VRL an eye

Joint Source-Channel Hiding

People

K. Solanki, O. Dabeer, U. Madhow, B.S.Manjunath, S. Chandrasekaran

Objective

We consider the problem of image-in-image hiding in this project, where, the basic design criteria are as follows:

  1. Perceptual criteria: The degradation to the host image is imperceptible.
  2. Robustness: It should be possible to recover the hidden, or signature, image under a variety of attacks.
  3. Graceful improvement:  The quality of the recovered signature image should be better if the attack is milder.

In recent work (see image-adaptive data-hiding project), it has been shown that digital data can be effectively hidden in an image so as to satisfy criteria (1) and (2) by hiding in the choice of quantizer for the host data. Unfortunately, these schemes do not satisfy the design criterion (3) - they exhibit the threshold effect: if the actual attack is more severe than the attack the scheme was designed for, there is a catastrophic failure in recovering the hidden image, while if the actual attack is less severe, then we are still stuck with the design attack image quality. In practice, the attack level is seldom known apriori, and ideally, we would like a scheme that results in graceful improvement and degradation in the image quality with less and more severe attacks respectively. In this project, our objective is to propose a method that allows graceful improvement of the hidden image using joint source-channel coding principles.

Joint Source-Channel Coding for Image-in-Image Hiding

We have designed a simple joint source-channel coding technique to achieve graceful improvement in the recovered image quality with milder attacks. The signature image is compressed efficiently (using JPEG) into a sequence of bits, which is hidden using the scheme described in the image-adaptive data-hiding project. The residual error between the original and compressed signature image is then hidden using a new analog information hiding technique. The results show perceptual as well as mean-square error improvement in the recovered signature image as the attack becomes milder. See the presentation below for more details.

» Presentation: "A Joint Source-Channel Coding Scheme for Image-in-Image Data Hiding"

Examples: Gracefully Improving Image-in-Image Hiding

Hiding 128x128 Peppers Image in
512x512 Harbor Image

(Designed to survive JPEG compr. with QF = 25 onwards)

 

Hiding 256x256 Lenna Image in
512x512 Bridge Image

(Designed to survive JPEG compr. with QF = 50 onwards)

 
Recovered signature images at varying attack levels   Recovered signature images at varying attack levels

Acknowledgements

This research is supported by a grant from ONR #N00014-01-1-0380. Program manager: Dr. Ralph Wachter

Publications

    2003

  1. K. Solanki, O. Dabeer, U. Madhow, B. S. Manjunath and Shiv Chandrasekaran,
    "Robust Image-Adaptive Data Hiding: Modeling, Source Coding, and Channel Coding"
    41st Allerton Conference on Communications, Control, and Computing, Oct. 2003.
    VRL ID: 125, [abstract] [PDF] [BibTex]

    Abstract preview: "This paper provides a summary of our work over the past two years on robust, high-volume data hiding in images. We first present a basic framework for image-adaptive hiding, which allows selection of ..." [more]

  2. K. Solanki, O. Dabeer, B. S. Manjunath, U. Madhow and Shiv Chandrasekaran,
    "A Joint Source-Channel Coding Scheme for Image-in-Image Data Hiding"
    IEEE International Conference on Image Processing (ICIP), Barcelona, Spain, vol. 2, pp. 743-746, Sep. 2003.
    VRL ID: 119, [abstract] [PDF] [BibTex]

    Abstract preview: "We consider the problem of hiding images in images. In addition to the usual design constraints such as imperceptible host degrada-tion and robustness in presence of variety of attacks, we impose the ..." [more]