Contents

1. Introduction2
1.1 Problem Specification..2
1.2 Objective..2
2. Literature review....4
2.1 Image....4
2.2 MATLAB.6
2.3 Secure Image Transmission..7
3. Design9
3.1 Mosaic Image Creation.. 10
3.2 Image Recovery. 11
4. Partial Results... 15
5. Conclusion. 24
References... 26

1. INTRODUCTION
Currently, images from various sources are frequently utilized and transmitted through the
internet for various applications, such as online personal photograph albums, confidential
1

enterprise archives, document storage systems, medical imaging systems, and military image
databases. These images usually contain private or confidential information so that they should
be protected from leakages during transmissions.
The common approaches for secure image transmission are
1. Image Encryption
2. Data Hiding
Image Encryption is a technique where the data is modified using some key. The data is
modified according to the technique. In this approach there is a possibility of hacking the data
because the hacker may observe the pattern of change in the data.
To avoid that problem data hiding technique is being used. In data hiding technique the
secret image is being hidden into a cover image so that no one can realize the existence of data
behind that. The various data hiding techniques are LSB substitution, Histogram shifting,
prediction error expansion, recursive histogram modification, discrete cosine transformation.
All the above mentioned data hiding techniques results in distortion of the images.

1.1 Problem Specification

In all the above techniques large amount of data can not be embedded into a single image. If
we want to hide a secret image into a cover image with same size, we must compress the secret
image priorly. But this results into distortions in case of medical images, military images etc

1.2 Objective
The objective is to transmit the images securely with the same size and provide
confidentiality to the images.
For this purpose we transform the secret image into a meaningful mosaic image with the
same size and which looks similar as a target image which is selected priorly.
At first we select an image for transforming the secret image into that image. The
preselected image is called as Target Image. The secret image is divided into tiles and the target
image is divided into blocks. Color transformations are done on the tiles so as to resemble as the
blocks. Later one-one mapping is done and a mosaic image is formed.
2

Some information is embedded into the mosaic image for better security and also to recover
the image losslessly.

2. LITERATURE SURVEY
2.1 Image

An image is essentially a 2-D signal processed by the human visual system. The signals
representing images are usually in analog form. However, for image processing, storage and
transmission, they are converted from analog to digital form. A digital image is basically a 2-D
array of pixels. Images are formed of the significant part of data, particularly in remote sensing,
biomedical and video conferencing applications. The use of and dependence on information and
computers continue to grow, so does our need for efficient ways of storing and transmitting large
amounts of data.
2.1.1 Pixel
In digital image, a pixel is a single point in a raster image. It is the smallest unit of picture
that can be controlled, and is the smallest addressable screen element as shown in Fig. 2.1. Each
pixel has its own address. The address of a pixel corresponds to its coordinates. They are usually
arranged in a 2-D grid, and are often represented with dots or squares. Each pixel is a sample of
an original image. More samples typically provide more accurate representations of the original.
The intensity of each pixel is variable. In color image systems, a color is typically represented by
three or four component intensities such as red, green, and blue.

Figure 2.1 Pixel is smallest element of an image

2.1.2 RGB
When the eye perceives an image on a computer monitor, it is in actually perceiving a
large collection of finite color elements, or pixels. Each of these pixels is in and of itself
composed of three dots of light; a green dot, a blue dot, and a red dot. The color the eye
perceives at each pixel is a result of varying intensities of green, red, and blue light emanating
from that location. A color image can be represented as 3 matrixes of values, each corresponding
to the brightness of a particular color in each pixel. Therefore, a full color image can be
reconstructed by superimposing these three matrices of RGB, as shown in Fig. 2.2.
4

Figure 2.2: A color image is made of 3 matrices

2.1.3 GRAYSCALE
If an image is measured by an intensity matrix with the relative intensity being represented
as a color between black and white, it would appear to be a grayscale image, as shown in Fig. 2.3.

Figure 2.3: A grayscale image

The intensity of a pixel

is expressed within a given range

between a minimum and a

maximum.

represented in a range from 0 to

with

This

range

incremental

is
steps

according to the bit-depth of the image. The greater the number of steps is, the larger the bitdepth is. If each intensity value is represented as an 8-bit number, then there are 256 variations.
If the intensity values are represented as 16-bit numbers, there are 32,000 variations between
absolute black and pure white. Fig. 2.4 demonstrates a black to white gradient in 4-bits of
intensity.

Figure2.4: bit black to white gradient

2.2 MATLAB
MATLAB is a multi-paradigm numerical computing environment and fourth generation
programming language. A proprietary programming language developed by MATHWORKS.
MATLAB allows matrix manipulations, plotting of functions and data, implementation of

algorithms, creation of user interfaces, and interfacing with programs written in other languages
including C, C++, Java, Fortran and Python.

