Reversible Data Hiding In Encrypted Images And Its Application To Secure Missile Launching

Radhika R. Patil, Deepali Y. Loni
International Journal of Image Processing (IJIP), Volume (12) : Issue (3) : 2018 63
Reversible Data Hiding In Encrypted Images And Its Application
To Secure Missile Launching
Radhika R. Patil radhikap589@gmail.com
Department of Electronics
DKTE Society’s Textile & Engineering Institute, Ichalkaranji
Kolhapur-416115, Maharashtra, India
Deepali Y. Loni deepaliloni@rediffmail.com
Department of Electronics
DKTE Society’s Textile & Engineering Institute, Ichalkaranji
Kolhapur-416115, Maharashtra, India
Abstract
This paper proposes reversible data hiding in encrypted images for secure missile launching. The
work is presented in two stages: one involves encryption of cover image by block cipher algorithm
and other is embedding secure data related to missile launching. For embedding data, vacant
pixels are identified by Slepian-Wolf encoding method along with embedding key to hide the data.
At the other end by using decryption algorithm the original cover image is recovered and the
secret data is extracted. The performance analysis is presented by calculating parameters MSE,
PSNR and SSIM.
Keywords: Image Encryption, Data Embedding, Reversible Data Hiding.
1. INTRODUCTION
In many applications, like law forensics, military imagery and medical imagery the information
vendor requires to transmit data to a distant server for future processing. Now a day, internet is
the prime medium to transfer information from one end to another across the world. The
additional secret information can be hacked in a lot of different ways. This is the major problem
with sending information over the internet. Therefore it becomes very important to take data
security into consideration, during the procedure of data transferring. The intruder may also
capture image, and view the significant contents and then alter the image before transferring it to
receiver [1]. This is the way by which original image contents will be modified and receiver cannot
have an idea about it. In general, a bit of content distortion is typically imperceptible to human
imaginative. However, such distortion is not favored in some applications, like legitimate
documentation, medical imaging, military observations, high-accuracy scientific investigation,
since it might prompt risk of wrong decision making. Data security basically means given that
safety to information from unauthorized users or hackers and imparting excessive level of
protection to prevent information from modification. Data hiding is one kind of approach to secure
data in cover media but there exists some distortion. In data hiding method the private and secret
information is hidden into cover (host) image.
A large volume of data sent over internet is private and secret. Encryption is technique which
transmits the secret data. The reversible data hiding is also treated as the new watermarking
method which is used to validate an image by embedding some data on it as a watermark [2].
Most multimedia system data hiding method insert the extra information and modify original
content [3], and thus distortion in cover image occurs. Data hiding activity insert information bits
by changing the cover image, but enable the precise re-establishment of the original cover image
after getting the embedded secret data. Within majority of applications, the little distortion

