Lightweight digital archive management method and system for small and micro non-permitted projects

Abstract

The invention provides a lightweight digital archive management method and system for small and micro non-abandoned projects. The mobile terminal is utilized to conveniently collect information such as non-abandoned objects, performances and inheritors' spokes, and metadata is automatically added. Through preprocessing such as cutting and denoising, key information is extracted by using a lightweight identification technology; distributed storage is combined with block chain encryption to guarantee data security, multi-mode retrieval of keywords, categories and the like is supported, and results are visually presented. Archives can be updated through a mobile terminal or a computer terminal, and the system records logs and periodically backs up for testing. According to the method, the current situation of small and micro non-abandoned resources is met through lightweight design, efficient and convenient management is achieved, data reliability is ensured, non-abandoned inheritance development is powerfully promoted, and a low-cost and high-availability innovative scheme is provided for small and micro non-abandoned project digital archive management.

Description

Technical Field

[0001] The present invention relates to the field of digital archive management, and specifically to a lightweight digital archive management method and system for small and micro intangible cultural heritage projects. Background Art

[0002] Intangible cultural heritage is a treasure of human civilization, carrying rich historical, cultural, and artistic value. As an integral part of this heritage, small and micro-scale intangible cultural heritage projects play an irreplaceable role in preserving regional cultural characteristics and enriching cultural diversity. However, due to limited resources and lack of technology, these small and micro-scale intangible cultural heritage projects face numerous challenges in archival management.

[0003] Traditional intangible cultural heritage archive management relies primarily on paper files, which present issues such as inconvenient storage, easy damage, and difficult retrieval. With the development of digital technology, some intangible cultural heritage projects have begun experimenting with digital archive management. However, for small and micro-scale intangible cultural heritage projects, existing digitization solutions are often too complex and costly, making them difficult to adapt to their small scale, limited funding, and weak technical capabilities. Therefore, there is an urgent need for a lightweight, low-cost, and easy-to-use digital archive management method and system to meet the archive management needs of small and micro-scale intangible cultural heritage projects. Summary of the Invention

[0004] The purpose of the present invention is to provide a lightweight digital archive management method and system for small and micro intangible cultural heritage projects to solve the problems raised in the above background technology.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a lightweight digital archive management method for small and micro intangible cultural heritage projects, comprising the following steps:

[0006] Step 1: Data collection: Using mobile devices, collect relevant information about small and micro-scale intangible cultural heritage projects by taking photos, recording videos, and recording audio. This includes physical objects, performances, and oral accounts of inheritors. At the same time, scan paper documents to digitize text materials. During the collection process, the system automatically adds data on time, location, and the person who collected the data.

[0007] Step 2: Data preprocessing: The data collected in step 1 is preliminarily processed through the data processing module, including image cropping and denoising, video editing and transcoding, audio noise reduction, and format conversion; lightweight image recognition and speech recognition technology are used to extract and annotate key information in the data, including the name of the intangible cultural heritage project, the name of the inheritor, and the name of the performance action;

[0008] Step 3, Data Storage: Using distributed storage technology, the data processed in Step 2 is stored in a storage architecture that combines cloud servers and local servers through a storage management module. Frequently used data and recently collected data are stored on local servers for quick access. Historical data and backup data are stored on cloud servers to save local storage space. At the same time, blockchain technology is used to encrypt the data storage and transmission process to ensure data security and non-tampering.

[0009] Step 4: Data retrieval and query: Users can use the retrieval and query module to search and query digital archives by keywords, time range, and intangible cultural heritage item categories. The retrieval and query module uses full-text search technology to quickly search the text content in the archives, and uses content-based image and video retrieval technology to search based on the characteristics of images and videos. The search results are displayed in a list format, and users can click to view detailed information.

[0010] Step 5. Data update and maintenance: When new data is generated or the original data changes in small and micro intangible cultural heritage projects, the digital archives are updated through mobile terminal devices or computer management platforms; the system automatically records the update log, including update time, update content, and updater information; at the same time, data is regularly backed up and restored to ensure data integrity and availability.

[0011] Preferably, the data collection in step 1 by taking photos, video recordings, and audio recordings to collect relevant information about the small and micro intangible cultural heritage projects specifically includes the following: the video recording function is used to focus on recording the performance process of the intangible cultural heritage project; from the opening ceremony of the performance to the core skill display link, and then to the closing action at the end of the performance, the entire process is recorded uninterruptedly, and the performance process and rhythm are completely preserved; during recording, the zoom function of the mobile phone is used to capture close-ups of key performance actions in a timely manner to ensure that the performance details are clearly presented;

[0012] Recording equipment is used to capture the oral narratives of inheritors; specifically, the narratives of inheritors should include the historical origins, technical skills, and stories of inheritance of the intangible cultural heritage items. During the recording process, a quiet environment should be selected to reduce external noise interference, and the volume of the recording equipment should be adjusted to ensure that the inheritor's voice is clear and audible. To make the oral narrative more logical, an interview outline should be prepared in advance to guide the inheritor in telling the story in an organized manner.

[0013] For paper documents, use scanning software paired with mobile devices to convert them into digital text materials. When scanning, ensure that the document is placed flat. The software automatically identifies the document boundaries and crops it to optimize the scanning effect. In other words, the scan is converted into a digital format that is easy to store and retrieve.