2.2.1 Image representation in MATLAB

An image is represented as a m*n matrix where m represents number of rows and n
represents number of columns. Every element in a matrix is called a pixel and the value
represents its intensity at that particular position.
Some of the basic MATLAB functions are
Imshow(): this command shows our image in standard 8 bit format
Eg: imshow(img);
Imread(): this command will read an image into matrix.
Syntax:- var=imread(file_name);
Eg: img=imread(a.png):
Rgb2gray(): this command converts color image into gray scale image.
Syntax;- i=rgb2ray(RGB image);
Eg; img=imread(a.png);
Gray=rgb2gray(img);
Imshow(gray);

2.3 Secure Image Transmissions

There are many digital images are present in digital communication system which being
sent over computer networks. With the increasing growth of multimedia applications like audio,
video and images. Security is an important aspect in digital communication. There is one of the
obvious way to ensure security is Image encryption.

This technique converts original image to another image which is hard to understand and
to keeps the image confidential between users, in other word, its important that without
decryption key no one can access the content. Image encryption has applications in internet
communication, multimedia systems, medical imaging, telemedicine, military communication
etc. For privacy protection of digital images, encrypted databases is an important technological
capability in multiparty information management.
And the other is data hiding. In data hiding we cover our image with other image so that
no one can realize the existence of other image behind that cover image.

2.3.1 Works on Secure Image Transmission Techniques

W. Zhang, X. Hu, X. Li, and N. Yu,[4] have proposed a high capacity reversible image
data hiding scheme based on a generalization of prediction-error expansion (PEE) and an
adaptive embedding strategy. For each pixel, its prediction value and complexity measurement
are firstly computed according to its context. Then, a certain amount of data bits will be
embedded into this pixel by the proposed generalized PEE. Here, the complexity measurement is
partitioned into several levels, and the embedded data size is determined by the complexity level
such that more bits will be embedded into a pixel located in a smoother region. The complexity
level partition and the embedded data size of each level are adaptively chosen for the best
performance with an advisable parameter selection strategy. In this way, the proposed scheme
can well exploit image redundancy to achieve a high capacity with rather limited distortion.
The major drawbacks are low embedding capacity and its inability to control the capacity.
C. C. Chang, C. C. Lin, C. S. Tseng, and W. L. Tai[5] have proposed a robust reversible
data hiding scheme by exploiting the differences between a center pixel and its neighboring
pixels in each sub-block of the image to embed secret data into extra space. Moreover, scheme
enhances the embedding capacity and can recover the embedded data from the stegno image
without causing any perceptible distortions to the cover image.
C. H. Yang[1] have proposed an optimal pixel adjustment process (OPAP) to enhance the
quality of the stego-image by the simple LSB substitution method through raster scan. In this
method, the OPAP tries to vary the value of Most Significant Bit (MSBi) next to kth bit up to
which the secret data is embedded. C. H. Yang has proposed an LSB substitution method using
7

raster scan to improve the stego image quality by adapting an inverted pattern (IP) approach. In
this technique, secret message has been processed prior to embedding .
The main disadvantage of this is its low embedding capacity.

3. DESIGN
In the design phase actual flow of the process will be shown. In this technique there are two
phases. One is the creation of mosaic image and the other is image recovery.
The flow of the process can be shown in the below figure

Target Image

1.

Mosaic creation

Secret image
1.2 Transform the color
characteristics of the
image blocks

1.1 fit the images into

target blocks

1.3 rotate the images

into directions with
minimum RMSE

1.4 Embed relevant

information

Mosaic
image

2. Image Recovery
2.2 recover secret
image

2.1 extract the

eembeded information

Fig 3.1 Design of the process

3.1 Mosaic Creation

Target image

Secret image

Fit the images

into target blocks

Transform the
color
characteristics

Rotate the
blocks in
direction with
minimum
RMSE
Embed relevant
information

Mosaic image

Fig 3.2 Flow of mosaic image creation

3.2 Image Recovery

Mosaic image

10

Extract the
embedded
information

Recover the secret

image

Secret image

Fig 3.3 Flow of secret image recovery

The design includes two main phases

1. Mosaic image creation
2. Secret image recovery

1. Mosaic Image Creation

This phase includes four stages. They are
i.

Fitting the tile images of the secret image into the target blocks of the pre-selected target

ii.

image
Transforming the color characteristics of each tile image to become that of the

iii.

corresponding target block

Rotate each tile into a direction with minimum RMSE value with respect to its

iv.

corresponding target block

Embed the relevant information into the created mosaic image for future recovery

2. Secret Image Recovery

This phase includes two stages. They are

11

i.
ii.

Extracting the embedded information

Recovering the secret image

Algorithm flow
Phase 1
Stage 1 . Fitting the tile images into the target blocks.
Step 1: Here first we need to compare the sizes of the target and secret image sizes if they are
not equal we need to resize and equalize them and divide the secret image into tile
images {T1,T2,T3.........................Tn} and also the target image as {B1,B2BN}
and each Ti, Bj belongs to size of Nt.
Step 2: Then calculate both mean and standard deviation for each tile image

Ti

and target

image
Bj

respectively.

c =

1
c (3.1)
n i=1 i

'c =

1
c , (3.2)
n i=1 i

c = 1/n ( c ic )
i=1

.. ..(3.3)

