Font Size: a A A

Research On Traceability Technology Of Paper Documents

Posted on:2021-09-30Degree:MasterType:Thesis
Country:ChinaCandidate:H WangFull Text:PDF
GTID:2518306308969769Subject:Computer technology
Abstract/Summary:PDF Full Text Request
With the development of informatization and digitization in all works of life,more and more information is stored,modified,and transmitted in digital forms,but most important documents still need to be printed and kept in paper files for backup.Some confidential documents also need to be delivered by paper,so the protection and traceability of paper documents is particularly important.In addition,in the field of copyright protection for paper works,it is also necessary to combat piracy and infringement by being able to trace the origin of paper works.This paper mainly studies the traceability of paper documents,designs and implements the traceability system based on a variety of text watermarking algorithms.The main works and achievements are as follows:1.Aiming at the influence of deformation and distortion during the propagation process,we analyze the characteristics of paper files that do not change with printing,scanning and other operations,and study the use of traditional digital watermarking technology to achieve the traceability of paper files,that is,an anti-printing scanning watermarking algorithm based on document structure and character pixels.The relative width of the line spacing in the file and the relative black pixel value of each character are encoded with invariable features.Finally,the traditional watermarking algorithm based on the characteristics of these two files is combined to implement a preliminary paper file traceability tracking system.2.After analyzing the limitations of the above-mentioned traceability tracking system using traditional watermarking algorithms in terms of embedded capacity,a watermarking algorithm based on deep learning is proposed to achieve the traceability of paper documents,that is,a watermarking algorithm based on fusion fonts.The algorithm uses deep learning technology to realize the style transfer of font images,generate fusion fonts that human eyes cannot distinguish,and then establish a mapping relationship with embedded traceability information.This algorithm not only has high robustness,but also greatly improves the capacity of embedded traceability information.3.Based on the research of fused fonts digital watermarking algorithm,the corresponding traceability tracking system is designed and implemented.The system can embed more traceability information in paper documents,and improve the system’s robustness and traceability performance.Through the comparative analysis of performance,it also proves that the system can better fulfill the requirements of traceability of paper documents.
Keywords/Search Tags:Traceback, text watermarking, anti-print scanning, deep learning
PDF Full Text Request
Related items