Steganography in images

STEGANOGRAPHY IN IMAGES

Presented byAishwarya Korde.Roll no.: 322026

Guided byAmit Barve Sir

INTRODUCTION

MOTIVATION

TYPES OF STEGANOGRAPHY

IMAGE STEGANOGRAPHY

LSB TECHNIQUE

GRAY SCALE COVER IMAGES

EFFICIENCY ADAPTIVE STEGANOGRAPHY

DATA HIDING IN FLASH MEMORY

CONCLUSION AND FUTURE SCOPE

REFERENCES

Steganography is defined as the study of invisible communication. Steganography usually deals with the ways of hiding the existence of the communicated data in such a way that it remains confidential. Steganography is a Greek word which means concealed writing. The word “steganos” means “covered “ and “graphial “ means “writing” . Thus, steganography is not only the art of hiding data but also hiding the fact of transmission of secret data. Steganography hides the secret data in another file in such a way that only the recipient knows the existence of message. In order to safely transmission of confidential data, the multimedia object like audio, video, images are used as a cover sources to hide the data .

1)The internet allows for easy dissemination of information over large areas.

2)This is both a blessing and a curse since friends all over the world can view your information but so can everyone else. Encrypting data has been the most popular approach to protecting information but this protection can be broken with enough computational power.

3) An alternate approach to encrypting data would be to hide it by making this information look like something else. This way only friends would realize its true content.

4)In particular, if the important data is hidden inside of an image then everyone but your friends would view it as a picture. At the same time your friends could still retrieve the true information. This technique is often called data hiding or steganography

IMAGE

AUDIO

VIDEO

TEXT

NETWORK

Digital Medium to Achieve Steganography

Image steganography is method of information hiding into cover-image and generates a stego-image.

This stego-image then sent to the other party by known medium, where the third party does not know that this stego-image has hidden message.

After receiving stego-image hidden message can simply be extracted with or without stego-key (depending on embedding algorithm) by the receiving end .

Without stego-key, where embedding algorithm required a cover image with message for embedding procedure.

Output of embedding algorithm is a stego-image which simply sent to extracting algorithm, where extracted algorithm unhides the message from stego-image.

Terminologies

1) Cover-Image: Original image which is used as a carrier for hidden information.

2) Message: Actual information which is used to hide into images. Message could be a plain text or some other image.

3)Stego-Image: After embedding message into cover image is known as stego-image.

4) Stego-Key: A key is used for embedding or extracting the messages from cover-images and stego-images.

Spatial Domain Methods

Transform Domain Technique

Distortion Techniques

Masking and Filtering

It is a type of spatial domain methods.

This method is most commonly used for hiding data. In this method the embedding is done by replacing the least significant bits of image pixels with the bits of secret data.

The image obtained after embedding is almost similar to original image because the change in the LSB of image pixel does not bring too much differences in the image.

In the LSB matching, the choice of whether to add or subtract one from the cover image pixel is random.

Example:

00100111 11101001 11001000

00100111 11001000 11101001

11001000 00100111 11101011

00100111 11101000 11001000

00100110 11001000 11101000

11001001 00100111 11101011

The message embedding is performed for the two cover image pixels at a time. The gray-level values of those two pixels are x(i) and x(i+1).

After the message embedding, the value of the ithmessage bit m(i) is equal to the LSB of stegoimage’s ith pixel y(i) . The value of the (i+1)thmessage bit m(i+1) is a function of y(i) and y(i+1).

This method allows a selection of addition/subtraction of to carry information, because the selection can set a binary function(y(i),y(i+1)) to the desired value.

If a binary function has the following property

f(l-1,n) f(l+1,n) ….1

If a binary function f(l,n) is of the form

f(l,n) f(l,n+1) ….2

Then both an increase and a decrease of by n one will change the value of the function f(l,n).

For pair of pixels

input: a pair of cover image pixels x(i),x(i+1)

message bits: m(i),m(i+1)

Output: a pair of stego image pixels y(i) ,y(i+1)

If m(i)=LSB(x(i))

if m(i+1) f(x(i),x(i+1))

y(i+1) = x(i+1) +1

OR

y(i+1) = x(i+1) -1

else

y(i+1) = x(i+1)

end

y(i) = x(i)

else

if m(i+1) = f(x(i)-1,x(i+1))

y(i)=x(i)-1

else

y(i)=x(i)+1

end

y(i+1) = x(i+1)

end

1)Data Embedding Algorithm

STEP 1: Separate RGB component

STEP 2: Choose the component

STEP 3: Parameter Initialization

STEP 4: Preprocess

STEP 5:The resulting image is rearranged as a row vector V.

