3. Steganography !
encrypt msg cover file
watermarking
Advantages
secrecy
Imperceptibility
High capacity
Accurate extraction
files such as audio,video,image,etc..
5. Classical Methods
Kirchoff’s golden rule
Adaptive Steganography
4LSB substitution technique
Model authentication technique
6. Objective of proposed method
Enhance embedding capacity of image .
An adaptive number of least significant bits
substitution method with private stego-key.
verify whether the attacker has tried to
modify the secret hidden information in the
stego-image.
embed the hidden information in the cover
image and use digital signature using a key
to verify the integrity from the stego-image.
7. Method
key1 consists of some gray-level ranges.
Each range substitute different fixed number of
bits into LS part of the 8-bit gray value of the
pixels.
Pixel gray value “g” that fall within the range Ai-
Bi is changed by embedding message bits of
secret information into new gray value “g’ ”.
OPAP is used that make the new gray value “g’
” fall within the range Ai-Bi.
Digital signature of the secret information with
the key2 were obtained and appended with the
information
8. Private stego-key
generation
For a gray scale image 8-bit is used
to represent intensity of pixel, so there
are only 256 different gray values any
pixel may hold.
Different pixels in image may hold
different gray values
let four ranges of gray levels are < A1-
B1, A2-B2, A3-B3, A4-B4 > each
range starting and ending value are in
8-bits.
9. Method to decide Bits
insertion in each range
Let the four gray ranges decided by
the stego-key are <A1-B1, A2-B2, A3-
B3, A4-B4>
number of pixel count from cover
image in each range are < N1, N2,
N3, N4 >.
Let the ranges be 0-64, 65-127, 128-
191, 192-255
Let ranges hold no of pixels 34,
13238, 17116, 35148.
12. Example :
consider a gray level range 0-32, 3 – bits
substitution are 111
1 : let 00100000 (32) be pixel value of g,
2 : after LSB method, g’ = 001000111 (39)
3 : g’ < 32, K+1 bit of g’ changed from 0 to
1 or via- versa (00101111) and checked
again to fall within range if not K+2 bit is
changed (00111111) and so on until gray
value fall within range 00011111(31).
13. IMPLEMENTATION OF ALOS
ENCODING ALGORITHM
Input: Cover-image, secret message, keys K1, K2.
Output: Stego-image.
Step1: Read key K1 based on gray-Level ranges.
Step2: Read cover image
Step3: Decide No. of bits insertion into each range described in
section 2.3
Step4: Read the secret message and Convert it into bit stream form.
Step5: Read the key K2.
Step6: Find the signature using K2 and append with the message
bits.
Step7: For each Pixel
7.1: Find gray value g.
7.2: Decide the K-bits insertion based on gray ranges.
7.3: Find K-message bits and insert using method given in section
2.4
7.4: Decide and adjust new gray Value g’ using method described in
Optimum pixel adjustment process.
14.
15. EXTRACTION ALORITHM
Input: Stego-image, keys K1, K2;
Output: Secret information;
Step1: Read key K1 based on gray-level ranges.
Step2: Read the stego image.
Step3: Decide No. of bits extraction into each range
described in section 2.3.
Step4: For each pixel, extract the K-bits and save into
file.
Step5: Read the key K2 and find the signature of bit
stream
Step6: Match the signature.
Step7: End
16.
17. Advantages :
High hiding capacity compared to LSB Substitution
technique.
Robust in nature, i.e., highly secure algorithm since two
keys (key-1 and key-2) are used.
We get good quality of the stego image.
High water marking level.
Provides maximum possible payload.
Embedded data is imperceptible to the observer.
Limitations :
High computational complexity.
Requires a lot of overhead to hide a relatively bits of
information.
This can be overcome by using HIGH SPEED
18. Applications
In secret communication system.
Military applications.
Hiding and protecting of secret data in
industry.
Navy and Air force.
Business Deals
19. Results and Discussions
Lena and baboon 256 × 256 × 3 colour digital images
Range Cover image
Max bits that can
be embedded
(payload)
No of bits
embedded
Capacity
(bits/pixel
)
PSNR
Range1
Lena 653149
51360 3.8016 44.9412
4768 3.141 55.5705
115360 3.2268 40.8688
Baboon 609524
51360 3.2927 44.7668
4768 3.2 54.8287
115360 3.0352 41.7503
Range2
Lena 693700
51360 3.624 43.494
4768 3.7338 53.4812
115360 3.6078 40.3313
Baboon 700087
51360 3.6488 43.2479
4768 3.7192 53.6941
115360 3.5019 40.2084
20. The input cover image
0
200
400
600
The histogram of input cover image
0 100 200
The output stego image
0
200
400
600
800
The histogram of stego image
0 100 200
Experimental result using Range1 for Baboon cover image
21. The input cover image
0
200
400
600
800
The histogram of input cover image
0 100 200
The output stego image
0
500
1000
The histogram of stego image
0 100 200
Experimental result using Range2 for Lena cover image
22. Conclusion
Enhanced embedding capacity of image .
An adaptive number of least significant bits substitution
method with private stego-key.
whether the attacker has tried to modify the secret
hidden information in the stego-image.
embeded the hidden information in the cover image and
use digital signature using a key to verify the integrity
from the stego-image.
Experimental results verify that the proposed model is
effective and efficient.
23. References
S. Dumitrescu, W. X. Wu and N. Memon, “On steganalysis of
random LSB embedding in continuous-tone images”,
Proceeding of International conference on image Processing,
Rochester, NY, pp. 641-644, 2002.
A. Cheddad, J. Condell, K. Curran and P. McKevitt,
“Enhancing Steganography in digital images”, IEEE - 2008
Canadian conference on computer and Robot vision, pp. 326-
332,2008.
Ko-Chin Chang, Chien-Ping Chang, Ping S. Huang, and Te-
ming Tu, “A novel image steganographic method using Tri-
way pixel value Differencing”, Journal of multimedia, vol. 3,
issue 2, June 2008.
K. S. Babu, K. B. Raja, K. Kiran Kumar, T. H. Manjula Devi, K.
R. Venugopal, L. M.Pataki, “Authentication of secret
information in image steganography”, IEEE Region 10
Conference, TENCON-2008, pp. 1-6, Nov. 2008.
S. K. Moon and R.S. Kawitkar, “Data Security using Data
Hiding”, IEEE International conference on computational
intelligence and multimedia applications, vol. 4, pp. 247251,
Dec 2007