See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/329756200

A Fast Image Encryption Technique Using Henon Chaotic Map: Proceedings of

ICACIE 2017, Volume 1

Chapter in Advances in Intelligent Systems and Computing · January 2019

DOI: 10.1007/978-981-13-1708-8_30

CITATIONS READS

21 2,823

2 authors:

Kapil Mishra Ravi Saharan

Indian Institute of Information Technology Allahabad Central University of Rajasthan
5 PUBLICATIONS 36 CITATIONS 28 PUBLICATIONS 123 CITATIONS

SEE PROFILE SEE PROFILE

All content following this page was uploaded by Ravi Saharan on 24 January 2020.

The user has requested enhancement of the downloaded file.

A Fast Image Encryption Technique
Using Henon Chaotic Map

Kapil Mishra and Ravi Saharan

Abstract Recent advancements in the networking technologies led to increase in

network bandwidth, hence allowing transfer of files of large size. A major portion
of files being transferred through various networks all over the world consists of
multimedia data, particularly digital images. In this project, we propose a new image
encryption scheme that may be used for securing the digital images. The scheme
uses Henon chaotic map and 128-bit secret key in order to generate the cipher image.
Henon chaotic map is a two-dimensional iterated discrete dynamic system that shows
chaotic character on specific values of the constants used. Chaotic maps are very sen-
sitive to the initial parameters, i.e., a slight change in the initial conditions drastically
changes the overall output generated by the chaotic system. In our scheme, we use
Henon chaotic map along with an externally supplied 128-bit secret key is used to
encrypt the original image. After encrypting the image, pixel shuffling is performed
using a permutation matrix generated using the chaotic map. The algorithm is tested
on a standard set of images against various performance metrics like peak signal-
to-noise ratio (psnr), entropy, histogram etc. The algorithm was found to be robust
against plain-text, statistical attacks, chosen plain-text etc.

Keywords Information security · Image encryption · Image security

Henon chaotic map

1 Introduction

Vast improvement had been seen in the field of networking and technologies in the
recent years. Such improvements had allowed us to share and transfer large-size
media files, high-resolution digital images through various networks all over the

K. Mishra · R. Saharan (B)

Computer Science and Engineering Department, Central University of Rajasthan, NH-8,
Bandarsindri, Kishangarh, Ajmer 305817, Rajasthan, India
e-mail: [email protected]
K. Mishra
e-mail: [email protected]

© Springer Nature Singapore Pte Ltd. 2019 329

B. Pati et al. (eds.), Progress in Advanced Computing and Intelligent
Engineering, Advances in Intelligent Systems and Computing 713,
https://doi.org/10.1007/978-981-13-1708-8_30
330 K. Mishra and R. Saharan

world. Digital images being transferred through such networks may be of personal
or business nature which would be sensitive enough to require security from the
attackers and illegitimate users through the network as they may try to extract the
information which can be then leaked or used for wrong purposes. An increase had
been seen for such incidents about leaking of personal photos. So digital image
security is an increasing concern for the researchers all over the world and hence in
recent years, a lot of interest among researchers had been seen in the field of digital
image encryption [1–4].
We already have techniques for encrypting data being transferred through the
network. Various techniques including DES, IDEA, AES etc. are used for data
encryption through a secure network. A general information security model include
a sender side encryption module, network for data transfer and a decryption module
at receiver’s end. The information security is ensured by encrypting and decrypting
the data being transferred though the network. Conventional encryption techniques
discussed above provide proven security to the data. But due to the distinct char-
acteristics of digital image data as compared to conventional text data, we need to
design specialized algorithms in order to encrypt digital images [5, 6]. The distinct
characteristics of image data can be named as shown below:
• High Redundancy
• High Correlation
• High error tolerance or less sensitive to error
High redundancy and error tolerance exist in digital images due to the large number
of pixels presented in the image and inability of human eye to detect small change
in pixel values of the image [7, 8]. High correlation in digital images exist due to 8
immediate neighbors of each pixel while in conventional text data, there exist only
two immediate neighbors.
To adapt according to the different characteristics of digital image, we need to
develop dedicated encryption algorithm for images. Any such technique should be
efficient to deal with peculiar characteristics of images and simple enough to make
it easy to implement [9, 10]. Here, it is worth mention that information security
is considered to be formed by three pillars that are confidentiality, integrity and
availability. Image encryption techniques deal with ensuring the confidentiality of the
digital image data by utilizing cryptographic principles and techniques. Cryptography
is the branch of science that deals with the methods for secret communication in the
presence of third party (which may include other users, attackers etc. in our context).
Cryptographic methods are used to design encryption or decryption algorithms.
The main goal of all such methods is to ensure security of data in the presence of
untrusted users. Cryptographic methods for encryption are broadly classified on the
basis of key distribution policies as shown below:
1. Private key or symmetric key cryptography, and
2. Public key, also known as asymmetric cryptography.
In private key or symmetric key cryptography, same key is used to encrypt and
decrypt the data at senders and receivers side respectively (Fig. 3). In such methods,
A Fast Image Encryption Technique Using Henon Chaotic Map 331