[0014] While completing the data collection in the above-mentioned manner, metadata information is automatically added. The metadata information includes time, location, and collector data information. The time information is accurate to the specific moment of collection, which is used to sort out the development and changes of intangible cultural heritage projects according to the timeline; the location information is obtained through the positioning function of the device, and the geographical location of the data collection is recorded in detail, which is used to record the geographical distribution and dissemination path of the intangible cultural heritage projects; the collector information is automatically associated with the user account that logs into the system, clarifying the person responsible for data collection, and facilitating the subsequent tracing and verification of the data source.

[0015] Preferably, the data preprocessing in step 2 specifically processes the image, video and audio data respectively through the data processing module, and the specific contents are as follows:

[0016] Image data preprocessing steps:

[0017] Cropping: The data processing module uses the Canny operator for edge detection, and collects the amplitude and direction of the pixel gradient in the image by calculating step 1, marking the edge pixels. Then, for the need of focusing on the image display, the data processing module provides a graphical interface, and the user manually drags the cropping box to carefully adjust the cropping range, retaining only the key parts of the intangible cultural heritage or scene, and removing unnecessary background; Denoising: First, the mean filtering algorithm is used. With each image pixel as the center, the average value of the pixel values in its neighborhood is calculated, and the original value of the pixel is replaced with this average value to smooth the image and preliminarily reduce the impact of noise. For the residual salt and pepper noise, the median filtering algorithm is used. Specifically, based on the neighborhood of the image pixel point, the pixel values in the neighborhood are sorted by size, and the median value is replaced by the original pixel value to further remove isolated noise points and improve image clarity and detail performance; Key information extraction and annotation: A lightweight MobileNet convolutional neural network model is used for image recognition. A rich intangible cultural heritage image dataset is pre-built, covering typical objects and scenes of various intangible cultural heritage projects. The images in the dataset are then finely annotated, including the name of the intangible cultural heritage project and its category. The cropped and denoised images are then input into the trained MobileNet model. The model accurately identifies the name of the intangible cultural heritage project corresponding to the image by extracting and analyzing image features layer by layer. At the same time, face recognition technology based on local binary pattern features is used to search the image for parts that match the facial features of the inheritor already entered into the system. If a match is successful, the inheritor's name is automatically annotated. For images showing performance actions, the action name is annotated by comparing the position and angle of the human joints in the image with the help of the action posture database.

[0018] Video data preprocessing steps

[0019] Editing: The data processing module uses a motion analysis algorithm based on the optical flow method to process the video frame by frame and identify the key frames in the video. Users can manually edit the video based on the key frame prompts on the timeline interface provided by the data processing module, deleting shooting errors, excessive pauses, or clips irrelevant to the core content of intangible cultural heritage, so that the video content is compact and the key points are highlighted. Transcoding: Based on the system storage and playback requirements, the H.265 encoding format is selected for transcoding. Key information extraction and annotation: Based on the improved dynamic time warping algorithm, the performance action sequence in the video is analyzed, and the action sequence in the video is carefully matched with the pre-built performance action template library. The performance action name corresponding to each action clip is accurately identified and clearly marked on the video timeline. The face recognition technology based on the convolutional neural network is used to analyze each frame of the video and search for the facial features of the inheritor. Once the inheritor is identified, the time period in which he appears is marked and the corresponding inheritor's name information is associated. At the same time, through the semantic analysis of the overall content of the video and combined with the intangible cultural heritage project knowledge base, the name of the intangible cultural heritage project involved in the video is accurately determined and marked.

[0020] Audio data preprocessing steps

[0021] Noise reduction: The data processing module uses a noise reduction method based on Wiener filtering to filter the collected audio signals, effectively removing background noise. It also automatically adjusts the filter window size and median calculation method according to the local characteristics of the audio signal to further improve the noise reduction effect and highlight the voice of the inheritor. Format conversion: According to the system storage and playback requirements, the different audio formats collected in step 1 are converted into MP3 format for easy storage and playback on multiple devices. Key information extraction and annotation: The lightweight Kaldi speech recognition toolkit is used to build a speech recognition model. First, the collected audio is framed and the continuous audio signal is divided into multiple short frames. Then, the Mel-frequency cepstral coefficient features are extracted for each frame of audio signal. The extracted features are then input into the trained speech recognition model. The model converts the speech signal into text content. Through semantic analysis of the text content and combined with the intangible cultural heritage project knowledge base, the key information of the intangible cultural heritage project name and the inheritor's name is accurately extracted and annotated on the time axis corresponding to the audio to facilitate subsequent retrieval and query.

[0022] Preferably, the search query module in step 4 uses full-text search technology to quickly search the text content in the archive, and uses content-based image retrieval and video retrieval technology to search according to the characteristics of the image and video. The specific implementation steps are as follows:

[0023] Full-text search of text content

[0024] Index construction: During the data preprocessing phase in step 2, the data processing module indexes all text materials. The index structure uses an inverted index to record in which documents each keyword appears and where it appears. Keyword matching: The system searches the inverted index for documents containing the keyword based on the keyword entered by the user. For multi-keyword queries, the system matches based on the logical relationship between the keywords. Relevance ranking: For matched documents, the retrieval query module ranks the documents by relevance based on the degree of keyword matching and frequency of occurrence. Documents with higher keyword matching and higher keyword frequency will be ranked higher.