because of the data embedding is sometimes acceptable. For plaintext images many RDH
methods have been proposed [4]-[7], these are not applicable to encrypted images since the
redundancy in the original image cannot be used directly after image encryption.
Over the past few years, an excellent amount of schemes regarding to data hiding in encrypted
images has been developed. Even though, inside these schemes, the cover image is lastingly
distorted because of data embedding. In general, the cloud service supplier has no authority to
add everlasting distortion. This means that, the original plaintext content i.e. cover image should
be re-established without any mistake in image and data recovery for licensed receiver. To
overcome this problem in encrypted images, the solution is the use of reversible data hiding
(RDH).
The service supplier adds additional secret messages, e.g., notations, labels, verification
information, or image data in encrypted images even not accessing the original cover image
contents [8]. This is possible because of reversible data hiding technique in encrypted images.
The original cover image is compulsorily recovered totally and also the hidden secret message is
completely recovered at the receiving end. RDH in encrypted images is attractive. For example,
in medical application, a patient will not give permission to expose his/her medical images to any
outsider, whereas database manager may need to implant the medical records or patient’s
information in the encrypted image [9]. On the other end, the original cover image for diagnosis
should re-established without any error after decryption and revival of the hidden secret
information.
Strategies proposed in [10] - [12] makes use of the reversible data hiding is accomplished by
using LSB modification. First the original cover image is encrypted using special encryption
algorithm and then some of them embeds one bit of data into each block by a way of just flipping
the last three LSBs. The spatial relationship exists in natural images and the interfered block,
interfered block must be less smooth than the original block. Thus, original cover is recovered
along with secret information. If selection of block falls in inappropriate block size, during data
extraction and image recovery errors may occur. Thus block size is a factor which decides
embedding rate of this method. Some RDH methods use histogram modification [9]. A histogram
modification and n-nary data hiding scheme used to embed secret information into encrypted
image. At the receiver end, original cover image can be totally recovered and the additional
information can be extracted with the aid of the embedding key and the encryption key.
Another approach proposed in [8] uses the Slepian-Wolf encoding for data hiding. This idea is
inspired by distributed source coding (DSC) [13]-[14]. In this first the image encryption is done by
stream cipher algorithm then by using low density parity check codes the spare room is
generated to add secret data in that vacated room. The information extraction and image
recovery is with the aid of using distributed source coding technique. Along with RDH in
encrypted images algorithm the reserving room before encryption technique [2] is used. And
consequently it is straightforward for the data hider to reversibly hide data within the encrypted
image. This approach has been given an amazing amount of reversibility, that is, data extraction
and image recovery are not containing any error. The difference expansion method [4] calculates
the neighboring pixel values differences, and for the difference expansion (DE) selects some
difference values. This is applicable for audio and video as well. The information about original
content restoration, additional data, and message authentication code will be implanted into the
difference values.
This work proposes reversible data hiding in encrypted image for secure missile launching. The
original cover is entirely encrypted by using block cipher, and the secret data is embedded by
modifying a part of encrypted image. At receiver side, with the help of embedding key and
encryption algorithm, the embedded data are successfully extracted while the original image is
perfectly recovered.

2. SYSTEM DESCRIPTION
The general structure of the proposed method is presented in Figure 1, which consists of 4
phases, namely image encryption, data hiding, image recovery, data extraction, and the last unit
of missile launching.
The original cover image is first transformed into encrypted image using encryption algorithm.
The data hider embeds the additional secret information into encrypted image using an
embedding key to generate embedded image.
At the receiving end, receiver extracts the inserted secret information independently only if it has
an embedding key. If receiver has knowledge about encryption algorithm and embedding key
both, the inserted secret bits can be extracted and original cover image can be recovered.
Cover
Image
FIGURE 1: The Proposed System.
2.1 Image Encryption
Original image must be grayscale; if input image is color then we first convert it into grayscale (0
to 255). The image is preprocessed such as image resizing and converted into particular intensity
range. The mathematical operations on the image may results into negative value or may exceed
the upper boundary. At recovery stage this may result in receiving the random symbols like $, #,
etc. To avoid such circumstance at receiving end we set intensity range of image to 15 and 240.
Encryption is not directly applied on whole image; we select non-overlapping blocks from cover
image. We divided selected block into two sub regions and then calculate pixel difference value
and integrative component. Again we further divided the result obtained into two sub regions
resulting into total four sub regions. For these four sub regions we calculated the difference and
integrative components labeled as c1, c2, c3, and c4. We shuffled all the four components before
combining them. We have not used any shuffling key, shuffling is done by simply rearranging the
four components in different order (like c2, c1, c4, c3). The result of preprocessing and encryption
are shown in Figure 2 and 3 respectively.
Original cover image Preprocessed cover image
FIGURE 2: Preprocessing of Cover Image.
Block Cipher Slepian-Wolf
Encoding
Transmission
Reception
Decoding (Image
Recovery and Data
Extraction)
Missile
Launching
Unit

