A remote sensing data query system

Abstract

The application provides a remote sensing data query system, and relates to the technical field of remote sensing data.The system comprises a data acquisition module configured to acquire remote sensing data, including image data and metadata; a data processing module configured to process the image data, and obtain available domain data and label data based on the image data; an index creation module configured to create an index based on the image data, the metadata, the available domain data and the label data, and obtain an encoding index table, wherein the encoding index table at least comprises the following contents: the identification name of the metadata, the image data encoding, and the association encoding of the available domain data and the label data; and a data query module configured to query based on the encoding index table, and obtain a query result.The application improves the utilization rate of remote sensing data, and the query result is accurate, the user operation is convenient and clear, and the user experience and query efficiency are improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of remote sensing data, and particularly relates to a remote sensing data query system. BACKGROUND

[0002] In recent years, remote sensing satellite data has become a hot topic in the fields of national security, foreign policy, global environmental change scientific research, etc. With the accelerating pace of development of China's earth observation satellite series and the launch of the national high-resolution earth observation system project, the remote sensing satellite application industry has become a national strategic high-tech industry, and is changing from experimental application to business service. Satellite applications have become an important support for China's economic construction, social development and government decision-making. Over the past 10 years, China has formed a series of earth observation satellites such as China-Brazil Earth Resources Satellite, Environmental Disaster Mitigation Satellite, Surveying Satellite, and High-Resolution Satellite, built a national unified land observation remote sensing data receiving and processing ground system, accumulated rich remote sensing satellite data resources, and formed a satellite remote sensing service and application system.

[0003] At present, due to the limited resources of single satellite images and the influence of weather and other environmental factors, remote sensing applications usually involve the comprehensive application of multi-satellite multi-load data. For example, in the case of remote sensing flood emergency monitoring, satellite data not only includes high-resolution data, but also may include resource satellites, environmental disaster mitigation satellites, foreign satellites, military satellites, aerial images, etc. The sensors that may be used include visible light, radar, infrared, etc. For the research of multi-source remote sensing data information service system, the daily management and use of remote sensing data obtained by each satellite of multi-source satellite remote sensing data and the corresponding thematic product data produced by each industry department are extremely important work, which is the key to database construction in multi-source remote sensing data information service system, and is also the core part of further distinguishing the good and bad of service system and whether it can provide fast and convenient services for users.

[0004] Remote sensing data is huge and information is complex, and many data are invalid data. The traditional satellite remote sensing data technology and method cannot meet the performance requirements of basic query and visualized response data for these data. The current query system often has the problem that the query result cannot meet the user's demand. And when the user queries the data, he needs to input the formatted query statement according to the listed conditions to obtain the required data information. This method has small flexibility on the one hand, and each query needs to repeat a large number of formatted conditions, which leads to very inconvenient system use. On the other hand, when the data category is expanded, the query interface is too complex and difficult to understand intuitively, which seriously reduces the user experience and efficiency. SUMMARY

[0005] To solve the above technical problems, the application provides a remote sensing data query system, which forms available domain data by judging the availability quality of image data, and creates a hybrid index of image data, metadata, available domain data and label data, can support users to query natural sentences, can accurately query the required data of the user according to the coding index table, and can provide available domain data for the user to export, improves the utilization rate of remote sensing data, and the query result is accurate, the user operation is convenient and clear, improves the user experience and query efficiency.

[0006] To achieve the above technical purposes, the application provides a remote sensing data query system, which comprises:

[0007] A data acquisition module configured to acquire remote sensing data, including image data and metadata;

[0008] A data processing module configured to process the image data, obtain available domain data and label data based on the image data, the available domain data refers to data composed of regions with available performance reaching preset requirements in the image data, and the label data refers to data composed of labels carried by each pixel in the available domain data;

[0009] An index creation module configured to create an index according to the image data, metadata, available domain data and label data, and obtain a coding index table, the coding index table at least includes the following contents: identification name of the metadata, coding of the image data, and associated coding of the available domain data and the label data;

[0010] A data query module configured to query according to the coding index table, and obtain a query result.

[0011] In an embodiment of the application, the index creation module comprises:

[0012] A first creation unit configured to create dynamic coding of the image data, content identification coding of the metadata, and associated coding of the internal relationship of the available domain data and the label data;

[0013] A second creation unit configured to create a coding index table for establishing a hybrid index of remote sensing data according to the dynamic coding of the image data, the content identification coding of the metadata, and the relationship network coding of the internal relationship of the available domain data and the label data.

