operations.

[0040] Specifically, in embodiments of the present invention, preprocessing satellite remote sensing data may include:

[0041] The acquired satellite remote sensing images are processed into standard scene products to obtain corresponding standard satellite remote sensing images. The processing of standard scene products may include panchromatic data analysis, auxiliary data separation, field-of-view stitching, radiometric correction, and system-level geometric correction.

[0042] Standard satellite remote sensing images are processed to obtain preprocessed data.

[0043] Normalization can be a process of standardizing data according to a pre-defined format or standard.

[0044] Furthermore, in one embodiment of the present invention, preprocessing the aerial remote sensing data may include:

[0045] The original satellite remote sensing images are densified by aerial triangulation to obtain the exterior orientation elements of each image;

[0046] By using the exterior orientation elements of each photograph, a stereo pair of each photograph is formed;

[0047] Based on the stereo image pairs of each image, the corresponding DEM data is obtained through the Digital Elevation Model (DEM).

[0048] Based on the stereo image pairs of each photograph, a corresponding orthophoto map (DOM) is created.

[0049] The method for creating an orthophoto map DOM may include the following steps:

[0050] Step a: Import the aerial triangulation results to establish the survey area file and restore the model.

[0051] Step b: Define the working area for a single model, generate epipolar images, and match the epipolar images to form matching points and equal parallax curves. The working area should be determined as close as possible to the control point lines to prevent gaps between image pairs.

[0052] Step c: Check the matching results and perform interactive stereo editing (region editing, point editing) as needed. The focus of the check may include: areas with tall buildings, blurred areas, shadowed areas, large bodies of water, densely built-up areas, forested areas, and areas with terrain changes such as valleys and ridges. If local matching problems exist, revert to relative orientation, add relative orientation points to the problematic areas, or add feature points and feature lines during matching preprocessing.

[0053] Step d: Generate a single-model DEM.

[0054] Step e: Digital differential correction.

[0055] In one embodiment of the present invention, based on the single-model DEM, the internal and external orientation elements of the image, and the image resolution, a differential correction method is used for correction and image resampling to generate a single-model DOM.

[0056] Step f: Hue or color adjustment.

[0057] In one embodiment of the present invention, before image mosaicking, the hue or color deviation between adjacent images is checked, and image processing methods are used to adjust it as needed so that the resulting images are basically consistent.

[0058] Step g: Set the mosaic range according to the map frame coordinates, specify the file storage path, and execute the image mosaic command to automatically stitch together the entire DOM.

[0059] Step h: DOM format conversion.

[0060] In one embodiment of the invention, DOM format conversion may include converting the DOM from its internal format to GeoTIFF format.

[0061] Step i: Joint inspection.

[0062] The edge inspection can include model edge inspection and map sheet edge inspection. Specifically, it checks whether the transition of images at model edge inspection is natural and whether there are any gaps; and whether the brightness, contrast, and color of images at map sheet edge inspection are basically consistent, and whether the distance between corresponding points exceeds the limit.

[0063] Step j: Refer to the corresponding scale standard sheet digital topographic map format to input finishing annotations for each sheet, finish the DOM, and generate finishing data.

[0064] In an embodiment of the present invention, the map frame decoration annotation may include the map frame, map number (no map title annotation required), publishing authority, aerial photography operation time, publication time, plane coordinate system, and scale.

[0065] Step k: Overlay the finishing data and DOM data to output image map data containing complete finishing data.

[0066] Furthermore, in one embodiment of the present invention, preprocessing the ground video surveillance data may include:

[0067] The ground video surveillance data is processed by feature extraction and structured description to obtain the first preprocessed data;

[0068] In an embodiment of the present invention, ground video surveillance data can be processed by a structured analysis system to extract features and perform structured analysis on the raw video data, outputting first preprocessed data. The first preprocessed data may include video feature data and structured analysis data.

[0069] The first preprocessed data is automatically aggregated and standardized to obtain the second preprocessed data;

[0070] In one embodiment of the present invention, the big data aggregation gateway can import the first processed data in a standard format through a data interface, and realize the automatic aggregation and standardized processing of large-scale streaming video analysis results data through the big data aggregation gateway.

[0071] The second preprocessed data is analyzed. If blacklist data exists, metadata and data elements related to the blacklist data are retrieved to obtain more detailed information about the blacklist.

