Method, system, device and medium for digital restoration of defective ancient book scan image

The end-to-end restoration method using OCR processing and AI model assistance solves the problem of manual reliance in the digital restoration of ancient books, achieving efficient text completion and background restoration, and improving restoration efficiency and restoration capabilities.

CN121921213BActive Publication Date: 2026-06-19HEFEI INSTITUTE OF PHYSICAL SCIENCE CHINESE ACADEMY OF SCIENCES
View PDF 2 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
HEFEI INSTITUTE OF PHYSICAL SCIENCE CHINESE ACADEMY OF SCIENCES
Filing Date
2026-03-25
Publication Date
2026-06-19

AI Technical Summary

Technical Problem

In existing technologies, the verification of missing content in the process of digital restoration of ancient books relies on the knowledge reserves of industry experts, which is a large workload. Image restoration relies on manual operation, and the image processing technology provided by commercial systems is simple and lacks intelligent restoration methods.

Method used

A full-process assisted restoration method is adopted, including OCR processing, missing area segmentation, location enhancement inference model, text restoration and background restoration. By combining OCR results and AI models, the text completion and background restoration of missing ancient books are completed automatically.

Benefits of technology

It improves the efficiency of digital restoration of ancient books, simplifies the restoration process, enhances the ability to restore the original layout, is easy for industry professionals to use, and integrates a variety of tools for reasoning and proofreading.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN121921213B_ABST
    Figure CN121921213B_ABST
Patent Text Reader

Abstract

This invention discloses a method, system, device, and medium for digital restoration of scanned images of damaged ancient books, relating to the field of image processing technology. The method includes: importing the scanned image of the ancient book to be restored; performing OCR processing on the image to obtain OCR results; segmenting the damaged areas of the image to obtain a binary mask; constructing structured data based on the OCR results and inputting it into a location-enhanced inference model to predict the complete text at the damaged location and its predicted bounding box; performing text restoration based on the proportion of damage within the predicted bounding box and the information entropy of the corresponding text recognition results; separating the text pixels from the scanned image of the ancient book, generating background pixels at the damaged location based on the binary mask using an ancient book background restoration model, and then re-merging the restored text with the restored background; saving and outputting the final restored image. This method assists users in inference and verification throughout the entire process, improving restoration efficiency.
Need to check novelty before this filing date? Find Prior Art