[0014] In an embodiment of the application, the first creation unit comprises:

[0015] An image data coding part configured to construct a pyramid according to the image data, and create dynamic coding for the pyramid according to the distribution of the image data;

[0016] The identification coding part of the metadata is configured to parse and extract the content of the metadata, and to textually encode the name of the metadata and the corresponding content;

[0017] The association coding part is configured to divide the available domain data according to a preset division mode to obtain a plurality of grids, encode the grids according to the coordinate information corresponding to the grids, and establish the association between the grids and the corresponding encoding and label data.

[0018] In an embodiment of the present application, the creation of the encoding index table for establishing the hybrid index of remote sensing data includes:

[0019] The field name and data type of the encoding index table are determined, and the encoding index table at least includes the following fields: the identification of the image data, the identification of each content in the metadata, and the relationship description between the available domain data and the label data.

[0020] According to the data corresponding to the field name and data type, the data is imported into the encoding index table.

[0021] According to the dynamic encoding of the image data, the content identification encoding of the metadata, and the relationship network encoding of the internal relationship between the available domain data and the label data, the encoding index of the image data, the encoding index of the metadata, and the relationship encoding index of the available domain data and the label data are established.

[0022] The three indexes are formed into a dictionary library based on the encoding results, and are written into the encoding index table.

[0023] In an embodiment of the present application, the data query module includes:

[0024] The data analysis unit is configured to analyze and extract the query information of the user, and to obtain the key word and the associated word related to the key word.

[0025] The data generation unit is configured to query and match the key word and the associated word, to form the query option and the expanded query option.

[0026] The data query unit is configured to perform a first query according to the query option to obtain a first data set, and to perform a second query according to the expanded query option to obtain a second data set, and to combine the first data set and the second data set as the query result.

[0027] In an embodiment of the present application, the key word and the associated word are selected from the dictionary library.

[0028] In an embodiment of the present application, the query matching refers to querying the corresponding fields according to the keywords in the coding index table, forming the query options of the fields corresponding to the keywords, and querying the corresponding fields according to the associated words in the coding index table, forming the augmented query options of the fields corresponding to the associated words.

[0029] In an embodiment of the present application, the one-to-one correspondence between the available domain data and the label data is based on pixels.

[0030] In an embodiment of the present application, the data processing module is further configured to determine the availability quality of the image data, extract the pixels meeting the availability quality requirements in the image data to form the available domain data.

[0031] In an embodiment of the present application, the image data includes multi-structure data, and the data processing module is further configured to perform format conversion on the image data to form image data in a unified format.

[0032] Compared with the prior art, the present application has the following beneficial effects:

[0033] (1) The system forms the available domain data based on the availability quality of the image data, and the available domain data corresponds to the label data, the obtained query result includes the available domain data for the user to download, which can improve the utilization rate of remote sensing data, and the query result can also include the original image data according to the user selection, improving the user experience.

[0034] (2) The coding index table of the present application includes the identification of the image data, the identification of each content in the metadata, the relationship description of the available domain data and the label data, which can support the user to input natural sentences for query, form the keywords and associated words by analyzing the natural sentences in the dictionary library of the coding index table, form the query options and augmented query options, and query in the coding index table through the query options and augmented query options to obtain the query result. The present application reduces the complexity of user operation, improves the accuracy of the query result through multiple queries, and further improves the user satisfaction.

[0035] (3) When querying, the present application can also mine the query information by automatic analysis to form additional query options, increase the accuracy of the query result, and provide a fine screening function, so that the user can select the required data in the query result, improving the query efficiency. BRIEF DESCRIPTION OF DRAWINGS

[0036] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings needed in the embodiments. Obviously, the drawings described below only illustrate some of the embodiments of the present application, and for those skilled in the art, other drawings can also be obtained from these drawings without creative effort.

[0037] Figure 1 The structure diagram of the query system of the embodiment of the present application;

[0038] Figure 2 The structure diagram of the identification coding part of the metadata of the embodiment of the present application;

[0039] Figure 3 The structure diagram of the data query module of the embodiment of the present application. DETAILED DESCRIPTION

[0040] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only some of the embodiments of the present application, but not all the embodiments. It should be noted that, as long as there is no conflict, each embodiment in the present application and each feature of each embodiment can be combined with each other, and the formed technical solutions are all within the protection scope of the present application.