secure exchange of keys is essential. These kinds of methods are computationally

low cost and require less resources. While in public key or asymmetric cryptography,
different keys, one for encryption at sender side while other for decryption at receiver
side are used. These methods are of computationally high cost and require greater
resources. Apart from above-mentioned division, image encryption algorithms are
classified based upon nature of techniques used as chaotic techniques that use chaotic
maps and non-chaotic techniques that do not use chaos at all.

2 Literature Survey

Prior to our work, we performed an exhaustive literature survey; some of the previous
works are described briefly in this section. In [1], the proposed method is based upon
SCAN language, which generates a large set of unique patterns based upon a small
set of predefined patterns. The method shuffles the pixels of the original image based
upon the encryption keys generated using the SCAN language. Encryption keys are
nothing but patterns generated by a SCAN word. Encryption keys are used in such a
way that no pixel is accessed more than once. The SCAN word is a combination of
two patterns:
• SCAN pattern
• Partition pattern
A SCAN pattern is further dependent upon four general patterns, each one of
which contains eight transformations numbered from 0 to 7. These are: Continuous
Raster C, Continuous Diagonal D, Continuous Orthogonal O, and Continuous Spiral
S. Each of the partition patterns is dependent on 3 general patterns B, Z and X.
Each one of these also depends upon eight transformation patterns. The input image
is divided into four subregions. The partition pattern decide which sub region is
traversed first by scanning path. Scanning is done for each subregion in a separate
manner. Huge no. of possible patterns makes the method resistible against the brute-
force attack.
In [2], various chaos-based methods are described. Chaotic maps are very sensitive
to the initial conditions and hence are extensively used in cryptographic methods.
Various image encryption techniques use chaotic maps. A general approach of a
chaotic technique consists of the following phases:
• Pixel Shuffling (Confusion) phase which involves changing location of pixels
and hence decorrelation of pixels. After this phase, statistical information like
histogram of the image does not change.
• Pixel modification (Diffusion) phase involves modification of pixel values.
Based upon above techniques, chaotic methods are classified into three categories:
pixel permutation or transposition techniques, pixel modification only techniques,
i.e., only changing values of the pixels, and visual transformation technique that
involves both transposition as well as pixel modification operations.
332 K. Mishra and R. Saharan

In [3], a survey paper was presented that reviews major chaotic encryption tech-
niques proposed in recent years. Different techniques are discussed along with their
respective problems and strength factors with probable application areas of the tech-
nique. A hybrid image encryption and authentication technique using hashing and
digital signature technique is discussed. Another technique using error correcting
codes was also presented. Different algorithms were presented and concluded that
each one of them is suitable for different applications. It is also concluded that if
the algorithm was not designed properly then the image may be insecure and can be
forged.
In [4], authors described an image encryption method based upon Henon chaotic
map and w7 cipher using 128-bit external secret key. The method consists of follow-
ing phases:
• At the first stage, the original image is shuffled using a permutation map generated
by the Henon chaotic map.
• At second stage, XOR operation is applied between the earlier generated shuffled
image and the cipher image generated using w7 cipher.
Permutation matrix is a matrix that consists of single one in each row and column.
It is used for shuffling because it avoids computation cost at the decryption side
as its inverse is nothing but its transpose, so it prevents heavy computation which
otherwise would be needed at the decryption side. As the algorithm uses both phases
viz. pixel shuffling and modification in order to generate the final encrypted image,
the algorithm is secure enough to thwart various types of attacks. But since it uses
w7 cipher for encryption which is a stream cipher, it needs to generate a huge cipher
stream of size m * n * 8 (for an image of size m * n) and then reshape operation needs to
be performed, the execution time of the algorithm is increased. The schematic model
is shown in Fig. 1.

Fig. 1 Image encryption by w7 cipher and Henon chaotic map

A Fast Image Encryption Technique Using Henon Chaotic Map 333

3 Proposed Algorithm

After performing an exhaustive literature survey of various image encryption tech-

nique, we realize a need of an efficient algorithm which is secure as well as is less
complex and fast. It is concluded that the algorithm should perform well on various
security parameters so that it may sustain various kinds of attacks. In our work, we
designed an approach based upon the Henon chaotic map and externally supplied
128-bit secret key.

