Reversible JPEG-based hiding method with high hiding-ratio

Lee Shu Teng Chen, Sian Jheng Lin, Chih-Ching Lin

Research output: Contribution to journalArticlepeer-review

8 Scopus citations

Abstract

The stego-images generated by many existing hiding techniques are not economic in size, and hence need compression. Unfortunately, compression usually destroys the secret content hidden inside. To solve this dilemma, some hiding methods based on compression code (rather than the image itself) are reported. This paper proposes a high-capacity and high-hiding-ratio "reversible" steganography method based on JPEG-compression code. In the proposed method, the JPEG compression code of an image is used as the cover media. An 8×8 hiding-capacity table is firstly evaluated, which is then utilized to modify the quantization table attached to the given JPEG code. The two quantization tables (modified and original) together can map the DCT coefficients of each block to some larger DCT coefficients, with secret data hidden inside these larger DCT coefficients. In the decoding process, after lossless extraction of the hidden secret data, the proposed method can also recover the original JPEG-compression code. Experimental results show that our method outperforms other JPEG-based hiding methods (reversible or not) regarding both hiding-ratio and stego-image's quality.

Original languageEnglish
Pages (from-to)433-456
Number of pages24
JournalInternational Journal of Pattern Recognition and Artificial Intelligence
Volume24
Issue number3
DOIs
StatePublished - 1 May 2010

Keywords

  • JPEG
  • Reversible data hiding
  • hiding ratio
  • stego code

Fingerprint Dive into the research topics of 'Reversible JPEG-based hiding method with high hiding-ratio'. Together they form a unique fingerprint.

Cite this