Watermarking of hiding or embedding an imperceptible signal

Watermarking
has been invoked as a tool for the protection of Intellectual Property Rights
(IPR) of multimedia contents. Because of their digital nature, multimedia
documents can be duplicated, modified, transformed, and diffused very easily.
In this context, it is important to develop a system for copyright protection,
protection against duplication, and authentication of contents. For this, a
watermark is embedded into the digital data in such a way that it is
indissolubly tied to the data itself. Later on, such watermark can be extracted
to prove ownership to trace the dissemination of the marked work through the
network. In this proposed scheme mainly focus on the transform domain,
particularly Discrete Wavelet Transform (DWT) and Singular Value Decomposition
(SVD). Apart from this, also protect the copyright of virtual views which is
generated by using a technique called Depth-Image-Based-Rendering. DIBR has
become an important technology in 3D content representation. In this technique,
watermark is embedded in the center view using two mathematical transform, DWT
and SVD. Then, the left and right views are generated automatically from the
watermarked center view and their corresponding depth map at the receiver side.
Finally watermark can be extracted from the center, left or right views in a
blind fashion without using the original unwatermarked center, left or right
views. In this watermarking schemes are robust against image compression, noise
addition and geometric distortion such as rotation, scaling, and cropping.
Moreover, the proposed scheme has good performance in terms of depth image variation
and baseline distance adjustment.

Keywords—3D
video, blind watermarking, depth-image-based-rendering (DIBR), discrete wavelet
transfrom (DWT), singular value decomposition (SVD), robustness.

We Will Write a Custom Essay Specifically
For You For Only $13.90/page!


order now

                                                                                                                                                     
I.      Introduction

Due to the rapid development of the 3D display market, the
protection and authentication of the intellectual property rights of 3D
multimedia data has become an essential concern. Without adding security
information it impossible to automatically verify the authenticity of the
uploaded multimedia objects. So the digital watermarking is used to protect the
information against the illegal changes in the form of images, videos and
audios. Watermarking is the process of hiding or embedding an imperceptible
signal (data) into the given signal (data). This imperceptible signal (data) is
called watermark or metadata and the given signal (data) is called cover work.
This cover work can be an image, audio or video file. Video Watermarking is a
young and rapidly evolving field in the area of multimedia.  It refers to embedding watermarks in a video
sequence in order to protect the video from illegal copying and identify
manipulation. Video is becoming more important in different applications such
as, video broadcast, video surveillance, video on demand, video conferencing,
where two factors of video data are very important i.e., authenticity and
integrity.

In this paper, propose an imperceptible and robust video
watermarking algorithm based on cascading two mathematical transforms; the
Discrete Wavelet Transform (DWT), and the Singular Value Decomposition (SVD).
The proposed algorithm is also ‘blind’ in the sense that it does not require
the original data to extract the embedded watermark. Relevance of DIBR
technique to video watermarking is described in the next section. The proposed
DWT-SVD watermarking algorithm is described in section three. Concluding
remarks are given in the last section.

 

                                                                                                                    
II.    DEPTH-IMAGE-BASED-RENDERING

In this section, the DIBR 3D image production chain is
described. It consists of three stages: image capturing, 3D warping, and the
post-processing operation of hole-filling. In depth-image-based-rendering is
the process of synthesizing virtual views of the scene captured from color
images or video associated with depth information.

A.    Pre-Processing
of Depth Image

As the first step, pre-processing of the depth map by
smoothing filter can reduce the hole occurrences in the virtual left and right
views, especially in the depth regions with sharp edges and
discontinuities.  In this process, the
gray values of an 8-bit depth image. After that, different kinds of filters can
be adopted to smooth the depth image. For efficient implementation, Gaussian
filter is employed in the proposed scheme because of its adjustable window
size.

B.    3D
Image Wraping

Three-dimensional image warping
is very important in generating the virtual left and right views. In the image
warping step, the virtual left and right view can be synthesized from the
center view and its corresponding depth map by the values of baseline distance
and focal length. 

 

C.    Hole
Filling

Due to the sharp
changes in the depth map and different viewpoints, there are new exposed hole
areas revealed in the rendered left and right views after 3D warping process.
It is because that some new pixels appeared in the virtual image are occluded
by the foreground object in the center image. The hole occurrences can be
reduced by pre-processing of the associated depth map.

 

 

                                                                                       
III.   PROPOSED
DWT-SVD BASED BLIND WATERMARKING

First, the input image is transformed using a transform
such as the DWT. The watermark data is embedded to a transformed coefficient.
Finally, inverse transform is performed on the transformed watermarked image.
Second, the watermark data is extracted from the watermarked image. The
watermarked and the original image are transformed using the DWT. The
transformed image is subtracted from the transformed watermarked image.

Discrete wavelet transform (DWT) is any wavelet transform
for which the wavelets are discretely sampled. It can exploit the
characteristics of the HVS when compared to DCT and DFT; it is possible to hide
watermarks with more energy in an image, which makes watermarks more robust.
The DWT separates an image into a lower resolution approximation image (LL) as
well as horizontal (HL), vertical (LH) and diagonal (HH) detail components.
These attractive properties are not sufficient since the DWT itself does not
exhibit the shift invariance property.

On the other hand, the SVD transform is a numerical
technique for diagonalizing matrices and also a factorization of a real or
complex matrix. In SVD-based watermarking, an image is treated as a matrix
decomposed by SVD into the three matrices; U, S and V. By virtue of the fact
that slight variations in the elements of matrix S does not affect visual perception
of the quality of the cover image. These transform also possess the
shift-invariance which is a property required for the development of effective
DIBR watermarking schemes.