[0041] It should be noted that the terms used herein are only for describing specific embodiments, and are not intended to limit the exemplary embodiments according to the present application. As used herein, the singular form is intended to include the plural form, unless the context clearly indicates otherwise, and it should also be understood that, when the terms "comprise" and / or "include" are used in the specification, there is a feature, step, operation, device, component and / or combination thereof.

[0042] Referring to Figure 1 The present application provides a query system for remote sensing data, which comprises:

[0043] A data acquisition module configured to acquire remote sensing data, including image data and metadata.

[0044] Specifically, the module is used to acquire a large amount of satellite remote sensing data from different sources. The obtained remote sensing data all include image data and metadata. The image data refers to satellite remote sensing whole scene image, which is data expressing various types of remote sensing features of an area in the form of image. The metadata is a description of the content, type, representation method, spatial reference, management method and other characteristics of the image data, which is an important means to realize remote sensing data management and sharing.

[0045] The data processing module is configured to process the image data, obtain available domain data and label data based on the image data, the available domain data refers to data composed of regions in the image data with available performance reaching preset requirements, and the label data refers to data composed of labels carried by each pixel in the available domain data.

[0046] Specifically, the data processing module is further configured to determine the availability quality of the image data, and extract pixels in the image data meeting the availability quality requirement to form the available domain data.

[0047] Specifically, the image data includes multi-structure data, and the data processing module is further configured to perform format conversion on the image data to form image data in a uniform format.

[0048] In the data processing module, the input image data is extracted to form the available domain data, specifically as follows:

[0049] The extracted available domain data in the module includes available domains of various quality detection items.

[0050] First, quality indicators of various quality detection items of the image data are determined, for example, the quality detection items include cloud detection, shadow detection, edge detection, etc., and each quality detection item is provided with a corresponding indicator.

[0051] Then, the image data is subjected to quality detection of the quality detection items, and the image data is subjected to quality evaluation according to the quality indicators, and whether the quality detection result of each quality detection item reaches the corresponding quality indicator is taken as implicit information, the implicit information is in units of pixels, that is, each pixel contains implicit information of its quality detection result, and the implicit information is arranged as label data.

[0052] Meanwhile, the image data can also be subjected to availability determination according to the quality evaluation results of all the quality detection items, that is, the quality detection result of the quality detection item of the pixel represented in the implicit information reaches the corresponding quality indicator, which is taken as the determination basis of availability, then the pixel is extracted and included in the available domain data of the quality detection item, and the available domain data of all the quality detection items is arranged to form the available domain data of the image data.

[0053] It needs to be clear that the available domain data can be saved in the form of a data table, for example, the available domain data is specifically an availability quality table, after the availability of the image data is determined for each quality detection item, the available image data is formed, which specifically refers to the determination result of whether the scene image data is available on the quality detection item. It includes the available image data corresponding to all quality detection items, then the header field of the availability quality table is determined, the name of each quality detection item is taken as the table header, and the corresponding available image data is imported into the availability quality table to form the final available domain data.

[0054] It needs to be noted that after the quality detection of the image data is performed on the quality detection item, the quality is evaluated according to the quality index, and the evaluation result is implicit information, from which the label data can be obtained, and in the implicit information, the pixels reaching the quality index are extracted in units of pixels, and then the available domain data is formed, therefore, the implicit and corresponding relationship between the available domain data and the label data can be clear, and the internal relationship between the available domain data and the label data is associated and expressed, so that the label data can be used to assist the establishment of subsequent indexing of the available domain data.

[0055] An index creation module is configured to create an index according to the image data, the metadata, the available domain data and the label data, and obtain an encoding index table, the encoding index table at least includes the following contents: the identification name of the metadata, the image data encoding, the association encoding of the available domain data and the label data.

[0056] Specifically, the index creation module includes:

[0057] A first creation unit is configured to create a dynamic encoding of the image data, a content identification encoding of the metadata, and an association encoding of the internal relationship of the available domain data and the label data.

[0058] Specifically, the first creation unit includes:

[0059] An image data encoding part is configured to construct a pyramid according to the image data, and create a dynamic encoding of the pyramid according to the distribution of the image data.

[0060] The specific steps of this part for encoding the image data are as follows:

[0061] Step one, construct a multi-source image data composite pyramid model for describing the multi-source, multi-scale and multi-temporal characteristics of multi-source image data.

[0062] In the implementation of the step, by adding sensor and time information in the traditional pyramid structure, the multi-source image data composite pyramid can quickly realize the local update of different sensors data at different time, reduce the difficulty of multi-source image data maintenance, and improve the application value of multi-source image data.