[0025] Content-based image retrieval

[0026] Feature extraction: During the data preprocessing phase, the data processing module extracts preset features from the collected images. These features can reflect the visual information of the image. Feature matching: When a user enters a search request, the search query module extracts the corresponding features based on the user's input requirements and matches them with the features of the images in the archive using cosine similarity. Similarity sorting: Based on the results of feature matching, the search query module sorts the images by similarity, placing the images most similar to the user's requirements at the top.

[0027] Content-based video retrieval

[0028] Keyframe extraction and feature extraction: In the data preprocessing stage, the data processing module extracts keyframes from the video. Keyframes can represent the main content of the video. Then, features similar to image retrieval are extracted from the keyframes. At the same time, the motion features of the video are extracted to reflect the dynamic information in the video; Feature matching and scene analysis: When the user enters a retrieval request, the retrieval query module extracts corresponding features according to the user's needs and matches them with the keyframe features of the video in the archive. At the same time, the retrieval query module also performs scene analysis to determine whether the scene in the video meets the user's needs; Similarity sorting and clip positioning: Based on the results of feature matching and scene analysis, the retrieval query module sorts the videos by similarity. At the same time, the retrieval query module locates the clips in the video that are related to the user's needs, so that the user can quickly view relevant content;

[0029] Search results integration and display

[0030] The search query module integrates the results of text content retrieval, image retrieval and video retrieval, removes duplicate records, and uniformly sorts them according to relevance and similarity. The search results are displayed in a list on the user interface. The list contains the title, brief description, collection time, and intangible cultural heritage project category information of the archive. Users can click on a record in the list to view detailed information about the archive.

[0031] Preferably, the data update and maintenance in step 5 are specifically performed by the staff or inheritors of the small and micro intangible cultural heritage projects opening the user management module on a mobile phone or computer. On the main interface of the user management module application, there is a "data update" function entrance, which is clicked to enter the update page; the update specifically includes data entry: if there is new picture data, new photos of the intangible cultural heritage production process are taken, directly through the camera function in the device, or photos that have been taken are selected from the device album and uploaded; for video, new clips of intangible cultural heritage performances are also recorded on the spot, or relevant videos stored locally are selected for uploading; in terms of audio, a new oral content of the inheritor is recorded, and submitted after recording through the recording function of the device; if there is any text data update, the new content is entered in the text edit box; metadata supplement: when submitting the updated data, the user management module automatically pops up the metadata supplement window, and the staff supplements the shooting time, location, and brief description information of the data for better subsequent management and retrieval. After completing the entry of data and metadata, click the "Submit Update" button, and the updated data begins to be uploaded to the system server.

[0032] Preferably, a lightweight digital archive management system for small and micro intangible cultural heritage projects includes a mobile acquisition terminal module, a data processing module, a storage management module, a search and query module, and a user management module, characterized in that: the mobile acquisition terminal module is installed on a mobile terminal device and provides data acquisition functions, including photo taking, video recording, audio recording, and scanning function entrances. The user selects the corresponding acquisition method according to needs, and the collected data is uploaded to a local server in real time or temporarily stored in the mobile terminal device, and uploaded again when network conditions permit;

[0033] Data processing module: responsible for preprocessing the collected data and extracting key information, including image, video, and audio processing submodules, as well as an artificial intelligence-based information extraction submodule; the data processing module runs on a local server or a cloud server and can be flexibly configured based on the data volume and computing resources;

[0034] Storage Management Module: This module implements data storage and management. It includes a local storage management submodule and a cloud storage management submodule. It is responsible for data storage, migration, and backup between local and cloud servers. It also uses blockchain technology to encrypt and verify data to ensure data security and reliability.

[0035] Search and query module: provides users with data search and query functions. Users can access this module through a computer management platform or mobile terminal devices, enter search conditions such as keywords, time range, and intangible cultural heritage project categories, and the search and query module will quickly return search results;

[0036] User management module: manage users, including user registration, login, and permission allocation functions; assign different operation permissions based on the user's role, including data editing, viewing, and downloading permissions.

[0037] Compared with the existing technology, the beneficial effects of the present invention are: Lightweight design: The present invention adopts mobile terminal equipment for data collection, combined with lightweight data processing and storage technology, which reduces the hardware requirements and costs of the system, and is suitable for the actual situation of small and micro intangible cultural heritage projects.

[0038] Fast data collection and upload: Data can be collected anytime and anywhere using mobile devices, allowing for timely recording of intangible cultural heritage performances, production processes, and even the daily storytelling of inheritors. The collected data can be uploaded to the server in real time via a convenient network connection, eliminating the need for tedious manual collation and transmission processes, significantly shortening the time from data collection to storage. For example, at a temporary intangible cultural heritage performance, staff can immediately use their mobile phones to capture videos and photos, and upload the data to the digital archive system within minutes of the event.

[0039] Precise Search and Query: Leveraging full-text search technology for text content, as well as content-based image and video retrieval, users can quickly locate the intangible cultural heritage archival materials they need. Whether searching for a detailed introduction to a specific intangible cultural heritage item, records of a specific inheritor, or even video clips of a specific performance, accurate results are obtained in a short period of time. Compared to traditional paper archives or complex digital archival systems, search efficiency is increased several times, significantly increasing the value of archival materials and facilitating rapid access to information for researchers, inheritors, and general enthusiasts.

