A collection system of archival metadata without description

Abstract

The invention provides an archive metadata free-recording collection system, belonging to the technical field of archive management. The archive metadata free-recording collection system comprises a classification scheme management module, a collection and analysis module, a data storage and reading module, an archive management module, a statistical analysis module, an identification and disposal module and a data scheme management module; the classification scheme management module is used for managing classification schemes of archive materials of various categories; the collection and analysis module is used for collecting original metadata free of recording and decoupling; the data storage and reading module is used for retrieving and analyzing original metadata to obtain usable metadata and storing usable metadata; the archive management module is used for archiving and arranging original metadata and usable metadata and performing completeness and standardization inspection; the statistical analysis module is used for counting usable metadata; the identification and disposal module is used for identifying and disposing usable metadata; the invention can solve the problem of inconvenience in archiving caused by inconsistent terms in archive metadata.

Description

Technical Field

[0001] The present invention belongs to the technical field of archive management, and in particular, relates to an archive metadata free-recording collection system. Background Art

[0002] The Chinese invention patent with publication number CN113112229A (application number: CN202110398874.7) proposes an intelligent archive organization system and method based on the pre-positioning of archive elements, including a file model module, a business module, an archive handover module and an archive module; in the process of archive file formation, the system automatically captures the electronic file metadata of the entire life cycle of the archive file through various modules to form a node file, thereby ensuring the authenticity and integrity of the archive file, and realizing the automatic collection, organization and archiving of the archive file, greatly reducing the cost of archive management.

[0003] The Chinese invention patent with publication number CN107633037A (application number: CN201710813230.3) proposes a file arrangement system, including a centralized storage module, a rights management module and a workflow module. The centralized storage module includes data backup, scanned files and cloud upload, which are all installed in the centralized storage module. The rights management module includes ordinary rights and management rights, which are all installed in the rights management module. The workflow module includes version management functions, automated office and search functions, which are all installed in the workflow module. The invention realizes the basic creation, modification, storage, transmission, version control, approval process, query and repeated use, deletion and recovery, data backup and other functions of files. It meets the needs of team-oriented document collaboration and knowledge management.

[0004] In different types of archives, the descriptions of the same content are often similar in meaning or the same but with different terms. The above inventions cannot solve the problem of inconsistent terms in archive metadata, which makes it inconvenient to organize and archive. Summary of the invention

[0005] In view of this, the present invention proposes a system for collecting archive metadata without recording, which can solve the problem of inconsistent terms in archive metadata and facilitate organization and archiving.

[0006] The present invention is achieved in that:

[0007] The present invention provides an archive metadata free-recording collection system, which includes a classification scheme management module, a collection and analysis module, a data storage and reading module, an archive management module, a statistical analysis module, an identification and disposal module, and a data scheme management module;

[0008] The classification scheme management module is used to manage the classification schemes of various categories of archive materials, and supports navigation classification of archive categories, years, institutions or issues, and storage periods;

[0009] The collection and analysis module is used to collect the original metadata in the electronic files, traditional carrier archive files or traditional archive digital copies without recording, and decouple the content of the collected original metadata;

[0010] The data storage and reading module is used to retrieve and analyze the original metadata to obtain usable metadata, and store the original metadata and the usable metadata;

[0011] The archive management module is used to archive and organize the original metadata and the available metadata, and to check the completeness and standardization of the archive directory and metadata items, and to guide the archiving process and automatically generate an inspection report, and to check the completeness and accuracy of the archive content according to the workflow, and to automatically generate an inspection report and remind the archive inspection content;

[0012] The statistical analysis module is used to perform statistics on the available metadata, including the resource room storage, archive utilization, data sources and quadrants;

[0013] The identification and disposal module is used to identify and dispose of the available metadata, and to perform statistical management on various categories of archival materials and related business conditions;

[0014] The data scheme management module is used to customize the available metadata and provide metadata scheme management, and configure metadata schemes for each category according to the archive category, and support validity verification of metadata.

[0015] The technical effects of the archive metadata free-description collection system provided by the present invention are as follows: by decoupling the original metadata and utilizing a method of separating directory contents, the strong correlation between files and directories in the original metadata is reduced, thereby facilitating unified and efficient management of the archive metadata. Through the set acquisition and analysis module, centralized, unified and comprehensive management of archive materials of different types and formats can be achieved; and by utilizing the data storage and reading module, when collecting and reading archive metadata, automatic rapid identification and conversion from the original metadata can be achieved to form usable metadata, unify the terms in the archive metadata, improve the quality of archive metadata collection, and improve reading efficiency.

