Reversible Watermarking Based on Invariant Image
Classification and Dynamic Histogram Shifting
Abstract
We propose a new reversible watermarking scheme. One first contribution is a histogram shifting
modulation which adaptively takes care of the local specificities of the image content.By applying it to
the image prediction-errors and by considering their immediate neighborhood,the scheme we propose
inserts data in textured areas. This classification is based on a reference image derived from the image
itself, a prediction of it, which has the property of being invariant to the watermark insertion. our method
can insert more data with lower distortion than any existing schemes.
In this project, we propose a new reversible watermarking scheme. One first contribution is a histogram
shifting modulation which adaptively takes care of the local specificities of the image content. By
applying it to the image prediction-errors and by considering their immediate neighborhood, the scheme
we propose inserts data in textured areas where other methods fail to do so. Furthermore, our scheme
makes use of a classification process for identifying parts of the image that can be watermarked with the
most suited reversible modulation. This classification is based on a reference image derived from the
image itself, a prediction of it, which has the property of being invariant to the watermark insertion. In
that way, the watermark embeddeder and extractor remain synchronized for message extraction and
image reconstruction.
The proposed scheme makes use of a classification process for identifying parts of the image that can be
watermarked with the most suited reversible modulation of histogram shifting.The Matlab will be used
for project development.
Introduction
Reversible watermarking was a mile stone in the development of secure data hiding and digital
watermarking. Several methods have been proposed to protect highly sensitive images like military
images and medical images. The main feature of reversible data hiding is that both the hided data as well
as the original medium before hiding is recoverable correctly. There are mainly three classifications for
reversible data hiding: Expansion embedding, Histogram shifting, a combination of both.
Existing System
Several reversible watermarking schemes have been proposed for protecting images of sensitive content,
like medical or military images, for which any modification may impact their interpretation. These
methods allow the user to restore exactly the original image from its watermarked version by removing
the watermark. Thus it becomes possible to update the watermark content, as for example security
attributes (e.g., one digital signature or some authenticity codes), at any time without adding new image
distortions. However, if the reversibility property relaxes constraints of invisibility, it may also introduce
discontinuity in data protection. In fact, the image is not protected once the watermark is removed. So,
even though watermark removal is possible, its imperceptibility has to be guaranteed as most applications
have a high interest in keeping the watermark in the image as long as possible, taking advantage of the
continuous protection watermarking offers in the storage.
Limitations:
 Not efficient.
 Image is not protected in correct way.
 Allows discontinuity in data protection.
Proposed System
Our scheme relies on two main steps. The first one corresponds to an “invariant “classification process for
the purpose of identifying different sets of image regions. These regions are then independently
watermarked taking advantage of the most appropriate HS modulation. From here on, we decided
distinguishing two regions where HS is directly applied to the pixels or applied dynamically to pixel
prediction-errors respectively.
We will refer the former modulation as PHS (for “Pixel Histogram Shifting”) and the later as DPEHS (for
“Dynamic Prediction-Error Histogram Shifting”).Our choice is based on our medical image data set, for
which PHS may be more efficient and simple than the DPEHS in the image black background,while
DPEHS will be better within regions where the signal is non-null and textured (e.g., the anatomical
object). In the next section we introduce the basic concept of the invariance property of our classification
process before detailing how it interacts with PHS and DPEHS. We also introduce some constraints we
imposed on DPEHS in order to minimize image distortion and then present the overall procedure.
Advantages:
 It provides robustness

The image is well protected.

Better pixel prediction.

Directly applying HS on pixels may be more efficient and of smaller complexity than applying it
on prediction-errors.