[0072] In an embodiment of the present invention, the second preprocessed data can be accessed through a secure access platform and then analyzed. Furthermore, during the data analysis process, it can be compared with data in an existing risk database. If blacklist data exists, metadata and data elements related to the blacklist data are retrieved to obtain more detailed information about the blacklist.

[0073] Furthermore, in one embodiment of the present invention, preprocessing the network information data may include:

[0074] Sensitive information is declassified and feature information is extracted from network information data to obtain third preprocessed data;

[0075] The third preprocessing operation is subjected to security scanning, security auditing, and intrusion detection to obtain the fourth preprocessing data;

[0076] The fourth preprocessed data is analyzed. If blacklist data exists, metadata and data elements related to the blacklist data are retrieved to obtain more detailed information about the blacklist.

[0077] In an embodiment of the present invention, the fourth preprocessed data can be accessed through a secure access platform and then analyzed. Furthermore, during the data analysis process, it can be compared and analyzed with data in an existing risk database. If blacklist data exists, metadata and data elements related to the blacklist data are retrieved to obtain more detailed information about the blacklist.

[0078] Furthermore, in one embodiment of the present invention, preprocessing the business information data may include:

[0079] Feature information is extracted from business information data, and then the data is output through a unified API interface or WEB interface via a data server. The output data is compared with an existing risk database. If blacklist data is found during the data comparison and analysis process, the relevant metadata and data elements of the blacklist are retrieved from the data server in a timely manner to obtain more detailed information about the blacklist.

[0080] Furthermore, in one embodiment of the present invention, preprocessing the geographic information data may include:

[0081] Geographic information data is processed through feature extraction, then through a data server, and finally output via a unified API or WEB interface. The data is then compared with an existing risk database. During the data comparison and analysis, if blacklist data is found, the relevant metadata and data elements of the blacklist are retrieved from the data server in a timely manner to obtain more detailed information about the blacklist.

[0082] Furthermore, it should be noted that, in one embodiment of the present invention, the data preprocessing operations can be performed separately at the corresponding data acquisition terminals. And, in another embodiment of the present invention, different preprocessing operations can be performed uniformly for different data sources.

[0083] Step 103: The preprocessed data is processed through a unified spatiotemporal benchmark, data quality verification, and data attribute extraction to obtain the underlying feature data corresponding to the multi-source heterogeneous data.

[0084] In this embodiment of the invention, a unified time base needs to define a standard time, to which any other time can be effectively and accurately converted. For example, the date in the time base uses the Gregorian calendar, and the time uses Beijing time. Through a unified time base, multi-source heterogeneous data can be consistent and meaningful in the time dimension, allowing different data to be used uniformly, thereby improving data utilization efficiency.

[0085] Furthermore, in embodiments of the present invention, the purpose of a unified spatial reference is to provide an efficient and consistent spatial positioning reference for data with spatial attributes, ensuring the consistency, compatibility, and sharing of spatial data. For example, CGCS2000 can be used as the basis and standard for the unified spatial reference.

[0086] Specifically, in the embodiments of the present invention, the aforementioned satellite remote sensing data, aerial remote sensing data, and geographic information data contain time and location information, requiring coordinate transformation and time format conversion based on a unified spatiotemporal reference. The aforementioned ground video data, network information data, and business information data do not directly possess geographic coordinates or other location information, but they do possess indirect location information such as IP addresses and checkpoint numbers, requiring mapping based on a unified spatiotemporal reference to convert the indirect location information into coordinate positions.

[0087] For example, a road traffic security checkpoint monitoring system (security checkpoint system) refers to a road traffic on-site monitoring system that relies on specific locations on the road, such as toll stations, traffic or security checkpoints, to photograph, record, and process all motor vehicles passing through the checkpoint. When unifying the spatiotemporal reference of the checkpoint monitoring system data, the correspondence between checkpoint numbers and coordinates, the coordinates corresponding to the data in the checkpoint monitoring system, and then the unified spatiotemporal reference can be used for data unification.