[0063] The traditional pyramid structure mainly refers to arranging an image set into a pyramid shape, gradually changing the resolution. The bottom of the pyramid is the original image, which has relatively high resolution, and the resolution and size of the pyramid decrease successively upwards, and the image resolution at the top is the lowest.

[0064] The composite pyramid in the embodiment is still constructed in a pyramid structure in a decreasing manner from bottom to top after adding sensor and time information.

[0065] Step two, dynamically encode the multi-source image data composite pyramid model structure according to the different distribution of multi-source image data. To express the longitude, latitude, elevation and other geographic information of image data.

[0066] Specifically, in the process of generating the multi-source image data composite pyramid structure, a data compression algorithm based on entropy coding is used to improve the storage and transmission efficiency of multi-source image data.

[0067] The identification coding part of the metadata is configured to parse and extract the content of the metadata, and text encode the name and corresponding content of the metadata.

[0068] Please refer to Figure 2 The specific steps of encoding the metadata are as follows:

[0069] Step one, metadata content information decomposition. According to the geographic information metadata standard, the metadata is layered into multiple metadata items, and the metadata items are further decomposed into structures associated with data and data names. The specific steps are as follows:

[0070] (1) Data standard layering. According to the geographic information metadata standard, the metadata is layered, and the metadata is decomposed into multiple metadata items of different data types, such as data output format, production date, cloud coverage, grid size, etc. Each object is a separate whole.

[0071] (2) Data object decomposition. As shown in Figure 2 , the metadata is decomposed into multiple metadata items, each of which is further decomposed into a set composed of data and data names, and the data and data names form a structure with an associated relationship.

[0072] Step two, metadata content information coding. The metadata information is classified as subject and object, and the subject corresponds to the data name and the object corresponds to the data. The metadata information is coded in a standard form.

[0073] Step three, in the geometric algebra space, the metadata information can be represented as a high-dimensional element, and the specific steps are as follows:

[0074] (1) In the geometric algebra space, it is composed of space base vectors and geometric operators, wherein the space base vectors are vector structures of any dimension, and the geometric product operation is the core of supporting vector expression and algorithm construction in the space.

[0075] (2) The expression of specific information relationship can be carried out in the geometric algebra space. The metadata information can be represented as a vector of different dimensions in the geometric algebra space. To correspond to the subject and object of the metadata, the concepts of vector subject and object are introduced. The subject of the vector, i.e. the coefficient of the vector, corresponds to the subject of the metadata; the object of the vector, i.e. the base of the vector, corresponds to the object of the metadata.

[0076] Step four, the data and data name of the metadata item are connected by geometric product, and different metadata items are connected by multiple vectors. The specific steps are as follows:

[0077] (1) The data and data name of the metadata item are different dimension vectors, and are connected by geometric product.

[0078] (2) Different metadata items are connected by multiple vectors, and the metadata can be coded as a linear combination of vectors of different dimensions.

[0079] The association coding part is configured to divide the available domain data according to a preset division method to obtain a plurality of grids, encode the coordinates corresponding to the grids, and establish the association relationship between the grids and the corresponding coding and label data.

[0080] The specific steps of this part for coding the association relationship are as follows:

[0081] Step one, establish a grid model based on available domain data, and preliminarily divide the available domain data. The division method can be to establish the relationship between the available domain data and the grid model according to the spatial resolution, geographical position and other information of the available domain data in the pyramid type hierarchical structure, and generate the grid according to the corresponding relationship.

[0082] Step two, for the available domain data grid, extract the pixel points contained in each grid, since the available domain data and the label data are in a one-to-one correspondence at the pixel level, that is, each pixel point in the available domain data corresponds to a label in the label data, then introduce the label data into the grid model of the available domain data, classify the pixel points according to the label type, and then redivide the available domain data grid. The classification of the pixel points in this step can be realized by a classification network, such as an SVM classification network.

[0083] Step three, determine the encoding string set of the available domain data under the grid model, and the encoding string set includes a plurality of encoding strings.

[0084] Specifically, based on the Hilbert curve, a plurality of encoding strings of the available domain data in each grid in the grid model are obtained, and all the encoding strings constitute the encoding string set.

[0085] The second creating unit is configured to create an encoding index table for establishing a hybrid index of remote sensing data according to the dynamic encoding of image data, the content identification encoding of metadata, and the relationship network encoding of the internal relationship between available domain data and label data.

