A Blockchain-Based Workload Measurement Method for Extracting Ancient Book Cultural Materials

Abstract

A blockchain-based method for workload assessment in extracting ancient text cultural materials includes image processing, ancient text character localization, ancient text character recognition, ancient text text identification, and ancient text digitization. Adopting this structure, the present invention offers the following advantages: it systematically integrates original ancient text resources and transforms them into digital resources, facilitating scholarly retrieval and enabling material extraction. This system accelerates the digitization process of ancient texts and improves the accuracy of ancient text character recognition by utilizing human-in-the-loop processing.

Description

Technical Field

[0001] This invention relates to the field of ancient book text recognition technology, specifically to a method for workload calibration of ancient book cultural material extraction based on blockchain. Background Technology

[0002] Ancient books, as an important part of human cultural heritage, possess extremely high academic research and artistic appreciation value. Due to their rarity and exclusivity, these values ​​cannot be widely utilized by the public. Even within strictly defined areas, the security and sustainable preservation of original ancient books remain difficult to guarantee. The discovery and effective utilization of ancient texts has become one of the main goals of digital library projects in various countries. However, in the area of ​​ancient book collation and digitization, there is a lack of a complete system to realize the process from original ancient book resources to modern material extraction. Summary of the Invention

[0003] This invention aims to address the serious deficiency in the digitization, datafication, and knowledge-based representation of ancient book collections, and to develop a method for calibrating the workload of extracting ancient book cultural materials for the batch development and reuse of ancient book cultural resources with independent intellectual property rights.

[0004] To solve the above-mentioned technical problems, the technical solution provided by the present invention is as follows:

[0005] A blockchain-based method for workload assessment in extracting ancient text cultural materials includes image processing, ancient text character localization, ancient text character recognition, ancient text text identification, and ancient text digitization. The steps of the assessment method are as follows:

[0006] S1. Image Acquisition: Taking photos using an external camera;

[0007] S2. Preprocessing the image: Preprocess the photo from step S1 and perform region segmentation on the photo;

[0008] S3. Convert image colors: Convert the page image of the photo from step S2 into black and white;

[0009] S4. Character localization of ancient books: Character localization of photos is achieved by using an adaptive character detection coarse localization algorithm, coarse segmentation is performed by binarization, gridding and preprocessing techniques, and fine segmentation is performed by using Chinese character features, so as to realize the character localization of massive digital ancient book objects.

[0010] S5. Ancient Book Character Recognition: Through clustering algorithms, similar character shapes are automatically clustered, and characters with similar shapes are grouped into one category, resulting in a large number of characters in a series of single characters;

[0011] S6. Ancient Text Identification: Automatic word segmentation of data text is performed using N-grams and Hidden Markov Models, providing punctuation and part-of-speech information to obtain text with punctuation, word segmentation, and part-of-speech tags.

[0012] S7. Digitalization and collation of ancient books: Knowledge outputs resulting from the summarization and collation of information from texts.

[0013] With the above structure, the present invention has the following advantages:

[0014] This system systematically integrates original ancient text resources and transforms them into digital resources, facilitating scholarly research and enabling material extraction. It accelerates the digitization process of ancient texts and improves the accuracy of character recognition by utilizing human-in-the-loop technology.

[0015] As an improvement, the character positioning in step S4 can be displayed on the page, and the user can fine-tune the results to make the positioning more accurate.

[0016] As an improvement, the data text for ancient book character recognition in step S5 allows experts and the public to identify the overall cluster and correct character images with clustering errors.

[0017] As an improvement, the ancient text identification in step S6 allows experts and the public to revise the word segmentation results of the algorithm, add personal names, place names and tags to the text.

[0018] As an improvement, the ancient book digitization process in step S7 displays the digitization progress of each ancient book, making it convenient for users to search and query the ancient books. The digitization progress is presented as a percentage. Attached Figure Description

[0019] Figure 1 This is a flowchart of the present invention.

[0020] Figure 2 This is a cross-section showing the digital organization of ancient books in this invention.