[0016] On the basis of the above technical solution, the archival metadata free-description collection system of the present invention can also be improved as follows:

[0017] Wherein, the data storage and reading module includes a metadata module and a data module;

[0018] The data module caches the original metadata on the metadata module, and a plurality of the metadata modules form a metadata cluster. The metadata modules in the metadata cluster each store a portion of the original metadata, and each portion of the original metadata is different from each other.

[0019] The metadata module is used to retrieve the original metadata in the internal cache to obtain the available metadata and feed it back to the user;

[0020] The data module is used to retrieve the original metadata to obtain the available metadata and cache it in the metadata module; the data module is also used to perform similarity analysis on the original metadata to obtain the available metadata, store the available metadata and cache the available metadata in the metadata module.

[0021] Furthermore, the step of performing similarity analysis on the original metadata to obtain the available metadata is specifically as follows:

[0022] (1) Use the Word2Vec model to obtain word vectors for test text and standard text;

[0023] (2) calculating the word similarity value between the test text and the standard text according to the word vector;

[0024] (3) determining whether the word similarity value exceeds a set threshold,

[0025] If yes, the test text and the standard text are synonyms or near synonyms, and the standard text is used as available metadata;

[0026] If not, the test text is not a synonym or near synonym of the standard text, and a new standard text is selected to repeat the above steps.

[0027] Furthermore, the step of calculating the word similarity value between the test text and the standard text according to the word vector is specifically as follows:

[0028] The first step is to remove common stop words from the test text and the standard text;

[0029] The second step is to obtain the Pearson coefficient as the word similarity value based on the processed word vector of the test text and the word vector of the standard text. When the word similarity value exceeds a set threshold, the test text and the standard text are synonyms or synonyms.

[0030] Furthermore, the similarity is a synonymy degree, and the similarity is used to determine the synonymy or near-synonymy relationship between the original metadata and the words in the available metadata in combination with the synonymy degree of the words. The value range of the word similarity value is 0 to 1, and the set threshold is 0 to 0.6.

[0031] The acquisition and analysis module includes an image acquisition module, an image classification module, an OCR character detection and recognition module, an information matching module, a database module and a decoupling module;

[0032] The image acquisition module is used to collect video images in the audio and video files to be recorded and transmit them to the image classification module;

[0033] The image classification module is used to extract image features in the video and transmit them to the OCR character detection and recognition module;

[0034] The OCR character recognition module is used to frame the position of the text, identify the specific content of the text, and form the text information of the image;

[0035] The image classification module transmits the image features to the information matching module, and the OCR character detection and recognition module transmits the text information of the image to the information matching module, and the information matching module is used to match the image features with the text information of the image and store them in the database module to form original metadata;

[0036] The decoupling module is used to decouple the original metadata stored in the database module.

[0037] Furthermore, the decoupling of the original metadata stored in the database module specifically includes the following steps:

[0038] (1) Create a file or directory, and create all directory entries including the parent directory path of the file or directory at the node where the file or directory is created;

[0039] (2) If all or part of the directory entry has been created at the node, create the remaining directory entry and proceed to the next step; if the directory entry has not been created at the node, return to the previous step;

[0040] (3) separating the original metadata to obtain the original metadata of the directory, the original metadata of the directory item, and the original metadata of the file;

[0041] (4) setting the original metadata of the directory in the directory index node;

[0042] (5) Each of the directory entries is segmented according to the distribution of the original metadata, and the directory entries related thereto are stored in the node where the original metadata is stored, and a reverse index pointing to the original metadata of the directory is established.

[0043] Furthermore, the database module includes a mechanical hard disk and a storage control system.

[0044] The mechanical hard disk has a first storage area and a shingled magnetic recording SMR area;

[0045] The storage control system is used to place the original metadata and the available metadata stored in the shingled magnetic recording SMR area into the first storage area.

[0046] Furthermore, the OCR character detection and recognition module realizes the extraction of text features through a character detection and recognition algorithm; the character detection and recognition algorithm is one of the Faster-RCNN algorithm, the RRPN algorithm, the CRNN algorithm and the EAST algorithm.