Step 3: Now we have the set of tile images as

S title ={T 1 .T 2 ,T 3 T n

} and target blocks

as

12

S target =

B 1 , B 2 B3 . B N

} then by sorting of this two according to the mean

and
standard deviation values we need to map the two tile image set to the target blocks in
1-to-1 manner then resulting mapping sequence L of the form T1

Bj1.TN Bjn

Step 4: so create the mosaic image F by fitting the tile images into the corresponding blocks
according to L.
Stage 2: performing color transformation
For color transformation we use

} = {q} rsub {c} left ({c} rsub {i} - {} rsub {c} right ) + {} rsub {c} rsup {'}

ci
. (3.4)
Stage 3: Rotating Images
For each colored transformed tile image
corresponding target block

B Ji

T1

calculate the RMSE values in F with respected to

after rotating

Ti

into directios of

=0,90,180,270,360

respectively.
Stage 4: embedding
Now we embed the relevant information into the image.
Construct a bit stream Mi including
1.
2.
3.
4.

Index of corresponding target block

Optimal rotational angle
Means and standard deviations of all the color channels
Key

Perform the XOR operation of 1, 2, 3 with 4 and embed in the image

13

PHASE 2: secret image retrieval

Stage1: extracting the secret image recovery information.
Step 1: extract the bit stream I from the F by reversion scheme and decode them
Step 2: Decode

Mi

for each tile image

T1

to obtain the following data items: 1) the index

ji of
the block

B ij

in F corresponding to

Ti

2) the optimal rotation angle of

Ti

; 3)

the means
of

Ti

and

B ij

and the related standard deviation quotients of all color channels.

Stage2: recovering secret image

Step 1: rotate in the reverse direction the block indexed by j i, namely Bij in F through the
optimal angle and fit the resulting block content into
image

Ti

to form an initial tile

Ti

Step2: use the extracted means and related standard deviation quotients to recover the original
pixel values in

Ti

Step 3: combine the all final tile images to get desired secret image T.

14

4. RESULTS
1.

Selecting the secret image

Fig4.1. Secret image

The intensity values of the above image are

15

//* red component

//* green
component

Fig 4.2 intensities of red, green colors of secret image

//* blue
component

Fig4.3. Pixel intensities of blue color of secret image

Now we select the target image

16

Fig4.4. Target image

The pixel intensities of the target image are

//* red component

//* green component

//*blue component

17

Fig4.5. Pixel intensities of red, green and blue colors of target image

Divide the secret image into tiles with 4*4as the size

Fig4.6. Pixel intensities of red, blue, green colors of tile1 respectively

18

Fig4.7. Pixel intensities of red, blue, green colors of tile2 respectively

Fig4.8. Pixel intensities of red, blue, green colors of tile3 respectively

19

Fig4.9. Pixel intensities of red, blue, green colors of tile4 respectively

Dividing the target image into blocks with 4*4 size each

Fig4.10. Pixel intensities of red, blue, green colors of block1 respectively

20

Fig4.11. Pixel intensities of red, blue, green colors of block2 respectively

Fig4.12. Pixel intensities of red, blue, green colors of block3 respectively

21

Fig4.13 Pixel intensities of red, blue, green colors of block 4 respectively

Fig 4.14 standard deviation values of tiles

In order to get the standard deviation values we first calculate the standard deviation values of
every color component and then we calculate the mean of standard deviation values of those
components of each tile.
Mean is calculated by using the equation 3.1 and standard deviation is calculated by using the
equation 3.3

22

Fig 4.15 standard deviation values of blocks

The process which is used for calculating the standard deviation values of tiles is repeated for
blocks in order to get the standard deviation values for blocks.

Fig 4.16 Ascending order of standard deviation values of tiles and their indices

Fig 4.17 Ascending order of standard deviation values of blocks and their indices
23

5. CONCLUSION
For providing security to the images data hiding technique is being used where a mosaic image is
created from target image and secret image.
In this process first step of mosaic image creation is done i.e., diving the images into
tiles and blocks and calculating their standard deviation values is done.
Further steps of the process has to be implemented.

24

REFERENCES
[1] C. H. Yang Inverted Pattern Approach to improve image quality of information hiding by
LSB substitution, Pattern Recognition., vol. 41 no. 8. Pp. 2647-2683, 2008.
[2] C. k. Chan and L. M. Cheng. Hiding data in images by LSB substitution, Pattern
Recognition vol 37, pp.469-47 Mar. 2004.
[3] J. Lai and W.H.Tsai, , Secret-fragment-visible mosaic imageA new computer art and its
Application to information hiding,IEEE Trans. Inf. Forens. Secur., vol.6, no. 3, pp. 936945, Sep. 2011.
[4] W. Zhang, X. Hu, X. Li, and N. Yu, Recursive histogram modification:Establishing
Equivalency between reversible data hiding and lossless data compression, IEEE Trans.
Image Process., vol. 22, no. 7,pp.2775-2785, Jul. 2013.
[5] C. C. Chang.C. C. Linc, C. S. Tseng, and W. L. Tai, Reversible hiding in DCT-based
Compressed images, Inf. Sci., vol. 177, no. 13, pp. 2768-2786, 2007.
25

26