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

net/publication/290505567

Image Encryption Based on 2-D Zaslavskii Chaotic Map and Pseudo Hadmard
Transform

Article in International Journal of Hybrid Information Technology · July 2014

DOI: 10.14257/ijhit.2014.7.4.16

CITATIONS READS
28 2,314

2 authors, including:

Linganagouda Kulkarni
B.V. Bhoomaraddi College of Engineering and Technology (BVBCET)
50 PUBLICATIONS 355 CITATIONS

SEE PROFILE

All content following this page was uploaded by Linganagouda Kulkarni on 18 September 2018.

The user has requested enhancement of the downloaded file.

 International Journal of Hybrid Information Technology
Vol.7, No.4 (2014), pp.185-200
http://dx.doi.org/10.14257/ijhit.2014.7.4.16

Image Encryption Based on 2-D Zaslavskii Chaotic Map and Pseudo

Hadmard Transform

Gururaj Hanchinamani1 and Linganagouda Kulakarni1

1
Department of Computer Science and Engineering, BVB College of Engineering and
Technology, Hubli, Karnataka, India 580031
gs_hanchinamani@bvb.edu, linganagouda@yahoo.co.uk

.
Abstract

AL
Recently, several chaotic based image encryption schemes have been proposed, each of them
has its own strengths and limitations more or less in terms of security level and computational

G
speed. In this paper, a novel approach for image encryption based on a 2-D Zaslavskii map and

IL .
is nly
Pseudo Hadmard transform is proposed. The encryption process is composed of two stages, i.e.

LE
permutation and diffusion. The permutation is achieved by scrambling rows and columns using
chaotic values. This stage substantially reduces the correlation between neighboring pixels.
fil O
During diffusion, the avalanche effect is achieved with 2-D Pseudo Hadmard transforms
followed by diffusion in two directions (forward and backward) with multiple additions and xor
is n
operations. This stage ensures resistance against differential attacks. The security and
th rsio

performance of the proposed method is analyzed thoroughly by using key-sensitivity, key-space,

e
statistical, entropy, differential and performance analysis. The proposed scheme achieves the
required level of security with only one round of encryption operation. Hence the proposed
by e

method is computationally fast.

e eV

Keywords: Differential attacks, Image encryption, Statistical analysis, 2-D Zaslavskii map,
2-D pseudo Hadmard transform
ad in

1. Introduction
m nl

The intrinsic features of images are bulk volume of data, high correlation among adjacent
O

pixels, high redundancy and human perception of decrypted image with small distortions.
Hence images are considerably different from text information. The conventional encryption
methods such as AES, IDEA, DES, RSA etc., are computationally intensive hence consume
more time and are not suitable for images [3-7]. Several image encryption algorithms are
already available in the literature, however, some of these suffer with brute-force attack,
statistical attack, and differential attacks. In this paper, the computational performance and
ok

security of the encryption process is improved by using 2-D Zaslavskii map and Hadmard
transform.
Bo

During the last years, several image encryption schemes have been proposed in the spatial
domain, among which chaotic-based methods are most popular. The encryption algorithms
based on chaos offer the advantages to be very sensitive to the initial conditions, ergodicity,
randomness and simplicity [7]. Chaotic encryption systems generally have high speed with low
cost, which makes them better candidate than conventional methods for multimedia data
encryption.
The remaining of the paper is organized as follows. In Section 2, the literature survey is
presented. The 2-D Zaslavskii map is described in Section 3. In Section 4, the proposed
encryption scheme is discussed in detail. Simulation results and security analysis are presented

ISSN: 1738-9968 IJHIT

Copyright ⓒ 2014 SERSC
International Journal of Hybrid Information Technology
Vol.7, No.4 (2014)

in Section 5 to show the efficacy and validity of the algorithm. Finally, conclusions are drawn in
the last section.

2. Literature Survey
The general architecture of the chaotic based image encryption systems typically consists of
iteration of two stages (i) permutation and (ii) diffusion. The permutation is achieved by
scrambling all the pixels as a whole using 2D chaotic maps [1, 10, 11]. In diffusion stage, the
pixel values are altered sequentially and the change made to a particular pixel depends on the
accumulated effect of all the previous pixel values. However, as many rounds of permutation
and diffusion or iterations should be taken, the overall encryption speed is slow.
A brief overview of the recently proposed chaotic based encryption schemes is given

.
hereafter. To resist brute-force attacks the key space is increased by using multiple chaotic maps

AL
in [1, 10, 7]. Shatheesh Sam [2] proposed image encryption based on intertwining chaotic maps
to enhance security and keylength. Authors of [3] proposed symmetric encryption scheme based

G
on cyclic elliptic curve and chaotic system, which encrypts 256-bit of plainimage to 256-bit of

IL .
is nly
cipherimage within eight 32 bit registers. Guodong Ye [4] proposed an image encryption

LE
scheme with generalized Arnold map, as the key stream depends on the processed image the
method can resist known- and chosen-plain text attacks. Authors of [5] proposed a gray level
fil O
encryption scheme to eliminate image outlines and to disrupt the distributional characteristics of
gray level. In [6], the computational time is reduced by encrypting significant data in spatial
is n
domain and insignificant data in wavelet domain. Xiaoling Huang [8] proposed encryption
th rsio

scheme based on chaotic Chebyshev generator with multiple permutations to enhance the
e
decorrelation. The authors of [9] proposed image encryption with circle map and is resistant
against differential attacks. In [11], a hierarchy of 2-D piecewise nonlinear chaotic maps with an
invariant measure is introduced. In [12], an encryption scheme based on large pseudorandom
by e

permutation is proposed, which is combinatorially generated from small permutation matrices

e eV

based on chaotic maps. The authors of [13] combined the chaos-based image encryption with
pixel bit. This uses a single chaotic system applied directly to the position scrambling operation.
However, some of the chaotic based encryption schemes have been cryptanalyzed
ad in

successfully [14, 22]. Liang Zhao et. al., [22] presented differential attack on [13], and proposed
m nl

an improved scheme using self-correlations. Rhouma et. al., [14] presented attack on [23] with
only one pair of plaintext and ciphertext.
O