[0047] Furthermore, the image classification module is respectively connected to the image acquisition module, the OCR character detection and recognition module, and the information matching module, the OCR character detection and recognition module is connected to the information matching module; and the decoupling module is connected to the database module.

[0048] Compared with the prior art, the beneficial effects of the archive metadata free-description collection system provided by the present invention are: by decoupling the original metadata and utilizing the method of separating the directory content, the strong correlation between files and directories in the original metadata is reduced, which facilitates the unified and efficient management of the archive metadata. Through the set collection and analysis module, centralized, unified and comprehensive management of archive materials of different types and formats can be achieved; and by utilizing the data storage and reading module, when collecting and reading archive metadata, automatic rapid identification and conversion from the original metadata can be achieved to form usable metadata, standardize the descriptive terms in the archive content, improve the quality of archive archiving, and improve the reading efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments of the present invention will be briefly introduced below. Obviously, the accompanying drawings in the following description are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative labor.

[0050] Figure 1 This is a structural diagram of a system for collecting archive metadata without recording proposed by the present invention; DETAILED DESCRIPTION

[0051] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0052] Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the invention claimed for protection, but merely represents selected embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0053] It should be noted that similar reference numerals and letters denote similar items in the following drawings, and therefore, once an item is defined in one drawing, further definition and explanation thereof is not required in subsequent drawings.

[0054] In the description of the present invention, it should be understood that the terms "center", "longitudinal", "lateral", "length", "width", "thickness", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside", "clockwise", "counterclockwise" and the like indicate orientations or positional relationships based on the orientations or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the referred device or element must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be understood as a limitation on the present invention.

[0055] like Figure 1 As shown, it is an architecture diagram of a system for collecting archive metadata without cataloging provided by the present invention. The system includes a classification scheme management module, a collection and analysis module, a data storage and reading module, an archive management module, a statistical analysis module, an identification and disposal module and a data scheme management module;

[0056] The classification scheme management module is used to manage the classification schemes of various categories of archival materials, and supports navigation classification of archival categories, years, institutions or issues, and storage periods;

[0057] The collection and analysis module is used to collect the original metadata in electronic files, traditional carrier archive files or traditional archive digital copies without recording, and decouple the content of the collected original metadata;

[0058] The data storage and reading module is used to retrieve and analyze the original metadata to obtain the usable metadata, and store the original metadata and the usable metadata;

[0059] The archive management module is used to archive and organize the original metadata and available metadata, and to check the completeness and standardization of the archive directory and metadata items, and to guide the archiving process and automatically generate inspection reports. It also checks the completeness and accuracy of the archive content according to the workflow, and automatically generates inspection reports and reminds you of the archive inspection content.

[0060] The statistical analysis module is used to count the available metadata, including the resource room collection, archive utilization, data sources and quadrants;

[0061] The identification and disposal module is used to identify and dispose of available metadata, and to perform statistical management on various categories of archival materials and related business situations;

[0062] The data scheme management module is used to customize the available metadata and provide metadata scheme management, configure metadata schemes for each category according to the archive category, and support the validity verification of metadata.

[0063] It should be noted that the classification scheme management module performs classification settings based on the categories of archives and classification scheme requirements. If the established classification scheme is adjusted, the adjusted classification scheme will be used for classification navigation in the classification cataloging module of the archives.

[0064] Among them, in the above technical solution, the data storage and reading module includes a metadata module and a data module;

[0065] The data module caches the original metadata on the metadata module. Multiple metadata modules form a metadata cluster. The metadata modules in the metadata cluster each store a portion of the original metadata, and each portion of the original metadata is different from each other.

[0066] The metadata module is used to retrieve the original metadata in the internal cache to obtain the available metadata and feed it back to the user;

[0067] The data module is used to retrieve the original metadata to obtain the available metadata and cache it in the metadata module; the data module is also used to perform similarity analysis on the original metadata, obtain the available metadata, store the available metadata and cache the available metadata in the metadata module.

[0068] When the user reads, the metadata module first retrieves the available metadata from the metadata cluster and provides feedback. If no available metadata is found in the metadata cluster, the metadata module sends a request to the data module. After receiving the request from the metadata, the data module scans the mechanical hard disk. When available metadata is obtained, the available metadata is cached to the data module, which then provides feedback. After receiving the request from the metadata, the data module scans the mechanical hard disk. When no available metadata is obtained, a similarity analysis is performed on the scanned original metadata, and available metadata is obtained and cached to the metadata module, which then provides feedback.