The watermark embedder and extractor remain synchronized because the extractor will retrieve
the same reference image. Herein, we adapt this process to select the most locally appropriate
watermarking modulation.
Project Block Diagrams
Embedding process
Extraction process
Algorithm Details
(A)Embedding Procedure
Input: Original 8 bit grayscale image I, with MxN pixels and watermark Iw.
Output: Watermarked image Iw, the peak point a, the minimum point b, length of watermark and the
location map L.
Algorithm Steps:
Step 1: Scan the image I and construct it’s histogram H(x)N[0, 255]. In this histogram obtain peak point a
and less point b which is equal to (a-1).
Step 2: Record the position of pixel values whose values lies between point a and b.
Step 3: Scan the cover image I again. Set counter k for length of watermark.
If counter k is less than length of watermark
(a) If scanned pixel value lies within a and b, decrease it by 1.
(b) If pixel value lies below ‘b’ and, then don’t change that pixel value.
(c) If pixel values lies above ‘a’ then don’t change that pixel value.
(d) Scan the watermark, if scanned value is 0, then decrease pixel value of a by 1. If scanned pixel value
of watermark is 1, then do not decrease pixel value.
Step 4: Continue step 3 upto end of watermark. If counter k becomes greater than length of watermark, do
not change any value upto end of image scanning completes.
(B)Extraction and Restoration Procedure
Input: Watermarked Image Iw, the peak point a, the minimum point b, the location map L and the length
of the watermark Iw.
Output: Original 8 bit grayscale image I and the recovered watermark Iw.
Step 1: scan the image in the same order as in the embedding phase.
Step 2: Set counter k=0, k is used to indicate length of watermark. For k is less than length of watermark,
go to step 3 else step 4.
Step 3: (a) If image scanned pixel value is ‘a’ bit 1 is extracted. If the scanned value is a-1, extract 0 bit,
increase pixel value by (n-1), and increase counter k=k+1.
(b)If scanned pixel value lies between a and b then add ‘1’ in the scanned pixel value (optional).
(c) (i) If pixel value is greater than ‘a’ and lesser than ‘b’ then do not change these values.
Step 4: Continue step 3 upto end of watermark. If counter k becomes greater than length of watermark, do
not change any value upto end of image scanning completes.
Step 5: Go to location map L of b+1, and make it b+1.
Project Development Details
Project Systems Area Domain : Image processing
Tool
: Matlab 7.10(R2010a)
Operating System
: 32 bit Windows XP or 7
Conclusion
We will use Matlab for development. In this project, we will propose a new reversible watermarking
scheme which originality stands in identifying parts of the image that are watermarked using two distinct
HS modulations: Pixel Histogram Shifting and Dynamic Prediction Error Histogram Shifting (DPEHS).
Reference Papers
1. Gouenou Coatrieux, Member, IEEE, Wei Pan, Nora Cuppens-Boulahia, Member, IEEE, Frédéric
Cuppens, Member, IEEE, and Christian Roux, Fellow, IEEE “Reversible Watermarking Based on
Invariant Image Classification and Dynamic Histogram Shifting” -IEEE TRANSACTIONS ON
INFORMATION FORENSICS AND SECURITY, VOL. 8, NO. 1, JANUARY 2012.
2. G. Coatrieux, C. Le Guillou, J.-M. Cauvin, and C. Roux, “Reversible watermarking for knowledge
digest embedding and reliability control in medical images,” IEEE Trans. Inf. Technol. Biomed., 2009
Mar.,13(2):158-165.
3. F. Bao, R. H. Deng, B. C. Ooi, and Yanjiang Yang, “Tailored reversible watermarking schemes for
authentication of electronic clinical atlas”,IEEE Transactions on Information
4.L. Kamstra and H. J.A. M. Heijmans, ”Reversible data embedding into images using wavelet techniques
and sorting,” IEEE Transactions on Image Processing, vol. 14, no. 12, pp. 2082-2090, 2005.
5. Lixin Luo, Zhenyong Chen, Ming Chen, Xiao Zeng, and Zhang Xiong, “Reversible Image
Watermarking Using Interpolation Technique”, IEEE Transactions on Information Forensics and
Security, vol. 5, no. 1, pp. 187-193, mars 2010.