[0021] Figure 3 This is a schematic diagram of Embodiment 1 of the present invention. Detailed Implementation

[0022] The present invention will now be described in further detail with reference to the full text.

[0023] Combined with appendix Figures 1-3 A blockchain-based method for workload assessment in extracting ancient text cultural materials includes image processing, ancient text character localization, ancient text character recognition, ancient text text identification, and ancient text digitization. The steps of the assessment method are as follows:

[0024] S1. Image Acquisition: Taking photos using an external camera;

[0025] S2. Preprocessing the image: Preprocess the photo from step S1 and perform region segmentation on the photo. Remove the regions without ancient text characters by preprocessing the region segmentation.

[0026] S3. Image Color Conversion: Convert the page image of the photo from step S2 to black and white. This can be achieved by converting RGB images to grayscale, directly extracting the luminance signal, and using an autoencoder. The autoencoder, which limits the dimensionality of the latent space or has a smaller dimensionality than the input data (x), learns by minimizing the loss function L(x, g(f(x))), where L is a loss function that penalizes g(f(x)) for inconsistencies with x. g(f(x)) is the encoding and decoding process of the autoencoder (i.e., the output of the autoencoder), the loss function is a mean squared error function, x is the color image, and g(f(x)) is the reconstructed black and white image. The loss function penalizes the autoencoder for failing to reconstruct a black and white version of the color image.

[0027] S4. Character Localization in Ancient Books: Character localization in photos is achieved through an adaptive character detection coarse localization algorithm. Binarization, gridding, and preprocessing techniques are used for coarse segmentation, while fine segmentation is performed using Chinese character features. This enables character localization of massive amounts of digitized ancient book objects. The localization results are displayed on the page, and users can fine-tune the results to make the localization more accurate.

[0028] S5. Ancient Book Character Recognition: Through clustering algorithms, similar character shapes are automatically clustered. Characters with similar shapes are grouped into one category, resulting in a large amount of data text of a series of single characters. The character location results can be displayed on the page, and users can fine-tune the results to make the location more accurate.

[0029] S6. Ancient Text Identification: Automatic word segmentation of data text is performed using N-grams and Hidden Markov Models, providing punctuation and part-of-speech information to obtain text with punctuation, word segmentation, and part-of-speech tags. Ancient text identification allows experts and the public to revise the word segmentation results of the algorithm, add personal names, place names, and tags to the text.

[0030] S7. Digitization and Organization of Ancient Books: This involves summarizing and organizing information from texts to create knowledge outcomes. The digitization and organization of ancient books displays the progress of digitization for each book, making it convenient for users to search and query them. The digitization progress is presented as a percentage.

[0031] The character positioning in step S4 can display the positioning results on the page, and users can fine-tune the results to make the positioning more accurate.

[0032] The data text for ancient book character recognition in step S5 allows experts and the public to identify the overall cluster and correct character images with clustering errors.

[0033] The ancient text identification in step S6 allows experts and the public to revise the word segmentation results of the algorithm, add personal names, place names and tags to the text.

[0034] The S7 step of digitizing and organizing ancient books displays the digitization progress of each book, facilitating user retrieval and searching. Searches can be conducted by book number, section / category / genus / order, title, author's era, or author's name. The digitization progress is presented as a percentage, such as... Figure 2 As shown.

[0035] Example 1:

[0036] The steps for determining the workload of extracting cultural materials from ancient books are as follows:

[0037] S1. Image Acquisition: Taking photos via an external camera, such as... Figure 3 The first picture shown is a photograph of an ancient book;

[0038] S2. Preprocessing the image: Preprocess the photo from step S1 and perform region segmentation on the photo. Remove the regions without ancient text characters by preprocessing the region segmentation.