[0040] Facilitating inheritance and learning: Digital archival management makes intangible cultural heritage materials more accessible to inheritors and learners. Inheritors can review project history, key technical skills, and other materials at any time through the system, providing rich material for inheritance and teaching. Learners can also use convenient search functions to quickly find interesting intangible cultural heritage content for study, sparking greater interest and participation in intangible cultural heritage projects and promoting their inheritance.

[0041] Assisting cultural dissemination: The convenient sharing function provided by the system enables the widespread dissemination of intangible cultural heritage archival materials through the Internet. Whether on social media platforms, cultural exchange websites or online courses of educational institutions, it is easy to share the wonderful content of intangible cultural heritage projects, enhance the visibility and influence of small and micro intangible cultural heritage projects, and promote the dissemination and exchange of intangible cultural heritage culture on a wider scale. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1 Schematic diagram of the method flow of the present invention;

[0043] Figure 2 Schematic diagram of the system structure of the present invention. DETAILED DESCRIPTION

[0044] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only 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 making creative efforts are within the scope of protection of the present invention.

[0045] Example 1

[0046] See also Figure 1 The present invention provides a technical solution: a lightweight digital archive management method for small and micro intangible cultural heritage projects, comprising the following steps:

[0047] Step 1: Data collection: Using mobile devices, collect relevant information about small and micro-scale intangible cultural heritage projects by taking photos, recording videos, and recording audio. This includes physical objects, performances, and oral accounts of inheritors. At the same time, scan paper documents to digitize text materials. During the collection process, the system automatically adds data on time, location, and the person who collected the data.

[0048] The above-mentioned methods of collecting relevant information on small and micro-scale intangible cultural heritage projects through photography, video recording, and audio recording specifically include the following: the video recording function is used to focus on recording the performance process of the intangible cultural heritage project; from the opening ceremony of the performance to the core skill display, and then to the closing movements at the end of the performance, the entire process is recorded uninterruptedly, fully preserving the performance process and rhythm; when recording, the zoom function of the mobile phone is used to capture close-ups of key performance movements at the right time to ensure that the performance details are clearly presented;

[0049] Recording equipment is used to capture the oral narratives of inheritors; specifically, the narratives of inheritors should include the historical origins, technical skills, and stories of inheritance of the intangible cultural heritage items. During the recording process, a quiet environment should be selected to reduce external noise interference, and the volume of the recording equipment should be adjusted to ensure that the inheritor's voice is clear and audible. To make the oral narrative more logical, an interview outline should be prepared in advance to guide the inheritor in telling the story in an organized manner.

[0050] For paper documents, use scanning software paired with mobile devices to convert them into digital text materials. When scanning, ensure that the document is placed flat. The software automatically identifies the document boundaries and crops it to optimize the scanning effect. In other words, the scan is converted into a digital format that is easy to store and retrieve.

[0051] While completing the data collection in the above-mentioned manner, metadata information is automatically added. The metadata information includes time, location, and collector data information. The time information is accurate to the specific moment of collection, which is used to sort out the development and changes of intangible cultural heritage projects according to the timeline; the location information is obtained through the positioning function of the device, and the geographical location of the data collection is recorded in detail, which is used to record the geographical distribution and dissemination path of the intangible cultural heritage projects; the collector information is automatically associated with the user account that logs into the system, clarifying the person responsible for data collection, and facilitating the subsequent tracing and verification of the data source.

[0052] Step 2: Data preprocessing: The data collected in step 1 is preliminarily processed through the data processing module, including image cropping and denoising, video editing and transcoding, audio noise reduction, and format conversion; lightweight image recognition and speech recognition technology are used to extract and annotate key information in the data, including the name of the intangible cultural heritage project, the name of the inheritor, and the name of the performance action;

[0053] The data preprocessing is specifically carried out by processing the image, video and audio data respectively through the data processing module. The specific contents are as follows:

[0054] Image data preprocessing steps:

[0055] Cropping: The data processing module uses the Canny operator for edge detection, and collects the amplitude and direction of the pixel gradient in the image by calculating step 1, marking the edge pixels. Then, for the need of focusing on the image display, the data processing module provides a graphical interface, and the user manually drags the cropping box to carefully adjust the cropping range, retaining only the key parts of the intangible cultural heritage or scene, and removing unnecessary background; Denoising: First, the mean filtering algorithm is used. With each image pixel as the center, the average value of the pixel values in its neighborhood is calculated, and the original value of the pixel is replaced with this average value to smooth the image and preliminarily reduce the impact of noise. For the residual salt and pepper noise, the median filtering algorithm is used. Specifically, based on the neighborhood of the image pixel point, the pixel values in the neighborhood are sorted by size, and the median value is replaced by the original pixel value to further remove isolated noise points and improve image clarity and detail performance; Key information extraction and annotation: A lightweight MobileNet convolutional neural network model is used for image recognition. A rich intangible cultural heritage image dataset is pre-built, covering typical objects and scenes of various intangible cultural heritage projects. The images in the dataset are then finely annotated, including the name of the intangible cultural heritage project and its category. The cropped and denoised images are then input into the trained MobileNet model. The model accurately identifies the name of the intangible cultural heritage project corresponding to the image by extracting and analyzing image features layer by layer. At the same time, face recognition technology based on local binary pattern features is used to search the image for parts that match the facial features of the inheritor already entered into the system. If a match is successful, the inheritor's name is automatically annotated. For images showing performance actions, the action name is annotated by comparing the position and angle of the human joints in the image with the help of the action posture database.