[0086] Specifically, the creation process of the encoding index table is as follows:

[0087] First step, determine the field name and data type of the encoding index table, and the encoding index table at least includes the following fields: identification of image data, identification of each content in metadata, and relationship description of available domain data and label data.

[0088] Among them, the identification of image data, the identification of each content in metadata, and the relationship description of available domain data and label data can be correspondingly extracted from the dynamic encoding of image data, the content identification encoding of metadata, and the relationship network encoding of the internal relationship between available domain data and label data,

[0089] Second step, according to the data corresponding to each field name and data type, import it into the encoding index table.

[0090] Specifically, the data includes available domain data, image data, metadata and label data, and the data can be stored in a data collection table which stores all the data imported into the system.

[0091] Third step, respectively according to the dynamic encoding of image data, the content identification encoding of metadata, and the relationship network encoding of the internal relationship between available domain data and label data, establish the encoding index of image data, the encoding index of metadata, and the relationship encoding index of available domain data and label data.

[0092] Specifically, the three indexes are established based on the dynamic coding of the image data according to a non-uniform pyramid index algorithm, the one-dimensional coding index of the metadata is established based on the content identification coding of the metadata, and the two-dimensional grid coding index is established based on the relationship network coding of the internal relationship between the available domain data and the label data.

[0093] In the fourth step, the three indexes are formed into a dictionary library based on the coding results, and the dictionary library is written into a coding index table.

[0094] Specifically, the dictionary library stores the coding results in the form of characters, and establishes the association between the content in the dictionary library and the data and indexes when the dictionary library is written into the coding index table.

[0095] A data query module is configured to query according to the index table to obtain a query result.

[0096] Specifically, the data query module is further configured to query the remote sensing data according to the coding index table, to perform a first-level filtering on the preliminary query result, and to perform a second-level filtering on the preliminary query result in combination with the result of the first-level filtering, to obtain the query result.

[0097] Please refer to Figure 3 , the data query module comprises:

[0098] A data analysis unit is configured to analyze and extract the query information of a user, to obtain a keyword and an associated word related to the keyword.

[0099] A data generation unit is configured to query and match the keyword and the associated word, to form a query option and an expanded query option.

[0100] A data query unit is configured to perform a first query according to the query option, to obtain a first data set, and to perform a second query according to the expanded query option, to obtain a second data set, and to combine the first data set and the second data set as a query result.

[0101] The data query module of the above embodiment is described in detail as follows:

[0102] The data query module of the present application can specifically be provided with a data analysis unit, a data generation unit, and a data query unit.

[0103] (1) Data analysis unit

[0104] The unit is configured to obtain the query information of a user, to analyze and extract the query information of the user, to perform word segmentation on the query information according to a word segmentation rule, to obtain a keyword and an associated word related to the keyword.

[0105] Specifically, the word segmentation rule refers to a word segmentation symbol, and the query information refers to a natural language sentence. After the query information is segmented, the keywords in the dictionary library are read, and the associated words are formed according to the keyword heat weight and relevance.

[0106] (2) Data generating unit

[0107] The unit is configured to perform query matching according to the keywords and the associated words, to obtain query options corresponding to the keywords, and to obtain expanded query options corresponding to the associated words.

[0108] Specifically, the query matching refers to querying the corresponding fields in the encoding index table according to the keywords, forming query options corresponding to the keywords, and querying the corresponding fields in the encoding index table according to the associated words, forming expanded query options corresponding to the associated words.

[0109] (3) Data query unit

[0110] In an embodiment of the application, the unit is configured to perform a first query on the data in the encoding index table according to the query options. The first query result obtained by the first query is based on the analysis of the query information of the user. Then, prompt information is generated for the first query result, which is used to remind the user to check whether the first query result meets the user's expectation. If the user selects yes, the query is ended, the final query result is generated and exported. If the user selects no, a second query is entered, which is a second query on the data in the encoding index table according to the expanded query options. The result obtained by the second query is the second query result after the first query result is removed. Then, the second query result is analyzed, and second prompt information is generated. The user selects whether to be satisfied with the query result according to the prompt information. If the user selects yes, the query is ended, the final query result is generated and exported. If the user selects no, the user is automatically prompted to input new query information, and the data analysis unit, the data generating unit and the above process are re-executed until the user is satisfied with the query result.

