IJSTE - International Journal of Science Technology & Engineering | Volume 3 | Issue 05 | November 2016

ISSN (online): 2349-784X

A Comparative Study on Reversible Data Hiding

in Encrypted Images
Sruthy Raj
PG Scholar
Department of Electronics & Communication Engineering
MBC College of Engineering & Technology Peermade,
Kerala, India

Annmary Thomas
Assistant Professor
Department of Electronics & Communication Engineering
MBC College of Engineering & Technology Peermade,
Kerala, India

Abstract
Data hiding is a major issue of concern while exchanging secret information in an untrusted network. Reversible data hiding
plays a vital role in secure secret data transmission. For many applications such as medical image processing and military
application every bit of data is important .Even a small amount of distortion is unacceptable and any change can affect the
intelligence of the image. RDH-EI is the process of hiding secret information in a cover medium and the lossless extraction of
the same ensuring increased payload capacity, better protection and better visual quality of the recovered image. This paper
presents a survey on various data hiding techniques and related works.
Keywords: Difference expansion, Image encryption, Prediction Error Expansion, Reversible data hiding, Seperability
property, sparse coding
________________________________________________________________________________________________________
I.

INTRODUCTION

The idea of reversible data hiding in encrypted images (RDH-EI) has attracted researchers from centuries. While exchanging
heavy amount of data on internet, the security of the secret information is an issue of concern. . In application such as medical
imaging, covert military imaging and law forensic even a small amount of distortion is unacceptable and any change can affect
the intelligence of the image. RDH-EI is the process of hiding secret information in a cover medium and the lossless extraction
of the same ensuring increased payload capacity, better protection and better visual quality of the recovered image. A main
aspects of a data hiding scheme involves the following
Imperceptibility: Inability of a hacker to distinguish the original and the secret image.
Embedding Capacity: Amount of secret information that can be embedded without loss of the quality of the image.
Robustness: The degree of modification the cover image can withstand without loss of data.
The rest of the paper is organized as follows: Section II presents a detailed study on RDH. Section III presents the literature
survey on various data hiding techniques and Section IV deals with associated works in audio and video. Finally Section V
appeals the conclusion of the paper.
II. DATA HIDING IN DETAIL
Reversible data hiding
RDH-EI is the process of hiding secret information in a cover medium and the lossless extraction of the same ensuring increased
payload capacity, better protection and better visual quality of the recovered image. Data hiding is classified into 3 categories.
Pure data hiding where there is no stego key which is based on the assumption that other users are unaware of the
communication. The other category is secret key data hiding where prior to communication, the stego key is exchanged. Public
key data hiding is a type in which a public key and a private key is used for secure communication. Four major steps included in
reversible data hiding is as follows
1) Selection of the cover media in which the secret data is to be hidden.
2) The secret message which is to be transmitted with in the cover image.
3) A key that will be used to hide the data in the cover media and its inverse to retrieve the embedded secret data.
4) A key or a password to authenticate or to hide and unhide the data.

All rights reserved by www.ijste.org

90

A Comparative Study on Reversible Data Hiding in Encrypted Images

(IJSTE/ Volume 3 / Issue 05 / 017)

Fig. 1: Sketch of reversible data hiding in encrypted image.

The block diagram of RDH is shown in Fig.1 [1] which consist of image provider, data hider and a receiver. An encrypted
version of original image is given to data hider for hiding secret data by the image provider. The original image and additional
data is extracted losslessly at the receiver
III. LITERATURE SURVEY
X. Zhang [1] presented a scheme in which cover image is encrypted using a stream cipher. The encrypted image is divided into
various non-overlapping blocks and these each block again into two sets. The data hider embeds additional bits by flipping 3
LSB bits of these sets. The cover image is extracted lossless at the receiver end by using spatial correlation. These methods had
certain limitations such as each block is embedded only one bit payload. And if the block size is small, the error bits of data
extraction increase. This method did not consider the pixel correlations in the border of neighbouring blocks and did not fully
exploit the pixels in calculating the smoothness of each block.
In [2] reversibility property was improved using the technique of side match. This method uses spatial correlation between
adjacent pixel blocks. The cover image and secret data was recovered by concatenating the borders of the recovered blocks to the
unrecovered blocks. This method achieved better embedding rate by decreasing the error rate of extracted bits. In both [1] and
[2] the seperability property of RDH was not utilised.
Later in [3] presented a RDH scheme by using the histogram shifting. This simple method used the spare space for embedding
the secret information by shifting the some of the pixel values and others are expanded for lossless data embedding. Since the
PSNR is about 48.13 dB the visual quality of marked encrypted image itself is high. The major limitation of this method is the
time consumption than other conventional methods.
J.Tian [4] has proposed difference expansion method for embedding data in cover images. For each pixel pair the integer
average and difference is calculated. The difference is expanded in order to embed additional bit in the LSB of these pixel pair.
Quality degradation is comparatively low since correlation between two adjacent pixels is used. Compression and decompression
causes no loss of data in this method. The main limitation of this method is errors due to bit replacement which results in visual
quality degradation.
[5] Proposed a RDH technique called prediction-error expansion which is a modification of [4] .This method improves the
distortion performance at low embedding capacities and alleviates the problem of capacity control. This method utilises the local
correlation between in the larger neighbourhood. Prediction error is obtained using a suitable algorithm and this error is
expanded. The difference between prediction result and original image is used to embed additional data. Also [5] maximize the
embedding capacity compared to [4].
A hybrid method [6] is a modification of [5] introduced an algorithm consisting three algorithms adaptive embedding,
PredictiveError Expansion (PEE) and Pixel selection. Predictive Error expansion embed the data and provides authentication
and integrity to the user. The proposed system divides the image pixels into two parts. Later the required pixel is selected and
adaptive embedding is performed simultaneously.
The work done in [7] was a novel approach in which the cover data is first encrypted and then compressing it. The
compression is done using distributed source coding principles and is unaware of encryption key. This paper also discussed a
related problem of image hashing. Under some conditions the encrypted data can be compressed to the same rate as the original
cover image could have been compressed.
In [8] more secure RDH-EI method using the public key modulation scheme was introduced. This method used a SVM
classifier to distinguish between encrypted and non-encrypted image patches and jointly decoded the embedded message and the
original image signal which in turn increases the embedding capacity.[1]-[8] are implemented in PNG or TIF format. For JPEG
images in order to embed data strong compression is required which causes the loss of image content.[9] implemented RDH-EI