[0056] Video data preprocessing steps

[0057] Editing: The data processing module uses a motion analysis algorithm based on the optical flow method to process the video frame by frame and identify the key frames in the video. Users can manually edit the video based on the key frame prompts on the timeline interface provided by the data processing module, deleting shooting errors, excessive pauses, or clips irrelevant to the core content of intangible cultural heritage, so that the video content is compact and the key points are highlighted. Transcoding: Based on the system storage and playback requirements, the H.265 encoding format is selected for transcoding. Key information extraction and annotation: Based on the improved dynamic time warping algorithm, the performance action sequence in the video is analyzed, and the action sequence in the video is carefully matched with the pre-built performance action template library. The performance action name corresponding to each action clip is accurately identified and clearly marked on the video timeline. The face recognition technology based on the convolutional neural network is used to analyze each frame of the video and search for the facial features of the inheritor. Once the inheritor is identified, the time period in which he appears is marked and the corresponding inheritor's name information is associated. At the same time, through the semantic analysis of the overall content of the video and combined with the intangible cultural heritage project knowledge base, the name of the intangible cultural heritage project involved in the video is accurately determined and marked.

[0058] Audio data preprocessing steps

[0059] Noise reduction: The data processing module uses a noise reduction method based on Wiener filtering to filter the collected audio signals, effectively removing background noise. It also automatically adjusts the filter window size and median calculation method according to the local characteristics of the audio signal to further improve the noise reduction effect and highlight the voice of the inheritor. Format conversion: According to the system storage and playback requirements, the different audio formats collected in step 1 are converted into MP3 format for easy storage and playback on multiple devices. Key information extraction and annotation: The lightweight Kaldi speech recognition toolkit is used to build a speech recognition model. First, the collected audio is framed and the continuous audio signal is divided into multiple short frames. Then, the Mel-frequency cepstral coefficient features are extracted for each frame of audio signal. The extracted features are then input into the trained speech recognition model. The model converts the speech signal into text content. Through semantic analysis of the text content and combined with the intangible cultural heritage project knowledge base, the key information of the intangible cultural heritage project name and the inheritor's name is accurately extracted and annotated on the time axis corresponding to the audio to facilitate subsequent retrieval and query.

[0060] Step 3, Data Storage: Using distributed storage technology, the data processed in Step 2 is stored in a storage architecture that combines cloud servers and local servers through a storage management module. Frequently used data and recently collected data are stored on local servers for quick access. Historical data and backup data are stored on cloud servers to save local storage space. At the same time, blockchain technology is used to encrypt the data storage and transmission process to ensure data security and non-tampering.

[0061] Step 4: Data retrieval and query: Users can use the retrieval and query module to search and query digital archives by keywords, time range, and intangible cultural heritage item categories. The retrieval and query module uses full-text search technology to quickly search the text content in the archives, and uses content-based image and video retrieval technology to search based on the characteristics of images and videos. The search results are displayed in a list format, and users can click to view detailed information. The specific implementation steps are as follows:

[0062] Full-text search of text content

[0063] Index construction: During the data preprocessing phase in step 2, the data processing module indexes all text materials. The index structure uses an inverted index to record in which documents each keyword appears and where it appears. Keyword matching: The system searches the inverted index for documents containing the keyword based on the keyword entered by the user. For multi-keyword queries, the system matches based on the logical relationship between the keywords. Relevance ranking: For matched documents, the retrieval query module ranks the documents by relevance based on the degree of keyword matching and frequency of occurrence. Documents with higher keyword matching and higher keyword frequency will be ranked higher.

[0064] Content-based image retrieval

[0065] Feature extraction: During the data preprocessing phase, the data processing module extracts preset features from the collected images. These features can reflect the visual information of the image. Feature matching: When a user enters a search request, the search query module extracts the corresponding features based on the user's input requirements and matches them with the features of the images in the archive using cosine similarity. Similarity sorting: Based on the results of feature matching, the search query module sorts the images by similarity, placing the images most similar to the user's requirements at the top.

[0066] Content-based video retrieval

[0067] Keyframe extraction and feature extraction: In the data preprocessing stage, the data processing module extracts keyframes from the video. Keyframes can represent the main content of the video. Then, features similar to image retrieval are extracted from the keyframes. At the same time, the motion features of the video are extracted to reflect the dynamic information in the video; Feature matching and scene analysis: When the user enters a retrieval request, the retrieval query module extracts corresponding features according to the user's needs and matches them with the keyframe features of the video in the archive. At the same time, the retrieval query module also performs scene analysis to determine whether the scene in the video meets the user's needs; Similarity sorting and clip positioning: Based on the results of feature matching and scene analysis, the retrieval query module sorts the videos by similarity. At the same time, the retrieval query module locates the clips in the video that are related to the user's needs, so that the user can quickly view relevant content;

[0068] Search results integration and display

[0069] The search query module integrates the results of text content retrieval, image retrieval and video retrieval, removes duplicate records, and uniformly sorts them according to relevance and similarity. The search results are displayed in a list on the user interface. The list contains the title, brief description, collection time, and intangible cultural heritage project category information of the archive. Users can click on a record in the list to view detailed information about the archive.