[0039] S3. Image Color Conversion: Convert the page image of the photo from step S2 to black and white. This can be done by converting RGB images to grayscale, directly extracting the luminance signal, or using an autoencoder to change the image colors. Figure 3 As shown in the second image;

[0040] S4. Ancient Book Character Localization: This feature uses an adaptive character detection coarse localization algorithm to locate characters in photos. It employs binarization, meshing, and preprocessing techniques for coarse segmentation, followed by fine segmentation using Chinese character features. This enables character localization of massive amounts of digitized ancient book objects. The Chinese characters in the image content are segmented, and the individual character images are saved according to the original text's text order (from right to left). The localization results are displayed on the page, and users can fine-tune the results to make the localization more accurate. Figure 3 As shown in the third image;

[0041] S5. Ancient Text Character Recognition: This feature uses a clustering algorithm to automatically cluster similar character shapes, grouping characters with similar shapes into one category. This results in a large amount of text data containing a series of individual characters. Character location results can be displayed on the page. Figure 3 As shown in the fourth image, users can fine-tune the results to make the positioning more accurate.

[0042] S6. Ancient Text Identification: This method automatically segments the data text using unsupervised, dictionary-free N-grammar and Hidden Markov Models, providing punctuation and part-of-speech information to obtain text with punctuation, segmentation, and part-of-speech tags, such as... Figure 3 As shown in the last image, the ancient text tagging allows experts and the public to revise the word segmentation results of the algorithm, add names of people and places, and make tags to the text;

[0043] S7. Digitization and Organization of Ancient Books: This involves summarizing and organizing textual information into knowledge products. The digitization progress of each ancient book is displayed, facilitating user retrieval and searching. Digitization progress is presented as a percentage, such as... Figure 2 As shown.

[0044] This system systematically integrates original ancient text resources and transforms them into digital resources, facilitating scholarly research and enabling material extraction. It accelerates the digitization process of ancient texts and improves the accuracy of character recognition by utilizing human-in-the-loop technology.

[0045] The present invention and its embodiments have been described above. This description is not restrictive, and the embodiments shown throughout are only one of the embodiments of the present invention. The actual structure is not limited to this. In conclusion, if those skilled in the art are inspired by this description and design similar structures and embodiments without departing from the spirit of the present invention, they should all fall within the protection scope of the present invention.

Claims

1. A method for workload assessment of ancient book cultural material extraction based on blockchain, characterized in that, The calibration method, which includes image processing, ancient book character localization, ancient book character recognition, ancient book text identification, and ancient book digitization, follows these steps: S1. Image Acquisition: Taking photos using an external camera; S2. Preprocessing the image: Preprocess the photo from step S1 and perform region segmentation on the photo; S3. Convert image colors: Convert the page image of the photo from step S2 into black and white; S4. Character localization of ancient books: Character localization of photos is achieved by using an adaptive character detection coarse localization algorithm, coarse segmentation is performed by binarization, gridding and preprocessing techniques, and fine segmentation is performed by using Chinese character features, so as to realize the character localization of massive digital ancient book objects. S5. Ancient Book Character Recognition: Through clustering algorithms, similar character shapes are automatically clustered, and characters with similar shapes are grouped into one category, resulting in a large number of characters in a series of single characters; S6. Ancient Text Identification: Automatic word segmentation of data text is performed using N-grams and Hidden Markov Models, providing punctuation and part-of-speech information to obtain text with punctuation, word segmentation, and part-of-speech tags. S7. Digitalization and collation of ancient books: Knowledge outputs resulting from the summarization and collation of information from texts; The character positioning in step S4 can display the positioning results on the page, and users can fine-tune the results to make the positioning more accurate.

2. The method for workload assessment of ancient book cultural material extraction based on blockchain according to claim 1, characterized in that: The data text for ancient book character recognition in step S5 allows experts and the public to identify the overall cluster and correct character images with clustering errors.

3. The method for workload assessment of ancient book cultural material extraction based on blockchain according to claim 1, characterized in that: The ancient text identification in step S6 allows experts and the public to revise the word segmentation results of the algorithm, add personal names, place names and tags to the text.

4. The method for workload assessment of ancient book cultural material extraction based on blockchain according to claim 1, characterized in that: The S7 step of digitizing and organizing ancient books displays the digitization progress of each ancient book, making it convenient for users to search and query the ancient books. The digitization progress is presented in the form of a percentage.

Images