Preprocessed cover image Encrypted image
FIGURE 3: Encryption of Preprocessed Image.
This decomposition and combing process is applied for each and every non overlapping block of
cover image. Combined results (c2, c1, c4, c3) of each block, result into total encrypted image.
2.2 Data Hiding
The Slepian-Wolf encoding method is used for data hiding. To compress the selected bits from
encrypted image the Slepian-Wolf codes are useful. Defined low density parity check code
(LDPC) matrix H, it can be constructed in various different forms. The data-hider arbitrarily
chooses a parity-check matrix H corresponding to a regular or irregular LDPC code by setting the
numbers of variable nodes and the check nodes. The different algorithms have been proposed
for the matrix construction, for example, matrices used in Gallager codes, MacKey codes, and
finite geometry codes. For example by using MacKey method the matrix is constructed by
following steps
1. The H matrix is created by initially creating all zero matrix and then randomly flipping bits in
the matrix H. The flipped bits must not be necessarily distinct.
2. The matrix H is generated by randomly creating weight j columns.
3. The matrix H is generated with weight j per columns and uniform weigh per row and no two
columns are connected to the same row more than ones (avoiding four cycles)
4. Matrix H is generated as in step 3 with the girth condition further constrained so that the girth
is larger than six.
The MacKey’s algorithms were used to find good performing codes with the variety of length and
rates. The more details of LDPC matrix H can be found in [14].
We have selected a non-overlapping blocks of encrypted image for data embedding. We have
selected an embedding key and checked for embedding key bit equals to one. And where the
embedding key bit is one the value at that position in the selected block of encrypted image is
considered as coefficient. We performed matrix multiplication between selected coefficient and H
matrix, and the spare room is generated for data embedding. In the vacated room the data is
embedded and checked for if we have completed with all secret data to be embedded. Otherwise
data embedding process is continued. Along with this we have checked for one more condition,
whether we completed with selected block of encrypted image, if done then next block is
selected. Finally all non overlapping blocks are combined to form an embedded image which is
encrypted image containing secret data. Generated embedded image is transmitted and is shown
in Figure 4. The whole embedding process is shown in Figure 5.

Encrypted Image Embedded Image
FIGURE 4: Embedded Image.
2.3 Image Decryption
At the receiver side, using the received embedded image, the original cover image can be
recovered by decryption algorithm. First we divided the received embedded image into blocks
and those blocks are decomposed into four sub regions and applied decryption algorithm on it.
We performed exact reverse of the process applied at the transmission side. We have calculated
the pixel differences and integrative components for all four sub regions. We combined the
obtained results into two sub regions. Again by using these two sub regions we have calculated
pixel difference and integrative component and combined obtained results into single block. The
process was repeated for each block and all resulted blocks were combined to form
decrypted/recovered image. The recovered cover image is shown in Figure 6.
Embedded Image Recovered Cover Image
FIGURE 5: Image Extraction.
2.4 Data Extraction
In data extraction, we select a non overlapping block of embedded image. We used an
embedding key to extracts the embedded secret data. We checked for embedding key bit, if it is
one it means data is embedded at that position in selected block of embedded image. The value
at that position is considered as coefficient and LDPC matrix is applied on it. From this we get to
know the position where data is embedded and those bits are extracted. The extracted data is
shown in Figure 7.
2.5 Missile Launching Unit
The decrypted secret data are the coordinates required for missile launching. The launching of
missile involves coordinates corresponding to azimuth and elevation angle made by missile.