3.1 Henon Map

Henon map may be stated as a two-dimensional iterated discrete-time dynamical

system with a chaotic attractor as proposed by Henon in 1976 [11]. It can be stated
by following pair of equations:

X n+1  1 + yn − αxn2 . (1)

Yn+1  βxn . (2)

With x0, y0 as initial point, (x, y) denote the present state of the system. Henon
showed that if S is the area bounded by four points (−1.33, 0.42), (1.32, 0.133),
(1.245, −0.14) and (−1.06, −0.5), and if the initial point lies in the area S, then the
subsequent points—(xi, yi) for i ≥ 1, also lie in S [12].
The proposed work generates permutation matrix for shuffling of pixels of the
image (confusion phase) and cipher image for encryption of the shuffled image
(diffusion phase) using Henon chaotic map and the 128 bit externally supplied secret
key. As it is a private key algorithm, we assume the same key to be available at both
sender and receiver ends. The architecture of the proposed scheme is shown with the
help of a schematic diagram in Fig. 2. The algorithm is described below:

3.2 Encryption Algorithm

To encrypt a given image, following is the algorithm:

1. Take source image of size m*n as input.
2. Generate random variables from Henon chaotic map by following steps:
(a) X, Y  Henon(m*n) in order to generate random variables using Henon
chaotic map of size equal to no. of pixels.
(b) X  abs(floor(X(1:m*n)*1000000));
(c) Y  abs(floor(Y(1:m*n)*1000000));
(d) X  reshape(X,m,n); Y  reshape(Y,m,n);
334 K. Mishra and R. Saharan

Fig. 2 Architecture of proposed scheme

(e) D  X*Y;
(f) D  D*sum(key) sum of the digits of 128 bit key.
3. Generate permutation matrix P(m*n) by calculating position for each row i as
(a) pos  mod(D, n) + 1
(b) P(i, pos)  1
(c) Other entries being zero for the row i
4. Perform combined shuffling operation by first performing vertical shuffling and
then horizontal shuffling as shown below: For each i, j from 1 to n
(a) vI(1:n, j)  P*I(1:n, j)
(b) cI(j, 1:n)  vI(j, 1:n)*P
5. For i  1:m
(a) j  mod(i,16) + 1
(b) o D(i,1:m)  bitxor(D(i,1:m),okey(1,j))
6. D  mod(D,255) + 1. cipher image
7. Generate final encrypted image by applying XOR operation between the shuffled
image and the cipher image.
For decrypting, we follow the same algorithm in reverse order as of encryption
process just replacing the permutation matrix by its inverse which is nothing but its
transpose. All other steps remaining the same make the algorithm very simple to
implement it on encryption side as well as decryption side.
A Fast Image Encryption Technique Using Henon Chaotic Map 335

4 Results and Analysis

4.1 Experimental Setup

Proposed algorithm is tested and implemented on machine with following configu-

ration:

Operating system Windows 7 ultimate 64 bit

Processor Intel core i5
Memory 2 GB
Software used Matlab 2015a
Input image Standard gray scale images of size 512 * 512
are used.

4.2 Key Space Analysis

Large key-space is required for an efficient digital image encryption algorithm in

order to resist brute-force attack. In our proposed algorithm, we use 128-bit external
secret key making the key space 2128 and furthermore if we include two seed points
of the Henon map as part of secret key, then the key space becomes even larger. If
the floating point precision of the machine is 10–14, it makes the key space of the
algorithm as large as 2128 × 1014 * 2 which is enough to resist brute-force attacks.

4.3 Key Sensitivity Analysis

An efficient digital image encryption algorithm needs to be highly key sensitive. The
algorithm must give a totally different output even after a slight change of one bit in
the security key. In the proposed algorithm, Henon chaotic map is used which due
to its chaotic character, is highly sensitive to initial conditions. Also, we are using
128-bit external key for image encryption, which is highly sensitive as well.

