A novel and easy method to embed any form of secret messages into a cover image with controlled distortion is proposed. Any lossy image compressor may be applied first to a cover image to produce a lossily-processed result as the basis for embedding data in the cover image. The stego-image is produced by embedding data in each pixel of a cover image by changing its gray value without excessing the range of the gray value difference of the corresponding pixels of the cover image and its lossily-processed one. The quantity of distortion that is caused by embedding data is never in excess of that is caused by the lossy compressor. A multiple-based number system is proposed to convert the information in the secret bit stream into values to be embedded in the choosing pixels of the cover image. Pseudo-random mechanisms may be used to achieve cryptography. It is found from experiments that the values of the peaks of the signal-to-noise ratio of the stego-images are larger than those yielded by the chosen compressor, which means that the distortion in the embedding result is more imperceptible than that in the compressed one.
[1]
Allen Gersho,et al.
Vector quantization and signal compression
,
1991,
The Kluwer international series in engineering and computer science.
[2]
Ronald A. DeVore,et al.
Image compression through wavelet transform coding
,
1992,
IEEE Trans. Inf. Theory.
[3]
Eckhard Koch,et al.
Towards Robust and Hidden Image Copyright Labeling
,
1995
.
[4]
Ross Anderson,et al.
Information hiding terminology
,
1996
.
[5]
Walter Bender,et al.
Techniques for Data Hiding
,
1996,
IBM Syst. J..
[6]
Ingemar J. Cox,et al.
Secure spread spectrum watermarking for multimedia
,
1997,
IEEE Trans. Image Process..
[7]
Jian Zhao,et al.
In business today and tomorrow
,
1998,
CACM.
[8]
Ahmed H. Tewfik,et al.
Multiresolution scene-based video watermarking using perceptual models
,
1998,
IEEE J. Sel. Areas Commun..
[9]
Wenjun Zeng,et al.
Image-adaptive watermarking using visual models
,
1998,
IEEE J. Sel. Areas Commun..