Accordingly during embedding process we embedded these coordinates. For example xy120150;
where xy is used as an identifier which shows the start of valid bit frame, the next three digits
corresponds to azimuth angle and last three digits corresponds to elevation angle.
No
Yes
No
Yes
FIGURE 6: Data Embedding Process.
By using the serial communication the extracted data is send to missile launching hardware unit.
The hardware unit consists of two DC motors, motor driver circuitry and controller. If identifier is
received properly then the received data is treated as the valid frame and the DC motor rotation is
made accordingly. The missile is then positioned at the target location. If received identifier does
not match then it indicated invalid data received.
Start
Load Encrypted image and Embedding key
Select 8*8 Block from encrypted image
Read embedding key.
If
Embedding key bit==1
Generate vacant space by Slepian wolf encoding and
perform data embedding
If completed with
all secret data
Combine result of each 8*8 block
End

` Embedded Image Extracted Secret Data
FIGURE 7: Data Recovery.
3. PERFORMANCE PARAMETER
To evaluate the system performance we used mean square error (MSE), structural similarity
index matrices (SSIM), and peak signal to noise ratio (PSNR). These parameters give an idea
about how efficient the system is and from SSIM parameter we come to know how closer the
original cover image to the recovered image. The original cover image and recovered cover
image are used to calculate these performance parameters.
a. Mean Square Error
Let us consider two digital images X and Y. The MSE between these two images is expressed
mathematically as:
MSE(X, Y) = (1)
Where N is the number of pixels in digital image
The value of MSE ranges in between 0 and 1. The lower value of MSE indicates less error, as
MSE goes on increasing error also increases.
b. Peak Signal To Noise Ratio
The PSNR between two images is expressed mathematically as:
PSNR ( , ) =20 ( ) (2)
Where, MPP is Maximum Possible Pixel in an image, i.e. if the image of 8 bit then the MPP= –1
= 255 pixels.
As the lower PSNR the lower relative image quality. Higher PSNR indicates a good quality image.
c. Structural Similarity Index Metric
Structural Similarity Index quality is based on the computation of three terms namely luminance,
contrast, and structural term. The overall index is multiplicative combination of these three terms.
SSIM(x, y) = (3)
Where,
= , = , And =

Where, s
, and are local mean, standard deviation, and cross covariance for images x, y.
If = then
SSIM(x, y) = (4)
The range of SSIM is 0 to 1. The value 1 indicates two images are exactly same. As SSIM value
goes on decreasing, similarity between original and recovered image is less.
The obtained parameter MSE, PSNR, and SSIM for different images are listed in Table 1.
Image MSE PSNR SSIM
barb Color 0.140772 56.645653 0.999983
Cameraman 0.006736 69.846711 0.999979
Lena 0.003264 72.993450 0.999991
Peppers 0.004462 71.635697 0.999990
office_6 0.187708 55.395968 0.999966
TABLE 1: Performance Parameter Calculated for Different Images.
We have got MSE value approximately 0.1 which indicates less error and the PSNR value nearly
about 60 to 70 which indicates good quality of the recovered image. The last parameter we
calculated SSIM equals to 0.99 which indicates that original cover image and recovered cover
image are exactly same. From the Table 1 the MSE and PSNR value we achieved are good for
gray images than color. Also SSIM we attained is approximately same for color as well as gray
images.
The comparison the proposed work with other research approaches for the PSNR performance
parameter for Lena image is shown in Table 2.
Method used
Proposed
method
Reserving
room before
encryption
[2]
Distributed
source coding
[8]
LSB
modification
[10]
PSNR 72.993450 67.16 37.9 37.9
TABLE 2: Comparison of Performance Parameter for Lena Image.
From the Table 2 it is observed that the PSNR we achieved (72.99) is greater than those
obtained in [2], [8] and [10]. The higher value of PSNR indicates better system performance.
4. CONCLUSION
Here we propose a technique of reversible data hiding in encrypted images and its application to
secure missile launching. Initially encryption is applied on original image. Depending upon
embedding key, bits of encrypted image are selected and Slepian-Wolf encoding is applied to