All rights reserved by www.ijste.org

91

A Comparative Study on Reversible Data Hiding in Encrypted Images

(IJSTE/ Volume 3 / Issue 05 / 017)

in JPEG images. In this first JPEG image was realised as a bit stream and encrypted this bit stream. Embedded additional data
into the encrypted images by adjusting the arrangement of this bit stream. The major drawback of all these system was data
extraction can be done only after image encryption as depicted in fig 2[10].

Fig. 2: Sketch of non-separable reversible data hiding in encrypted image.

X. Zhang [10] introduced the concept of separable reversible data hiding in encrypted images. In this method hidden data is
extracted directly from the encrypted image. The encrypted image is permuted and this is again divided into segments. For each
segment the LSB layers are compressed in order to embed the secret data. At receiver part the original LSBs are recovered by
comparing the estimated bits with the compressed one. Three cases are introduced at the receiver side as in fig 3[10]. Only with
an encryption key or an embedding key and the recovery having both of them.

Fig. 3: Three cases at receiver side of the proposed separable scheme.

In [11] lossless compression of encrypted data using LDPC codes was introduced. Earlier method were limited in 3 LSB
layers. In this a fourth LSB layer was used for additional data embedding. An estimation error based data embedding scheme was
introduced in [12]. Some pixels are estimated before encryption. In [13] an error free approach in [10] was introduced by in
conjoining the concept of multi granularity encryption and additional secret data is embedded in several smoother blocks. The
main limitation of this system is the difficulty in room vacation. In order to overcome this limitation [14] was introduced. In this
method additional room for embedding secret data is emptied out before encryption and thus improved the embedding capacity
to more than ten times as large payloads. The limitation of these method is that the spare space emptied out is limited to at most 3
LSB planes per pixel.
A hybrid approach combining merits of [10] and [14] is introduced in [15].In these method the image is analysed as a patch.
The cover image is represented as sparse coefficients by sparse encoding using a over complete dictionary. The linear sparse
representation empty out a large sparse room for data embedding and thus capacity is improved. In [16] progressive recovery
based RDH-EI was introduced which is a modification of [10]. In this method content owner encrypts the cover image with any
of the encryption algorithm most preferably AES. Later the data hider divides this encrypted image into three sets. The additional
information is embedded into the LSB planes of this three sets. In the receiver phase 3 phases as in [10] is done but consist of an
additional progressive recovery stage for lossless extraction of cover image with all of its qualities. In progressive recovery stage

All rights reserved by www.ijste.org

92

A Comparative Study on Reversible Data Hiding in Encrypted Images

(IJSTE/ Volume 3 / Issue 05 / 017)