[0070] Step 5: Data update and maintenance: When new data is generated or existing data changes for small and micro intangible cultural heritage projects, the digital archives are updated through mobile devices or computer management platforms. The system automatically records update logs, including update time, update content, and updater information. At the same time, data is regularly backed up and restored to ensure data integrity and availability.

[0071] Data update and maintenance are specifically carried out by the staff or inheritors of small and micro intangible cultural heritage projects opening the user management module on their mobile phones or computers. On the main interface of the user management module application, there is a "data update" function entrance, which can be clicked to enter the update page; the update specifically includes data entry: if there is new picture data, new photos of the intangible cultural heritage production process are taken directly through the camera function in the device, or photos that have been taken are selected from the device album and uploaded; for video, new clips of intangible cultural heritage performances are also recorded on the spot, or relevant videos stored locally are selected for upload; in terms of audio, a new oral content of the inheritor is recorded through the recording function of the device and then submitted; if there is any text data update, the new content is entered in the text edit box; metadata supplement: when submitting the updated data, the user management module automatically pops up the metadata supplement window, and the staff supplements the shooting time, location, and brief description information of the data for better subsequent management and retrieval. After completing the entry of data and metadata, click the "Submit Update" button, and the updated data will begin to be uploaded to the system server.

[0072] Example 2

[0073] like Figure 2 A lightweight digital archive management system for small and micro intangible cultural heritage projects, including a mobile collection terminal module, a data processing module, a storage management module, a search and query module, and a user management module. The mobile collection terminal module is installed on a mobile terminal device and provides data collection functions, including photo taking, video recording, audio recording, and scanning function entrances. Users select the corresponding collection method according to their needs. The collected data is uploaded to the local server in real time or temporarily stored on the mobile terminal device, and will be uploaded when network conditions permit.

[0074] Data processing module: responsible for preprocessing the collected data and extracting key information, including image, video, and audio processing submodules, as well as an artificial intelligence-based information extraction submodule; the data processing module runs on a local server or a cloud server and can be flexibly configured based on the data volume and computing resources;

[0075] Storage Management Module: This module implements data storage and management. It includes a local storage management submodule and a cloud storage management submodule. It is responsible for data storage, migration, and backup between local and cloud servers. It also uses blockchain technology to encrypt and verify data to ensure data security and reliability.

[0076] Search and query module: provides users with data search and query functions. Users can access this module through a computer management platform or mobile terminal devices, enter search conditions such as keywords, time range, and intangible cultural heritage project categories, and the search and query module will quickly return search results;

[0077] User management module: manage users, including user registration, login, and permission allocation functions; assign different operation permissions based on the user's role, including data editing, viewing, and downloading permissions.

[0078] The lightweight design of this invention is highly compatible with the limited resources of small and micro-scale intangible cultural heritage projects, significantly reducing costs. Its efficient and convenient management process significantly improves archive utilization efficiency. Its data is secure and reliable, laying a solid foundation for the long-term preservation of intangible cultural heritage archives. It can also effectively promote the inheritance and development of intangible cultural heritage and widely promote cultural dissemination. It has broad application prospects in the fields of protection, inheritance, and research of small and micro-scale intangible cultural heritage projects, and is expected to bring revolutionary changes to the archive management of many small and micro-scale intangible cultural heritage projects, helping to better inherit and develop intangible cultural heritage in the digital age.

[0079] Through the above implementation methods, the lightweight digital archive management method and system for small and micro intangible cultural heritage projects of the present invention can effectively realize the digital management of small and micro intangible cultural heritage project archives, improve the efficiency and quality of archive management, and provide strong support for the inheritance and development of small and micro intangible cultural heritage projects.

[0080] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to these embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. A lightweight digital archive management method for small and micro intangible cultural heritage projects, characterized by: The following steps are involved: Step 1: Data collection: Using mobile devices, collect relevant information about small and micro-scale intangible cultural heritage projects by taking photos, recording videos, and recording audio. This includes physical objects, performances, and oral accounts of inheritors. At the same time, scan paper documents to digitize text materials. During the collection process, the system automatically adds data on time, location, and the person who collected the data. Step 2: Data preprocessing: The data collected in step 1 is preliminarily processed through the data processing module, including image cropping and denoising, video editing and transcoding, audio noise reduction, and format conversion; lightweight image recognition and speech recognition technology are used to extract and annotate key information in the data, including the name of the intangible cultural heritage project, the name of the inheritor, and the name of the performance action; Step 3, data storage: using distributed storage technology, the data processed in step 2 is stored in a storage architecture that combines cloud servers and local servers through a storage management module; Frequently used data and recently collected data are stored on local servers for quick access; historical data and backup data are stored on cloud servers to save local storage space. At the same time, blockchain technology is used to encrypt the data storage and transmission process to ensure data security and non-tampering. Step 4: Data retrieval and query: Users can use the retrieval and query module to search and query digital archives by keywords, time range, and intangible cultural heritage item categories. The retrieval and query module uses full-text search technology to quickly search the text content in the archives, and uses content-based image and video retrieval technology to search based on the characteristics of images and videos. The search results are displayed in a list format, and users click to view detailed information; Step 5. Data update and maintenance: When new data is generated or the original data changes in small and micro intangible cultural heritage projects, the digital archives are updated through mobile terminal devices or computer management platforms; the system automatically records the update log, including update time, update content, and updater information; at the same time, data is regularly backed up and restored to ensure data integrity and availability.