These vector is divided into nonoverlapping embedding units

with every two consecutive pixels, (xi ,xi+1)

where i=1,3..., m, n-1 assuming ‘n’ is an even number.

STEP 6: Encryption

STEP 7: Region Selection

EU (t) = {(xi ,xi+1) || xi - xi+1 | ≥ t,

for all (xi ,xi+1) € V }

STEP 8: Capacity EstimationT=max {2 x | EU (t)| ≥ | E(M) | }

STEP 9: Data HidingCase 1:LSB (xi) = mi & LSB (f (xi, xi+1) = mi+1then,(xi’ ,xi+1’) = (xi, xi+1)Case 2:LSB (xi) =mi & LSB (f (xi, xi+1)) ≠ mi+1 \ then(xi’ ,xi+1’) = (xi, xi+1+ r). where r = ±1.Case 3:LSB (xi) ≠ mi & LSB (f (xi-1, xi+1)) = mi+1then(xi’ ,xi+1’) = (xi-1, xi+1)Case 4:LSB (xi) ≠ mi & LSB (f (xi-1, xi+1)) ≠ mi+1then(xi’ ,xi+1’) = (xi+1, xi+1)where mi and mi+1 denote two secret bits to be embedded.

The function ‘f’ is defined as f(a,b) = (a/2)+b. r is a random value in{-1,+1} and ( xi’ , xi+1’)denotes the

pixel pair after data hiding. After the above modifications xi’ and xi+1’ may be out of [0,255], or the new difference | xi’ - xi+1’| may be less than the threshold T. In such cases, we need to readjust them as (x’’i , x’’i+1) by

x'’i = xi’ +4k1x’’i+1 = x’i+1 + 2k2k1 and k2 posses the value of either 0 or 1.we haveLSB (xi ‘’) = mi & LSB (f (x’’i , x’’i+1) = mi+1STEP 10: Post processSTEP 11: Combine RGB.STEP 12: Preset Region

Data Extraction Algorithm

STEP 1: Separate RGB component

STEP 2: Parameter Extraction

STEP 3: Preprocess

STEP 4: Region Identification

STEP 5: Data Extraction

STEP 6: Conversion of bits into ascii value

Part A – Composing the message

1) For each selected page in a block

2) Generate the group for each message bit via the page hiding key

3) Assign each group 0 or 1 according to the embedded data

4) For each bit

5) If its group will represent a message ”1”

6) Set it to be programmed 0

7) Else

8) Set it to be programmed 1

9) End if

10 )End for

11) End for

Part B – Writing the message to Flash

1) For each selected block

2) For i = 1, 2, ..,N (N is the number of Hiding PE cycles)

3) Erase the block

4) Program every selected page

5) End for

6) End for

Performance of edge adaptive steganography for color image with increasing values of parameters is analyzed. In addition to that, color image is separated into RGB layers and then data hiding is performed. The image quality after data embedding is very important for better performance of steganography methods.

When threshold is high only minimum set of characters can be embedded. As the threshold value decreases more characters can be implanted.

Separation of image into RGB component results increases the embedding capacity.

This method encrypts the message, which improves the security

This scheme can be applied to other covers such as audio and video which is taken as the future work.

Weiqi Luo, fangjun Huang and Jiwu Huang, “Edge Adaptive Image Steganography Based on LSB Matching Revisited”, IEEE transaction on Information forensics and security, Vol.5, No. 2, pp. 201-214, 2010.

J. Mielikainen, “LSB matching revisited,” IEEE Signal Process. Lett.,vol. 13, no. 5, pp. 285–287

X. Zhang and S. Wang, “Steganography using multiple-base notational

system and human vision sensitivity,” IEEE Signal Process. Lett., vol.12, no. 1, pp. 67–70, Jan. 2005.

A. Ker, et al “Improved detection of LSB steganography in grayscale images,” in Proc. Information Hiding Workshop, vol. 3200, Springer LNCS, 2004, pp. 97–115.