the LSB of first set is estimated which in turn is used for the estimation of other sets. The comparison on various data hiding
techniques was done on the basis of response of various method toward the aspects of data hiding. The parameter called PSNR is
used for evaluating the quality of the recovered image.
IV. RDH IN AUDIO AND VIDEO
Data Hiding Techniques in Audio Signal
Secret data can be transmitted over the network by generating a stego-audio signal which consist of secret data .This could be
accomplished by directly replacing the digitized sample of the audio with the hidden data or by hiding the secret data in the LSB
of audio by considering the parity of the sample [17][18].For that initially the parity of the sample is checked and then the data to
be hidden is embedded into the LSB. Another method is by XORing of the LSB s. Depending on the outcome of this operation
the LSB of the sample data is changed or left unchanged. [19] Uses an algorithm of polarity of reverberations which is added to
the high frequency signals. The data is embedded by replacing high frequency signals in one middle channel. Each channel
performs the algorithm to adjust the coherence between these channels. The embedded data is retrieved using the correlation
between the difference and sum of the stereo signal. The main advantage of this method is that the original data is not required to
extract the hidden data.
Data Hiding Techniques in Video Sequences
Adaptive embedding algorithm based RDH is the most commonly used scheme in video. This method uses 4x4 DCT residual
blocks and determines predefined threshold. Until the first non-zero coefficient is met the blocks are scanned in an inverse zigzag
fashion. The predefined threshold is compared with value of this coefficient and if it is greater than the threshold then secret data
is embedded in that pixel.
V. CONCLUSION
This paper gives a detailed comparative study of various data hiding techniques for secure data transmission over internet. The
paper mainly focus on RDH schemes in images which is the most common application of data hiding. Different encryption
methods and data embedding schemes are used in various papers in order to attain various aspects of RDH.A well knowledge of
data hiding techniques will unquestionably guide us to identify new areas and to improve its applications in the already existing
application areas also.
REFERENCES
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]

X. Zhang, "Reversible data hiding in encrypted images," IEEE Signal Process. Lett., vol. 18, no. 4, pp. 255258, Apr. 2011.
W. Hong, T. Chen, and H. Wu, An improved reversible data hiding in encrypted images using side match, IEEE Signal Processing Lett vol. 19,no. 4,
pp. 199202, Apr. 2012.
Z. Ni, Y. Shi, N. Ansari, and S. Wei, Reversible data hiding, IEEE Trans.Circuits Syst. Video Technol., vol. 16, no. 3, pp. 354 362, Mar.2006.
J. Tian, Reversible data embedding using a difference expansion Trans. Circuits Syst. Video Technol., vol. 13, no. 8, pp. 890 2003.
D.M. Thodi and J. J. Rodriguez, Expansion embedding techniques for reversible watermarking, IEEE Trans.Image Process., vol. 16,no. 3, pp. 721730,
Mar. 2007.
X. L. Li, B. Yang, and T. Y. Zeng, on adaptive prediction-error expansion and pixel selection Image Process., vol. 20, no. 12, pp. 3524.
M. Johnson, P. Ishwar, V.M. Prabhakaran, D. Schonberg and K. Ramchandran, On compressing encrypted data, IEEE Trans. Signal Process., vol. 52, no.
10, pp. 2992-3006, Oct. 2004
J. Zhou, W. Sun, L. Dong, et al. Secure reversible image data hiding over encrypted domain via key modulation, IEEE Transactions on Circuits and
Systems for Video Technology, vol. 26,no. 3, pp. 441452, Mar. 2016
Z. Qian, X. Zhang, and S. Wang, Reversible data hiding in encrypted JPEG bit stream, IEEE Transactions on Multimedia vol. 16, no. 5, pp. 14861491,
Aug. 2014.
X. Zhang, Separable reversible data hiding in encrypted image, IEEE Transactions Information Forensics and Security,vol. 7, no. 2, pp. 826832, Apr. 2012
X. Zhang, Z. Qian, G. Feng, and Y. Ren, Efficient reversible data hiding in encrypted images, J. Vis. Commun. Image Represent. vol. 25, no. 2, pp. 322
328, Feb. 2014.
W. Zhang, K. Ma, and N. Yu, Reversibility improved data hiding in encrypted images, Signal Process., vol. 94, pp. 118127, Jan. 2014.
Z. Yin, B. Luo, and W. Hong, Separable and error-free reversible data hiding in encrypted image with high payload, Sci. World J., vol. 2014, Mar. 2014,
Art. ID 604876
K. Ma, W. Zhang, et al. Reversible data hiding in encrypted images by reserving room before encryption, IEEE Transactions Information Forensics and
Security, vol. 8, no. 3,pp. 553562, Mar. 2013.
X. Cao, L. Du, X. Wei, et al. High capacity reversible data hiding in encrypted images by patch-level sparse representation, IEEE Transactions on
Cybernetics, 46(5): 1132-1143, 2016
Zhenxing Qian,, Xinpeng Zhang, Guorui Feng, Reversible Data Hiding in Encrypted Images Based on Progressive Recovery, IEEE Signal Processing
Letters ,vol. 23, no. 11, November 2016.
H. B. Kekre, Archana Athawale, Archana Athawale, Uttara Athawale, Information Hiding in Audio Signals, International Journal of Computer
Applications IJCA, Vol. 7, No. 9, Foundation of Computer Science, New York, USA, pp. 14-19.
B. Santhi, G. Radhika and S. Ruthra Reka, Information Security using Audio Steganography-A Survey, Research Journal of Applied Sciences,
Engineering and Technology, Vol. 4, No. 14, pp. 2255-2258.
K. Kondo, A Data Hiding Method for Stereo Audio Signals Using the Polarity of the Inter-Channel Decorrelator, IEEE Fifth International Conference on
Intelligent Information Hiding and Multimedia Signal Processing, IIH-MSP'09. 12-14 Sept. 2009, pp.86-89

All rights reserved by www.ijste.org

93