2. A lightweight digital archive management method for small and micro intangible cultural heritage projects according to claim 1, characterized in that: In step 1, the data collection process involves taking photos, recording videos, and recording audio to collect relevant information about small and micro-sized intangible cultural heritage projects, specifically including the following: The video recording function is designed to focus on documenting the performance process of intangible cultural heritage items. From the opening ceremony to the core skills demonstration and the closing movements, the entire process is recorded uninterruptedly, fully preserving the flow and rhythm of the performance. During recording, the phone's zoom function is used to capture close-ups of key performance movements at the right time, ensuring that the performance details are presented clearly. Recording equipment is used to capture the oral narratives of inheritors; specifically, the narratives of inheritors should include the historical origins, technical skills, and stories of inheritance of the intangible cultural heritage items. During the recording process, a quiet environment should be selected to reduce external noise interference, and the volume of the recording equipment should be adjusted to ensure that the inheritor's voice is clear and audible. To make the oral narrative more logical, an interview outline should be prepared in advance to guide the inheritor in telling the story in an organized manner. For paper documents, use scanning software paired with mobile devices to convert them into digital text materials. When scanning, ensure that the document is placed flat. The software automatically identifies the document boundaries and crops it to optimize the scanning effect. In other words, the scan is converted into a digital format that is easy to store and retrieve. While completing the data collection in the above-mentioned manner, metadata information is automatically added. The metadata information includes time, location, and collector data information. The time information is accurate to the specific moment of collection, which is used to sort out the development and changes of intangible cultural heritage projects according to the timeline; the location information is obtained through the positioning function of the device, and the geographical location of the data collection is recorded in detail, which is used to record the geographical distribution and dissemination path of the intangible cultural heritage projects; the collector information is automatically associated with the user account that logs into the system, clarifying the person responsible for data collection, and facilitating the subsequent tracing and verification of the data source.

3. A lightweight digital archive management method for small and micro intangible cultural heritage projects according to claim 1, characterized in that: The data preprocessing in step 2 is specifically performed by processing the image, video and audio data respectively through the data processing module, and the specific contents are as follows: Image data preprocessing steps: Cropping: The data processing module uses the Canny operator for edge detection. It calculates the magnitude and direction of the gradient of the pixels in the image collected in step 1 and marks the edge pixels. Then, to meet the need for highlighting the image, the data processing module provides a graphical interface. The user manually drags the cropping frame and fine-tunes the cropping range to retain only the key parts of the intangible cultural heritage object or scene and remove any excess background. Denoising: First, the mean filter algorithm is used. With each image pixel as the center, the average value of the pixel values in its neighborhood is calculated. This average value is used to replace the original value of the pixel to smooth the image and initially reduce the impact of noise. For the residual salt and pepper noise, the median filter algorithm is used. Specifically, based on the neighborhood of the image pixel point, the pixel values in the neighborhood are sorted by size, and the median value is used to replace the original pixel value to further remove isolated noise points and improve the image clarity and detail performance; Key information extraction and annotation: The lightweight MobileNet convolutional neural network model is used for image recognition, and a rich intangible cultural heritage image dataset is pre-built, covering various intangible cultural heritage projects. Typical objects and scenes are collected, and the images in the dataset are finely labeled, including the names of the intangible cultural heritage items and their categories. The cropped and denoised images are then input into the trained MobileNet model. The model accurately identifies the names of the intangible cultural heritage items corresponding to the images by extracting and analyzing the image features layer by layer. At the same time, face recognition technology based on local binary pattern features is used to search the image for parts that match the facial features of the inheritors entered into the system. If the match is successful, the inheritor's name is automatically labeled. For images showing performance actions, the name of the performance action is labeled by comparing the positions and angles of the human joints in the image with the help of the action posture database. Video data preprocessing steps Editing: The data processing module uses a motion analysis algorithm based on optical flow to process the video frame by frame and identify key frames in the video. Users can manually edit the video based on the key frame prompts on the timeline interface provided by the data processing module, deleting shooting errors, excessive pauses, or clips irrelevant to the core content of intangible cultural heritage, making the video content compact and highlighting the key points. Transcoding: Based on the system storage and playback requirements, the H.265 encoding format is selected for transcoding; Key Information Extraction and Annotation: Based on the improved dynamic time warping algorithm, the performance action sequence in the video is analyzed, and the action sequence in the video is carefully matched with the pre-built performance action template library. The performance action name corresponding to each action clip is accurately identified and clearly marked on the video timeline. Using face recognition technology based on convolutional neural networks, each frame of the video is analyzed to search for the facial features of the inheritor. Once the inheritor is identified, the time period of his / her appearance is annotated and the corresponding inheritor's name information is associated. At the same time, through semantic analysis of the overall content of the video and combined with the intangible cultural heritage project knowledge base, the name of the intangible cultural heritage project involved in the video is accurately determined and annotated; Audio data preprocessing steps Noise reduction: The data processing module uses a noise reduction method based on Wiener filtering to filter the collected audio signals, effectively removing background noise. It also automatically adjusts the filter window size and median calculation method based on the local characteristics of the audio signal to further enhance the noise reduction effect and highlight the voice of the inheritors. Format conversion: According to the system storage and playback requirements, the different audio formats collected in step 1 are converted into MP3 format for easy storage and playback on multiple devices; key information extraction and annotation: Use the lightweight Kaldi speech recognition toolkit to build a speech recognition model. First, the collected audio is framed and the continuous audio signal is divided into multiple short frames. Then, the Mel-frequency cepstral coefficient features are extracted for each frame of audio signal. The extracted features are then input into the trained speech recognition model. The model converts the speech signal into text content. Through semantic analysis of the text content and combined with the intangible cultural heritage project knowledge base, the key information of the intangible cultural heritage project name and the inheritor's name is accurately extracted and annotated on the time axis corresponding to the audio to facilitate subsequent retrieval and query.