[0088] Furthermore, in an embodiment of the present invention, a direct weighted comprehensive evaluation method is used to evaluate the quality of remote sensing images. Specifically, different quality level thresholds can be set for each evaluation indicator among the completeness, consistency, accuracy, and image quality of the remote sensing image; based on the actual situation, a weight is set for each indicator in the comprehensive evaluation, such that the sum of the weights is 1; according to the calculation formula of each evaluation indicator, the radiometric quality of the image is evaluated separately, and the evaluation results of each individual indicator are output; the comprehensive value of different quality levels is calculated based on the set weights, and the level corresponding to the maximum comprehensive value represents the level of the comprehensive evaluation. The method for calculating the comprehensive value of the level is as follows: first, all levels are set to 0; starting from the first indicator, if the evaluation result of the indicator is a certain level, the comprehensive value of the corresponding level is increased by the weight of that indicator, and so on until the last indicator.

[0089] In embodiments of the present invention, the integrity evaluation index of remote sensing images may include: whether the metadata records and attribute information of the remote sensing data are complete and whether there are any missing items.

[0090] In embodiments of the present invention, the consistency evaluation indicators for remote sensing images may include: whether the remote sensing data organization method, format, cloud cover size conform to the specifications, and whether they are consistent with previous and other datasets.

[0091] In embodiments of the present invention, the accuracy evaluation indicators of remote sensing images may include: spatial projection of remote sensing data, coordinate information, whether the data is accurate, and whether there are any abnormal or erroneous information.

[0092] In embodiments of the present invention, the image quality evaluation indicators for remote sensing images may include the following aspects:

[0093] ① Clarity index

[0094] Sharpness refers to the clarity of details and boundaries in an image, and is an important factor reflecting the quality of remote sensing images. The sharpness evaluation methods in this application may include at least one of the following: the average gradient method, the gray-level co-occurrence matrix method, and the gray-level-gradient co-occurrence matrix method.

[0095] ② Resolution

[0096] The spatial resolution of remote sensing images is objectively evaluated using the modulation transfer function.

[0097] ③ Noise

[0098] A uniform region of a certain size is selected in the image, and this is done through calculation. The purpose of selecting a uniform region is to eliminate the influence of details such as edges and textures.

[0099] ④ Cloud cover

[0100] Cloud cover can be detected using threshold methods, which can generally be divided into direct threshold methods and cloud index threshold methods.

[0101] ⑤ Invalid pixels

[0102] There are often a certain number of invalid pixels around remote sensing images, the most common being the black or white borders after image geometric correction and digital mosaicking.

[0103] Furthermore, in embodiments of the present invention, a hybrid file and database management approach is adopted to store the underlying extracted data. This hybrid management approach allows large volumes of data to be stored on a file system outside the database system, while the database system stores and manages their attributes and paths. Alternatively, small files such as text information can be directly stored in the database system, effectively solving the problem of increased management difficulty as various types of data increase.

[0104] Furthermore, in embodiments of the present invention, the entire storage architecture can adopt a cloud architecture, thereby effectively changing the storage mode of vertically storing data on one or more fixed devices. By integrating physical resources located on each individual storage device through technologies such as storage virtualization, distributed file systems, and underlying objectification, a logically unified storage resource pool is formed to provide external services, which is conducive to capacity expansion and service capability improvement.

[0105] Step 104: Extract attribute information from the underlying feature data to obtain the metadata corresponding to the underlying feature data.

[0106] In embodiments of the present invention, metadata may include attribute metadata, structural metadata, semantic metadata, target metadata, and event metadata.

[0107] Step 105: Construct the association relationship between multi-source heterogeneous data based on metadata, and obtain the automatic association strategy for multi-source data.

[0108] In embodiments of the present invention, the relationships between multi-source heterogeneous data can be constructed based on the metadata, data attributes, data content, and relationships between the data, resulting in temporal, spatial, attribute, and semantic relationships between the multi-source data. Furthermore, in embodiments of the present invention, an automatic multi-source data association strategy can be derived based on the relationships between multi-source heterogeneous data. This strategy can then establish associations between new data and existing data objects when new data arrives.

[0109] Step 106: Generate a knowledge graph based on the relationships to enable rapid retrieval and analysis of the input target object.

[0110] In one embodiment of the present invention, a knowledge graph is generated based on the association relationship. Then, when the user inputs the target object to be searched, the generated knowledge graph can be used to quickly retrieve and output all data related to the target object, so that the user can obtain comprehensive data on the target object.