[0069] It should be noted that when the data module is started, it also performs a breadth traversal of the disk, ends the traversal after traversing two levels of directories, and sends the traversal results to the metadata module. After the available metadata in the traversal results reaches the metadata cluster, the metadata cluster groups and stores the available metadata in the traversal results and shares the storage of the directory structure.

[0070] When the memory occupied by the original metadata and the available metadata cache in the metadata module exceeds a threshold, the metadata module releases part of the memory to control the memory usage of the metadata module.

[0071] Furthermore, in the above technical solution, the steps of performing similarity analysis on the original metadata to obtain the usable metadata are specifically as follows:

[0072] (1) Use the Word2Vec model to obtain word vectors for test text and standard text;

[0073] (2) Calculate the word similarity value between the test text and the standard text based on the word vector;

[0074] (3) Determine whether the word similarity value exceeds the set threshold,

[0075] If yes, the test text and the standard text are synonyms or near-synonyms, and the standard text is used as available metadata;

[0076] If not, the test text is not a synonym or near synonym of the standard text, and a new standard text is selected to repeat the above steps.

[0077] Word2Vec is a method for converting text into vectors. It includes two algorithms: skip-gram and CBOW. Skip-gram predicts the words around the central word through the central word; CBOW predicts the central word through the surrounding words.

[0078] The specific operations of using the Word2Vec model to obtain the word vectors of the test text and the standard text are as follows:

[0079] Select a test text and a standard text, and select k keywords of the test text and k keywords of the standard text according to the TF-IDF weight of the words. Take these keywords as the dimensions of the vector space to establish the vector space. Count the number of times each keyword appears in the test text and the standard text to form the corresponding k-dimensional vector, which is the word vector of the test text and the standard text.

[0080] Among them, TF-IDF weight is often used in information retrieval and text mining as a statistical measure to evaluate the importance of a word to a document collection or a document in a corpus. This importance is proportional to the number of times the word appears in the document. The main idea of ​​TF-IDF is: if word A appears frequently in a document B and rarely appears in other documents, it is considered that word A has good distinguishing ability and can be used to distinguish document B from other documents.

[0081] Furthermore, in the above technical solution, the step of calculating the word similarity value of the test text and the standard text according to the word vector is specifically as follows:

[0082] The first step is to remove common stop words from the test text and standard text;

[0083] The second step is to obtain the Pearson coefficient as the word similarity value based on the word vector of the processed test text and the word vector of the standard text. When the word similarity value exceeds the set threshold, the test text and the standard text are synonyms or synonyms.

[0084] It should be noted that stop words are the most commonly used words in a particular language, and are words that do not affect readability even if removed;

[0085] The Pearson coefficient measures the linear correlation. If the Pearson coefficient is zero, the correlation between the test text and the standard text is the weakest, that is, the similarity is the smallest. When the absolute value of the Pearson coefficient is close to 1, the correlation is the strongest, that is, the similarity is the largest.

[0086] The specific operation of obtaining the Pearson coefficient based on the word vector of the processed test text and the word vector of the standard text is as follows:

[0087] Calculate the cosin value of the angle between the word vector of the test text and the word vector of the standard text, and take the absolute value of the cosin value as the Pearson coefficient.

[0088] Furthermore, in the above technical solution, the similarity is the degree of synonymy, and the similarity is used to determine the synonymy or near-synonymy relationship between the original metadata and the words in the available metadata in combination with the degree of synonymy of the words. The value range of the word similarity value is 0 to 1, and the set threshold is 0 to 0.6.

[0089] Among them, in the above technical solution, the acquisition and analysis module includes an image acquisition module, an image classification module, an OCR character detection and recognition module, an information matching module, a database module and a decoupling module;

[0090] The image acquisition module is used to collect video images in the audio and video files to be recorded and transmit them to the image classification module;

[0091] The image classification module is used to extract image features in the video and transmit them to the OCR character detection and recognition module;

[0092] The OCR character recognition module is used to frame the position of the text, identify the specific content of the text, and form the text information of the image;

[0093] The image classification module transmits the image features to the information matching module, and the OCR character detection and recognition module transmits the text information of the image to the information matching module. The information matching module is used to match the image features with the text information of the image and store them in the database module to form original metadata;

[0094] The decoupling module is used to decouple the original metadata stored in the database module.