4. A lightweight digital archive management method for small and micro intangible cultural heritage projects according to claim 1, characterized in that: In step 4, the search query module uses full-text search technology to quickly search the text content in the archive, and uses content-based image search and video search technology to search according to the characteristics of images and videos. The specific implementation steps are as follows: Full-text search of text content Index construction: During the data preprocessing phase in step 2, the data processing module indexes all text materials. The index structure uses an inverted index to record in which documents each keyword appears and where it appears. Keyword matching: The system searches the inverted index for documents containing the keyword based on the keyword entered by the user. For multi-keyword queries, the system matches based on the logical relationship between the keywords. Relevance ranking: For matched documents, the retrieval query module ranks the documents by relevance based on the degree of keyword matching and frequency of occurrence. Documents with higher keyword matching and higher keyword frequency will be ranked higher. Content-based image retrieval Feature extraction: In the data preprocessing stage, the data processing module extracts preset features from the collected images. The features can reflect the visual information of the image; Feature matching: When a user enters a search request, the search query module extracts the corresponding features based on the user's input requirements and matches them with the features of the images in the archive using cosine similarity. Similarity sorting: Based on the result of feature matching, the search query module sorts the images by similarity, placing the images most similar to the user's requirements at the top. Content-based video retrieval Keyframe extraction and feature extraction: In the data preprocessing stage, the data processing module extracts keyframes from the video. Keyframes can represent the main content of the video. Then, features similar to image retrieval are extracted from the keyframes. At the same time, the motion features of the video are extracted to reflect the dynamic information in the video; Feature matching and scene analysis: When the user enters a retrieval request, the retrieval query module extracts corresponding features according to the user's needs and matches them with the keyframe features of the video in the archive. At the same time, the retrieval query module also performs scene analysis to determine whether the scene in the video meets the user's needs; Similarity sorting and clip positioning: Based on the results of feature matching and scene analysis, the retrieval query module sorts the videos by similarity. At the same time, the retrieval query module locates the clips in the video that are related to the user's needs, so that the user can quickly view relevant content; Search results integration and display The search query module integrates the results of text content retrieval, image retrieval and video retrieval, removes duplicate records, and uniformly sorts them according to relevance and similarity. The search results are displayed in a list on the user interface. The list contains the title, brief description, collection time, and intangible cultural heritage project category information of the archive. Users can click on a record in the list to view detailed information about the archive.

5. A lightweight digital archive management method for small and micro intangible cultural heritage projects according to claim 1, characterized in that: The data update and maintenance in step 5 is specifically done by the staff or inheritors of the small and micro intangible cultural heritage projects opening the user management module on their mobile phones or computers. On the main interface of the user management module application, there is a "data update" function entrance, which is clicked to enter the update page; the update specifically includes data entry: if there is new picture data, new photos of the intangible cultural heritage production process are taken, which can be taken directly through the camera function in the device, or photos taken from the device album can be uploaded; for videos, new clips of intangible cultural heritage performances can also be recorded on the spot, or relevant videos stored locally can be uploaded; In terms of audio, record the new oral content of the inheritor through the recording function of the device and submit it; If there is any text data update, enter the new content in the text edit box; Metadata supplement: When submitting updated data, the user management module automatically pops up the metadata supplement window. The staff can add the shooting time, location, and brief description information of the data to facilitate better management and retrieval. After completing the entry of data and metadata, click the "Submit Update" button and the updated data will begin to be uploaded to the system server.

6. A lightweight digital archive management system for small and micro intangible cultural heritage projects according to any one of claims 1 to 5, comprising a mobile acquisition terminal module, a data processing module, a storage management module, a search and query module, and a user management module, characterized in that: The mobile data acquisition terminal module is installed on the mobile terminal device and provides data acquisition functions, including photo taking, video recording, audio recording, and scanning. The user selects the corresponding acquisition method according to the needs. The collected data is uploaded to the local server in real time or temporarily stored on the mobile terminal device and uploaded again when the network conditions allow. Data processing module: responsible for preprocessing the collected data and extracting key information, including image, video, and audio processing submodules, as well as an artificial intelligence-based information extraction submodule; the data processing module runs on a local server or a cloud server and can be flexibly configured based on the data volume and computing resources; Storage Management Module: This module implements data storage and management. It includes a local storage management submodule and a cloud storage management submodule. It is responsible for data storage, migration, and backup between local and cloud servers. It also uses blockchain technology to encrypt and verify data to ensure data security and reliability. Search and query module: provides users with data search and query functions. Users can access this module through a computer management platform or mobile terminal devices, enter search conditions such as keywords, time range, and intangible cultural heritage project categories, and the search and query module will quickly return search results; User management module: manage users, including user registration, login, and authority allocation functions; Different operating permissions are assigned according to the user's role, including data editing, viewing, and downloading permissions.

Images