[0111] The management method, apparatus, and storage medium for multi-source heterogeneous data from the sky and ground proposed in this application involve acquiring multi-source heterogeneous data, performing corresponding preprocessing operations on different source data to obtain preprocessed data, and then using a unified spatiotemporal benchmark, data quality verification, and data attribute extraction to obtain underlying feature data corresponding to the multi-source heterogeneous data. Attribute information is extracted from the underlying feature data to obtain corresponding metadata. Based on the metadata, an automatic association strategy for multi-source data is used to obtain the relationships between the multi-source heterogeneous data. A knowledge graph is generated based on these relationships for rapid retrieval and analysis of input target objects. Therefore, this application obtains unified-format underlying feature data by using a unified spatiotemporal benchmark, data quality verification, and data attribute extraction on the preprocessed data, enabling analysis using the unified-format underlying feature data corresponding to multi-source heterogeneous data. Simultaneously, the knowledge graph generated based on the relationships effectively integrates the multi-source heterogeneous data to facilitate rapid retrieval and analysis of input target objects.

[0112] Example 2

[0113] Furthermore, Figure 2 This is a schematic diagram of a management device for multi-source heterogeneous data from the sky and ground, according to an embodiment of this application. Figure 2 As shown, it may include:

[0114] Module 201 is used to acquire multi-source heterogeneous data;

[0115] Preprocessing module 202 is used to perform corresponding preprocessing operations on different source data in multi-source heterogeneous data to obtain corresponding preprocessed data;

[0116] Processing module 203 is used to process the preprocessed data through a unified spatiotemporal benchmark, data quality verification, and data attribute extraction to obtain the underlying feature data corresponding to the multi-source heterogeneous data;

[0117] The extraction module 204 is used to extract attribute information from the underlying feature data to obtain the metadata corresponding to the underlying feature data.

[0118] The association module 205 is used to obtain the association relationship between multi-source heterogeneous data by using a multi-source data automatic association strategy based on metadata;

[0119] The generation module 206 is used to generate a knowledge graph based on the relationships, so as to quickly retrieve and analyze the input target object.

[0120] To implement the above embodiments, this disclosure also proposes a non-transitory computer-readable storage medium.

[0121] The non-transitory computer-readable storage medium provided in this disclosure embodiment stores a computer program; when the computer program is executed by a processor, it can achieve the following: Figure 1 The method for managing heterogeneous data from multiple sources (sky, ground, and air) is shown.

[0122] This application proposes a management method, apparatus, and storage medium for multi-source heterogeneous data based on space, air, and ground. The method involves acquiring multi-source heterogeneous data, performing corresponding preprocessing operations on different source data to obtain preprocessed data, and then using a unified spatiotemporal benchmark, data quality verification, and data attribute extraction to obtain underlying feature data corresponding to the multi-source heterogeneous data. Attribute information is extracted from the underlying feature data to obtain corresponding metadata. Based on the metadata, an automatic multi-source data association strategy is used to obtain the relationships between the multi-source heterogeneous data. A knowledge graph is generated based on these relationships for rapid retrieval and analysis of input target objects. Therefore, this application obtains unified-format underlying feature data from the preprocessed data through a unified spatiotemporal benchmark, data quality verification, and data attribute extraction, enabling analysis using this unified-format underlying feature data corresponding to multi-source heterogeneous data. Furthermore, the knowledge graph generated based on the relationships effectively integrates the multi-source heterogeneous data, facilitating rapid retrieval and analysis of input target objects.

[0123] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., refer to specific features, structures, materials, or characteristics described in connection with that embodiment or example, which are included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0124] Any process or method description in the flowchart or otherwise herein can be understood as representing a module, segment, or portion of code comprising one or more executable instructions for implementing custom logic functions or processes, and the scope of the preferred embodiments of this application includes additional implementations in which functions may be performed not in the order shown or discussed, including substantially simultaneously or in reverse order depending on the functions involved, as should be understood by those skilled in the art to which embodiments of this application pertain.

[0125] Although embodiments of this application have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting this application. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of this application.