[0095] Furthermore, in the above technical solution, decoupling the original metadata stored in the database module specifically includes the following steps:

[0096] (1) Create a file or directory, and create a directory entry containing all the parent directory paths of the file or directory at the node where the file or directory is created;

[0097] (2) If all or part of the directory entry has been created at the node, create the remaining directory entry and proceed to the next step. If the directory entry has not been created at the node, return to the previous step.

[0098] (3) Separating the original metadata to obtain the original metadata of the directory, the original metadata of the directory item, and the original metadata of the file;

[0099] (4) Set the original metadata of the directory in the directory index node;

[0100] (5) Each directory entry is segmented according to the distribution of the original metadata, and the directory entries related to it are stored in the node where the original metadata is stored, and a reverse index pointing to the original metadata of the directory is established.

[0101] Furthermore, in the above technical solution, the database module includes a mechanical hard disk and a storage control system.

[0102] The mechanical hard disk has a first storage area and a shingled magnetic recording SMR area;

[0103] The storage control system is used to place the original metadata and available metadata stored in the shingled magnetic recording SMR area into the first storage area.

[0104] Furthermore, in the above technical solution, the OCR character detection and recognition module realizes the extraction of text features through a character detection and recognition algorithm; the character detection and recognition algorithm is one of the Faster-RCNN algorithm, the RRPN algorithm, the CRNN algorithm and the EAST algorithm.

[0105] It should be noted that when using the Faster-RCNN algorithm for extraction, after inputting the test image, the selective search algorithm first extracts about 2,000 suggestion windows from top to bottom in the image, and then the entire image is input into CNN for feature extraction; the suggestion windows are then mapped to the last layer of convolutional feature maps of CNN; the RoI pooling layer is then used to generate a fixed-size f feature map for each suggestion window; finally, the classification probability and bounding box regression are jointly trained by Softmax Loss and Smooth L1 Loss;

[0106] When using the RRPN algorithm for extraction, a convolutional network is first used to generate a feature map, and then RRPN is used to generate a candidate region with an angle. Finally, RoI pooling is used to generate a feature vector of fixed length, followed by two layers of full connection for fine-tuning the category of the candidate region.

[0107] When using the CRNN algorithm for extraction, the main network structure used is CNN+RNN+CTC. From bottom to top, they are: convolution layer, using CNN, extracting feature sequences from the input image; recurrent layer, using RNN, predicting the label distribution of the feature sequence obtained from the convolution layer; transcription layer, using CTC, converting the label distribution obtained from the recurrent layer into the final recognition result through operations such as de-duplication and integration;

[0108] When using the EAST algorithm for extraction, the image is first sent to the FCN network structure and a single-channel pixel-level text score feature map and a multi-channel geometric feature map are generated. The text area uses two geometric shapes, RBOX and QUAD, and a different loss function is designed for each geometric shape; then, a threshold is applied to each predicted area, and those with a score exceeding the predetermined threshold are considered valid and saved.

[0109] Furthermore, in the above technical solution, the image classification module is respectively connected to the image acquisition module, the OCR character detection and recognition module, and the information matching module, the OCR character detection and recognition module is connected to the information matching module; and the decoupling module is connected to the database module.

[0110] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims

1. A system for collecting archive metadata without recording, characterized in that: It includes classification scheme management module, collection and analysis module, data storage and reading module, archive management module, statistical analysis module, identification and disposal module and data scheme management module; The classification scheme management module is used to manage the classification schemes of various categories of archive materials, and supports navigation classification of archive categories, years, institutions or issues, and storage periods; The collection and analysis module is used to collect the original metadata in the electronic files, traditional carrier archive files or traditional archive digital copies without recording, and decouple the content of the collected original metadata; The data storage and reading module is used to retrieve and analyze the original metadata to obtain usable metadata, and store the original metadata and the usable metadata; The archive management module is used to archive and organize the original metadata and the available metadata, and to check the completeness and standardization of the archive directory and metadata items, and to guide the archiving process and automatically generate an inspection report, and to check the completeness and accuracy of the archive content according to the workflow, and to automatically generate an inspection report and remind the archive inspection content; The statistical analysis module is used to perform statistics on the available metadata, including the resource room storage, archive utilization, data sources and quadrants; The identification and disposal module is used to identify and dispose of the available metadata, and to perform statistical management on various categories of archival materials and related business conditions; The data scheme management module is used to customize the available metadata, provide metadata scheme management, configure metadata schemes for each category according to the archive category, and support metadata validity verification; Wherein, the data storage and reading module includes a metadata module and a data module; The data module caches the original metadata on the metadata module, and a plurality of the metadata modules form a metadata cluster. The metadata modules in the metadata cluster each store a portion of the original metadata, and each portion of the original metadata is different from each other; The metadata module is used to retrieve the original metadata in the internal cache to obtain the available metadata and feed it back to the user; The data module is used to retrieve the original metadata to obtain the available metadata and cache it in the metadata module; the data module is also used to perform similarity analysis on the original metadata to obtain the available metadata, store the available metadata and cache the available metadata in the metadata module.

