Digital forensics of printed source identification for Chinese characters

Min Jen Tsai, Jung Liu, Jin Sheng Yin, Imam Yuadi

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

Recently, digital forensics, which involves the collection and analysis of the origin digital device, has become an important issue. Digital content can play a crucial role in identifying the source device, such as serve as evidence in court. To achieve this goal, we use different texture feature extraction methods such as gray-level co-occurrence matrix (GLCM) and discrete wavelet transform (DWT), to analyze the Chinese printed source in order to find the impact of different output devices. Furthermore, we also explore the optimum feature subset by using feature selection techniques and using support vector machine (SVM) to identify the source model of the documents. The average experimental results attain a 98.64 % identification rate which is significantly superior to the existing known method by 1.27 %. The superior testing performance demonstrates that the proposed identification method is very useful for source laser printer identification.

Original languageEnglish
Title of host publicationDigital-Forensics and Watermarking - 12th International Workshop, IWDW 2013, Revised Selected Papers
PublisherSpringer Verlag
Pages337-361
Number of pages25
ISBN (Print)9783662438855
DOIs
Publication statusPublished - 2014
Externally publishedYes
Event12th International Workshop on Digital-Forensics and Watermarking, IWDW 2013 - Auckland, New Zealand
Duration: 1 Oct 20134 Oct 2013

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume8389 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference12th International Workshop on Digital-Forensics and Watermarking, IWDW 2013
Country/TerritoryNew Zealand
CityAuckland
Period1/10/134/10/13

Keywords

  • Digital image forensics
  • Discrete wavelet transform (DWT)
  • Feature selection
  • Graylevel co-occurrence matrix (GLCM)

Fingerprint

Dive into the research topics of 'Digital forensics of printed source identification for Chinese characters'. Together they form a unique fingerprint.

Cite this