Optimum histogram pair based image lossless data embedding

Guorong Xuan, Yun Q. Shi, Peiqi Chai, Jianzhong Teng, Zhicheng Ni, Xuefeng Tong

Research output: Chapter in Book/Report/Conference proceedingConference contribution

19 Scopus citations

Abstract

This paper presents an optimum histogram pair based image reversible data hiding scheme using integer wavelet transform and adaptive histogram modification. This new scheme is characterized by (1) the selection of best threshold T, which leads to the highest PSNR of marked image for a given payload; (2) the adaptive histogram modification, which aims at avoiding underflow and/or overflow, is carried out only when it is necessary, and treats the left side and right side of histogram individually, seeking a minimum amount of histogram modification; and (3) the selection of most suitable embedding region, which attempts to further improve the PSNR of marked image in particular when the payload is low. Consequently, to our best knowledge, it can achieve the highest visual quality of marked image for a given payload as compared with the prior arts of image reversible data hiding. The experimental results have been presented to confirm the claimed superior performance.

Original languageEnglish (US)
Title of host publicationTransactions on Data Hiding and Multimedia Security IV
EditorsYun Q. Shi
Pages84-102
Number of pages19
DOIs
StatePublished - 2009

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume5510 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

All Science Journal Classification (ASJC) codes

  • Theoretical Computer Science
  • General Computer Science

Keywords

  • Adaptive histogram modification
  • Integer wavelets
  • Optimum histogram pair
  • Reversible (lossless) data embedding
  • Selection of best threshold
  • Selection of suitable embedding region

Fingerprint

Dive into the research topics of 'Optimum histogram pair based image lossless data embedding'. Together they form a unique fingerprint.

Cite this