make spare room for the secret data. At the receiver end, all hidden secret data is extracted using
embedding key, also original image is approximately recovered with good quality with the help of
decryption algorithm. To generate corresponding syndromes LDPC parity-check matrix is used.
On encrypted image we performed data embedding operation, so the data-hider cannot access
the contents of the original image. That ensures security of the contents in data hiding. As the
embedding and recovery is protected by the encryption and embedding keys, an adversary is
unable to break into the system without these keys. The future direction is to improve system
performance parameters by considering noisy channel.
5. REFERENCES
[1] Shilpy Mukherjee, A. Mahajan, “Review on Algorithms and Techniques of Reversible Data
Hiding” International Journal of Research in Computer and Communication Technology, Vol
3, Issue 3, March- 2014
[2] K. Ma, W. Zhang, X. Zhao, N. Yu, and F. Li, “Reversible data hiding in encrypted images by
reserving room before encryption,” IEEE Trans. Inf. Forensics Security, vol. 8, no. 3, pp. 553–
562, Mar. 2013.
[3] Mehmet U. Celik, Gaurav Sharma, A. Murat Tekalp and Eli Saber, “Reversible DATA Hiding”
IEEE ICIP pp. 157-160. 2002.
[4] J. Tian, “Reversible data embedding using a difference expansion,” IEEE Trans. Circuits
Syst. Video Technol., vol. 13, no. 8, pp. 890–896, Aug. 2003.
[5] Z. Ni, Y.-Q. Shi, N. Ansari, and W. Su, “Reversible data hiding,” IEEE Trans. Circuits Syst.
Video Technol., vol. 16, no. 3, pp. 354–362, Mar. 2006.
[6] D. M. Thodi and J. J. Rodriguez, “Expansion embedding techniques for reversible
watermarking,” IEEE Trans. Image Process., vol. 16, no. 3, pp. 721–730, Mar. 2007.
[7] L. Luo, Z. Chen, M. Chen, X. Zeng, and Z. Xiong, “Reversible image watermarking using
interpolation technique,” IEEE Trans. Inf. Forensics Security, vol. 5, no. 1, pp. 187–193, Mar.
2010.
[8] Zhenxing Qian, and Xinpeng Zhang, “Reversible data hiding in encrypted images with
distributed source encoding,” IEEE Transactions on Circuits and Systems for Video
Technology, vol. 26, no. 4,pp. 636-646 April 2016.
[9] Z. Qian, X. Han, and X. Zhang, “Separable reversible data hiding in encrypted images by n-
nary histogram modification,” in Proc. 3rdInt. Conf. Multimedia Technol. (ICMT), Guangzhou,
China, 2013, pp. 869–876.
[10]X. Zhang, “Reversible data hiding in encrypted image,” IEEE Signal Process. Lett., vol. 18,
no. 4, pp. 255–258, Apr. 2011.
[11]X. Zhang, “Separable reversible data hiding in encrypted image,” IEEE Trans. Inf. Forensics
Security, Vol. 7, No. 2, pp. 826–832, Apr. 2012.
[12]W. Hong, T.-S. Chen, and H.-Y. Wu, “An improved reversible data hiding in encrypted images
using side match,” IEEE Signal Process. Lett., vol. 19, no. 4, pp. 199–202, Apr. 2012.
[13]W. Liu, W. Zeng, L. Dong, and Q. Yao, “Efficient compression of encrypted grayscale
images,” IEEE Trans. Image Process., vol. 19, no. 4, pp. 1097–1102, Apr. 2010.
[14]S. S. Pradhan and K. Ramchandran, “Distributed source coding using syndromes (DISCUS):
Design and construction,” IEEE Trans. Inf. Theory, vol. 49, no. 3, pp. 626–643, Mar. 2003.

Reversible Data Hiding In Encrypted Images And Its Application To Secure Missile Launching

More Related Content

What's hot (20)

Similar to Reversible Data Hiding In Encrypted Images And Its Application To Secure Missile Launching (20)

Recently uploaded (20)

Reversible Data Hiding In Encrypted Images And Its Application To Secure Missile Launching