[0111] In another embodiment of the present application, the unit is configured to make a first query on the data in the encoding index table according to the query option to obtain a first data set, make a second query on the data in the encoding index table according to the augmented query option to obtain a second data set, list the first data set and the second data set in a query list; in addition, the unit analyzes and mines the query option and the augmented query option to generate an additional query option, makes a third query on the data in the encoding index table according to the additional query option to obtain a third data set, also lists the third data set in the query list, removes the duplicate data in the query list to form a preliminary query result. A prompt information is generated for the preliminary query result, and the user selects whether to be satisfied with the preliminary query result according to the prompt information; if the user selects yes, the query is ended, and the query result is exported; if the user selects no, the user is automatically prompted to input new query information, and the data analysis unit, the data generation unit and the above process are re-executed until the user is satisfied with the query result.

[0112] It should be noted that when the data is exported after the user is satisfied with the query result, an export option is generated, and the user can select the exported data according to the need. The export option includes image data and available domain data. It should be clear that the metadata information is included in all the export options, and the tag information is included in the available domain data.

[0113] The above is only a specific embodiment of the present application, but the protection scope of the present application is not limited to this. Any person skilled in the art can easily think of changes or replacements within the technical range disclosed by the present application, which should be covered within the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A remote sensing data query system, characterized in that, The system includes: The data acquisition module is configured to acquire remote sensing data, including image data and metadata; The data processing module is configured to process image data and obtain usable domain data and label data based on the image data. The usable domain data refers to the data composed of areas in the image data whose usability performance meets preset requirements. The label data refers to the data assigned to each pixel to describe the quality assessment result of the pixel after quality detection of the image data and quality assessment of the image data according to quality indicators. The usable domain data and the label data have a one-to-one correspondence based on pixels. The index creation module is configured to create an index based on image data, metadata, availability domain data, and tag data, and obtain an encoded index table. The encoded index table includes at least the following: the identifier name of the metadata, the image data encoding, and the association encoding between the availability domain data and the tag data. The data query module is configured to perform a query based on the encoding index table to obtain query results, which include image data and available domain data. The data processing module is further configured to determine the usability quality of the image data and extract the pixels in the image data that meet the usability quality requirements to form usability domain data. The index creation module includes: The first creation unit is configured to create dynamic encoding of image data, content identification encoding of metadata, and association encoding of the intrinsic relationship between available domain data and tag data; The second creation unit is configured to create a coded index table for building a hybrid index of remote sensing data based on the dynamic coding of image data, the content identification coding of metadata, and the relationship network coding of the inherent relationship between available domain data and label data. The first creation unit includes: The image data encoding part is configured to construct a pyramid based on the image data and create dynamic encoding for the pyramid according to the distribution of the image data. The metadata identification and encoding part is configured to parse and extract the metadata, and to text-encode the metadata name and its corresponding content. The associated coding section is configured to divide the available domain data into multiple grids according to a preset partitioning method, encode the grids according to the coordinate information corresponding to the grids, and establish the association between the grids, their corresponding codes, and the label data.

2. The system according to claim 1, characterized in that, The creation of the encoded index table for building a hybrid index of remote sensing data includes: Determine the field names and data types of the coding index table. The coding index table should include at least the following fields: the identifier of the image data, the identifier of each content in the metadata, and a description of the relationship between the available domain data and the tag data. Based on the data corresponding to each field name and data type, import them into the coding index table; Based on the dynamic encoding of image data, the content identification encoding of metadata, and the relationship network encoding of the inherent relationship between available domain data and tag data, respectively, establish the encoding index of image data, the encoding index of metadata, and the relationship encoding index of available domain data and tag data; The three indexes are used to form a dictionary based on the encoding results and written into the encoding index table.

3. The system according to claim 2, characterized in that, The data query module includes: The data parsing unit is configured to parse and extract user query information to obtain keywords and related terms. The data generation unit is configured to perform query matching on keywords and related terms to generate query options and expand query options; The data query unit is configured to perform a first query based on the query options to obtain a first dataset, and then perform a second query based on the expanded query options to obtain a second dataset, and finally merge the first dataset and the second dataset as the query result.

4. The system according to claim 3, characterized in that, The keywords and related terms are obtained by filtering from the dictionary database.

5. The system according to claim 3, characterized in that, The query matching refers to querying the corresponding fields in the coding index table based on keywords, forming query options from the fields corresponding to the keywords, and simultaneously querying the corresponding fields in the coding index table based on related words, forming expanded query options from the fields corresponding to the fields corresponding to the related words.

6. The system according to claim 1, characterized in that, The image data includes multi-structured data, and the data processing module is further configured to perform format conversion on the image data to form image data in a unified format.

Images