4.4 Histogram Analysis

Histogram of an image provides information about the frequency distribution of

its pixels and regarding density estimation. A cipher image should have a uniform
336 K. Mishra and R. Saharan

Fig. 3 Histogram analysis of plain baboon image

Fig. 4 Histogram analysis of encrypted baboon image

histogram to be secure from known plain-text attack. Figure 3 depicts the histogram
of the original baboon image and Fig. 4 shows the histogram of the encrypted image.
Since the histogram of the encrypted image is nearly uniform, the proposed algorithm
is expected to prevent known plain-text attack. As the histogram of the encrypted
image is uniform, it can be concluded that the proposed algorithm is highly resistant
to statistical attacks.
A Fast Image Encryption Technique Using Henon Chaotic Map 337

4.5 Information Entropy

Information entropy is a measurement of uncertainty or randomness in a signal or

image. A good encryption technique must incorporate randomness property and
follow uniform distribution [5]. It is calculated by the following formula:

2
N
−1
H (m)  − P(m i )log2 [P(m i )] . (3)
i0

where,
P (mi) = Probability of a pixel, and
N= Bit-depth of each pixel

4.6 Correlation Analysis

High correlation is one of the most important characteristics of data belonging to the
class of digital images. Each pixel is strongly correlated with its neighboring pixels
which may be horizontal, vertical or diagonal in position. Scatter plots are shown in
Figs. 5, 6, and 7 for depicting the correlation between randomly selected 5000 pixel
pairs of each of the horizontal, vertical and diagonal locations for both the original
plain image and the generated cipher image. The standard 512 * 512 sized gray scale
cameraman image was used for the correlation testing. Correlation coefficients are
calculated using the Eq. 1 shown. The correlation coefficients of the various plain
images and that of respective encrypted images are shown in Fig. 8. Generally, for
a normal image, pixels are highly correlated and the coefficients are very close to 1,
while for the encrypted image, the coefficients are close to 0.

Fig. 5 Correlation analysis of horizontal pixel pairs

338 K. Mishra and R. Saharan

Fig. 6 Correlation analysis of vertical pixel pairs

Fig. 7 Correlation analysis of diagonal pixel pairs

5 Conclusion

In this paper, a new method for image encryption is proposed. The technique is based
upon using chaotic properties of Henon map as pseudo-random number generator
along with 128 bit secret key to obtain permutation matrix for shuffling of the original
image and a cipher image that is used to finally encrypt the shuffled image. The
method is vigorously tested on standard test images based upon various security
parameters of digital image encryption. The focus is kept on keeping the mechanism
simple enough, making it easy to implement in practical applications.
The future scope of the work may constitute the optimization of the algorithms for
applications in sensor nodes and military applications where the processing ability of
the nodes is extremely low. As, any algorithm that is costly in terms of computational
cost, that can not be implemented in the discussed scenario.
 A Fast Image Encryption Technique Using Henon Chaotic Map 339

Fig. 8 Security analysis of the proposed algorithm

References

1. Chen, C.-S., Chen, R.-J.: Image encryption and decryption using SCAN methodology. In: 7th
International Conference on Parallel and Distributed Computing, Applications and Technolo-
gies, 2006. PDCAT06, pp. 61–66. IEEE (2006)
2. Sankpal, P.R., Vijaya, P.A.: Image encryption using chaotic maps: a survey. In: 2014 Fifth
International Conference on Signal and Image Processing (ICSIP), pp. 102–107. IEEE (2014)
3. Rajput, A.S., Mishra, N., Sharma, S.: Towards the growth of image encryption and authentica-
tion schemes. In: 2013 International Conference on Advances in Computing, Communications
and Informatics (ICACCI), pp. 454–459. IEEE (2013)
4. Jolfaei, A., Mirghadri, A.: An image encryption approach using chaos and stream cipher. J.
Theor. Appl. Inf. Technol. 19(2), 117–125 (2010)
5. Kumar, M., Aggarwal, A., Garg, A.: A review on various digital image encryption techniques
and security criteria. Int. J. Comput. Appl. 96(13) (2014)
6. Wei-bin, C., Xin, Z.: Image encryption algorithm based on Henon chaotic system. In: 2009
International Conference on Image Analysis and Signal Processing. IEEE (2009)
7. Ping, P., Mao, Y., Lv, X., Xu, F., Xu, G.: An image scrambling algorithm using discrete Henon
map. In: 2015 IEEE International Conference on Information and Automation, pp. 429–432.
IEEE (2015)
8. Nithin, N., Bongale, A.M., Hegde, G.P.: Image encryption based on FEAL algorithm. Int. J.
Adv. Comput. Sci. Technol. (2013)
9. Hamad, S., Khalifa, A., Elhadad, A., Rida, S.Z.: A modified playfair cipher for encrypting
digital images. Mod. Sci. (2013)
10. Soleymani, A., Nordin, M.J., Sundararajan, E.: A chaotic cryptosystem for images based on
Henon and Arnold cat map. Sci. World J. (2014)
11. Forre, R.: The Hnon attractor as a keystream generator. In: Advances in Cryptology-EuroCrypt,
vol. 91, pp. 76–81 (1991)
12. Hnon, M.: A two-dimensional mapping with a strange attractor. Commun. Math. Phys. 50(1),
69–77 (1976)

View publication stats