Claims

1. A method for managing heterogeneous data from multiple sources (sky, ground, and air), characterized in that, The method includes: Acquire multi-source heterogeneous data, including satellite remote sensing data, aerial remote sensing data, ground video data, network information data, geographic information data, and business information data. The satellite payloads of the satellite remote sensing data include panchromatic, multispectral, hyperspectral, infrared, and SAR. The aerial remote sensing data includes visible light, infrared, and SAR type aerial remote sensing image data, visible light and infrared type video data, target tracks, flight tracks, and text message telemetry data. The geographic information data includes basic support data and urban traffic data. The business information data includes personnel, vehicle, organization, and case information. Perform corresponding preprocessing operations on the different source data in the multi-source heterogeneous data to obtain corresponding preprocessed data, including performing preprocessing operations on the satellite remote sensing data, the aerial remote sensing data, the ground video data, and the network information data to obtain corresponding preprocessed data; The preprocessed data is processed through a unified spatiotemporal benchmark, data quality verification, and data attribute extraction to obtain the underlying feature data corresponding to the multi-source heterogeneous data. The underlying feature data is subjected to attribute information extraction to obtain the metadata corresponding to the underlying feature data. The metadata includes attribute metadata, structural metadata, semantic metadata, target metadata, and event metadata. Based on the aforementioned metadata, an automatic multi-source data association strategy is used to obtain the association relationships between the multi-source heterogeneous data. A knowledge graph is generated based on the aforementioned relationships, so that it can be used for rapid retrieval and analysis of the input target object; The preprocessing operation for the satellite remote sensing data includes: The acquired satellite remote sensing images are processed into standard scene products to obtain corresponding standard satellite remote sensing images. The processing of the standard scene products includes panchromatic data analysis, auxiliary data separation, field of view stitching, radiometric correction, and system-level geometric correction. The standard satellite remote sensing imagery is then normalized to obtain preprocessed data. The preprocessing operation for the aerial remote sensing data includes: The original satellite remote sensing images were subjected to aerial triangulation to obtain the exterior orientation elements of each image; A stereo pair of each image is formed using the exterior orientation elements of each image; Based on the stereo image pairs of each image, the corresponding DEM data is obtained through the Digital Elevation Model (DEM). Based on the stereo image pairs of each photograph, a corresponding orthophoto map (DOM) is created. Preprocessing of ground video surveillance data includes: The ground video surveillance data is processed by feature information extraction and structured description to obtain the first preprocessed data; The first preprocessed data is automatically aggregated and standardized to obtain the second preprocessed data; The second preprocessed data is analyzed. If blacklist data exists, metadata and data elements related to the blacklist data are retrieved to obtain more detailed information about the blacklist. Preprocessing of network information data includes: Sensitive information declassification and feature information extraction are performed on the network information data to obtain third preprocessed data; The third preprocessed data is subjected to security scanning, security auditing, and intrusion detection operations to obtain the fourth preprocessed data; The fourth preprocessed data is analyzed. If blacklist data exists, metadata and data elements related to the blacklist data are retrieved to obtain more detailed information about the blacklist.

2. The method as described in claim 1, characterized in that, The method further includes: The step of performing corresponding preprocessing operations on different source data in the multi-source heterogeneous data to obtain corresponding preprocessed data includes performing preprocessing operations on the geographic information data and the business information data respectively to obtain corresponding preprocessed data.

3. A management device for heterogeneous data from multiple sources (sky, ground, and air), characterized in that, The device is used to implement the management method for multi-source heterogeneous data from the sky and ground as described in claim 1, and the device includes the following modules: The acquisition module is used to acquire heterogeneous data from multiple sources. The preprocessing module is used to perform corresponding preprocessing operations on different source data in the multi-source heterogeneous data to obtain corresponding preprocessed data; The processing module is used to process the preprocessed data through a unified spatiotemporal benchmark, data quality verification, and data attribute extraction to obtain the underlying feature data corresponding to the multi-source heterogeneous data. The extraction module is used to extract attribute information from the underlying feature data to obtain the metadata corresponding to the underlying feature data; The mining module is used to obtain the association relationships between the multi-source heterogeneous data based on the metadata using a multi-source data automatic association strategy. The generation module is used to generate a knowledge graph based on the relationships, so as to quickly retrieve and analyze the input target objects.

4. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the method as described in any one of claims 1-2.

Images