2. The archival metadata collection system without recording according to claim 1 is characterized in that: The step of performing similarity analysis on the original metadata to obtain the available metadata is specifically as follows: (1) Use the Word2Vec model to obtain word vectors for test text and standard text; (2) calculating the word similarity value between the test text and the standard text according to the word vector; (3) Determine whether the word similarity value exceeds a set threshold, If yes, the test text and the standard text are synonyms or near synonyms, and the standard text is used as available metadata; If not, the test text is not a synonym or near synonym of the standard text, and a new standard text is selected to repeat the above steps.

3. The archival metadata collection system without recording according to claim 2 is characterized in that: The step of calculating the word similarity value of the test text and the standard text according to the word vector is specifically as follows: The first step is to remove common stop words from the test text and the standard text; The second step is to obtain the Pearson coefficient as the word similarity value based on the processed word vector of the test text and the word vector of the standard text. When the word similarity value exceeds a set threshold, the test text and the standard text are synonyms or synonyms.

4. The archival metadata collection system without recording according to claim 3 is characterized in that: The similarity is the degree of synonymy, and the similarity is used to determine the synonymy or near-synonymy relationship between the original metadata and the words in the available metadata in combination with the degree of synonymy of the words. The value range of the word similarity value is 0-1, and the set threshold is 0-0.

6.

5. The archival metadata collection system without recording according to claim 4 is characterized in that: The acquisition and analysis module includes an image acquisition module, an image classification module, an OCR character detection and recognition module, an information matching module, a database module and a decoupling module; The image acquisition module is used to collect video images in the audio and video files to be recorded and transmit them to the image classification module; The image classification module is used to extract image features in the video and transmit them to the OCR character detection and recognition module; The OCR character detection and recognition module is used to frame the position of the text, recognize the specific content of the text, and form the text information of the image; The image classification module transmits the image features to the information matching module, and the OCR character detection and recognition module transmits the text information of the image to the information matching module, and the information matching module is used to match the image features with the text information of the image and store them in the database module to form original metadata; The decoupling module is used to decouple the original metadata stored in the database module.

6. The archival metadata collection system without recording according to claim 5 is characterized in that: The decoupling of the original metadata stored in the database module specifically includes the following steps: (1) Create a file or directory, and create a directory entry containing all the parent directory paths of the file or directory at the node where the file or directory is created; (2) If all or part of the directory entry has been created at the node, create the remaining directory entry and proceed to the next step; if the directory entry has not been created at the node, return to the previous step; (3) separating the original metadata to obtain the original metadata of the directory, the original metadata of the directory item, and the original metadata of the file; (4) Setting the original metadata of the directory in the directory index node; (5) Each of the directory entries is segmented according to the distribution of the original metadata, and the directory entries related thereto are stored in the node where the original metadata is stored, and a reverse index pointing to the original metadata of the directory is established.

7. The archival metadata collection system without recording according to claim 5 is characterized in that: The database module includes a mechanical hard disk and a storage control system. The mechanical hard disk has a first storage area and a shingled magnetic recording SMR area; The storage control system is used to place the original metadata and the available metadata stored in the shingled magnetic recording SMR area into the first storage area.

8. The archival metadata collection system without recording according to claim 5 is characterized in that: The OCR character detection and recognition module realizes the extraction of text features through a character detection and recognition algorithm; the character detection and recognition algorithm is one of the Faster-RCNN algorithm, the RRPN algorithm, the CRNN algorithm and the EAST algorithm.

9. The archival metadata collection system without recording according to claim 5 is characterized in that: The image classification module is connected to the image acquisition module, the OCR character detection and recognition module, and the information matching module respectively; the OCR character detection and recognition module is connected to the information matching module; and the decoupling module is connected to the database module.

Images