Based on the above discussions, though there exist several encryption schemes, each of them
has its own strength and limitations more or less in terms of security level and computational
speed. To resist statistical, differential, brute-force attacks and to improve the computational
performance, this paper proposes a novel chaotic image encryption scheme based on 2-D
Zaslavskii map and Pseudo Hadmard transform. The proposed method is resistant to brute-force
ok

attacks, statistical attacks and differential attacks with high computational speed. The proposed
approach achieves the required level of security with only one round of encryption operation. It
can be easily implemented and is computationally simple.
Bo

3. Chaotic Maps
Chaotic maps are nonlinear maps that exhibit chaotic behavior. The chaotic maps generate
pseudo-random sequences, which are used during encryption process. Chaotic maps are
sensitive to initial conditions and parameters, non-convergent, non-periodic and topologically
mixing. The proposed scheme uses 2-D Zaslavskii map and is discussed hereafter. The 2-D
Zaslavskii map is a discrete-time dynamical system, and is defined as

186 Copyright ⓒ 2014 SERSC

 International Journal of Hybrid Information Technology
Vol.7, No.4 (2014)

( ( ) ( )) ( )
( ( )) ( )

and
Where , are current chaotic values and , are next chaotic values and
are control parameters and is exponentiation. The key set for Zaslavskii map is
{ , }. Commonly used values for the parameters are = 12.6695, = 9.1, = 3.0.
The propositions of chaotic maps [12] are given in Eq. (3-5). The Zaslavskii chaotic output
sequence is analyzed by computing mean and self-correlations according the propositions given
in Eq. (3-5). It is observed that the mean value is close to 0.5 and the self correlations within the

.
sequence and across two sequences are very close to 0.

AL
Proposition 1. The mean value of the chaotic sequence is

G
IL .
∑ ( )

is nly
LE
Proposition 2. Self-correlation of a chaotic sequence is
fil O
is n
( ) ∑( )( ) ( )
th rsio
e
Proposition 3. Self-correlation function between two chaotic sequences is
by e
e eV

( ) ∑( )( ) ( )
ad in

4. Proposed Encryption Scheme

m nl

The algorithm consists of two stages, i.e., permutation and diffusion.

O

4.1. Permutation
The purpose of permutation is to reduce the high correlation between adjacent pixels in the
plain image. Let be a gray original image of size , it is a matrix containing rows and
columns, and the gray values ranges from 0 to 255. In the process of permutation, initially
chaotic values ( ) ( )} are generated by using Eq. (1, 2), after doing
ok

iterations in chaos maps. Let and . Then and are

Bo

sorted, and the positions of sorted chaotic values in the original chaotic sequence are found and
stored in and . The next step is to scramble row position of all values from first column
to last column according to . Similarly scramble column position of all values
from first row to last row according to . This stage shuffles all pixels and
decorrelates the neighboring pixels.

4.2. Diffusion
The diffusion function is employed to modify the gray values of the image pixels to confuse
the relationship between the plain image and the encrypted image. The diffusion function is

Copyright ⓒ 2014 SERSC 187

International Journal of Hybrid Information Technology
Vol.7, No.4 (2014)

used to ensure the plain image sensitivity i.e., a very little change in any one pixel of plain
image should spread out to almost all pixels in the whole image. The diffusion process contains
two steps. In the first step, the avalanche effect is introduced by using 2-D Pseudo Hadmard
transform and the second step performs diffusion in two directions (forward and backward) with
chaotic values, modulo addition and xor operations.

The 2-D Pseudo Hadmard transform (2-PHT) is defined as,

( ) ( )
( ) ( )
Where and are inputs to 2-PHT and and are outputs.

.
and its inverse is defined as,

AL
( ) ( )
( )

G
( )

IL .
is nly
LE
The avalanche effect is achieved by applying 2-PHT according to the Figure 1 on a block of
eight pixels at a time and for the entire image.

fil O
is n
th rsio
e
by e
e eV
ad in
m nl
O

Figure 1. Application of 2-PHT for the Avalanche Effect

In the second step, initially chaotic values ( ) ( ) } are

ok

generated by using Eq. (1, 2) after doing iterations in chaos maps. The real chaotic sequences
are transformed to integer form by using equation (10).
Bo

( ) ( )
Where is real chaotic value, is transformed integer value, and is 256 for 8-bit gray
image.
The 2-D permuted image is transformed to 1-D array by scanning the image left to
right and top to bottom. Diffusion of the processed image is performed by using and
chaotic sequence of 2-D Zaslavskii map and also previously diffused pixels. The computed
encrypted pixel values depends on the previously encrypted pixels and chaotic sequences, hence

188 Copyright ⓒ 2014 SERSC

 International Journal of Hybrid Information Technology
Vol.7, No.4 (2014)

algorithm shows resistance to the differential attacks such as known plain-text attack and known
cipher-text attack.
The forward diffusion is performed by using following equation,
((((( ) ) ) ) ) ( )
Where indicates modulo addition, is bitwise , is current pixel, and are
previously encrypted pixels, is permuted pixels, and are the 2-D Zaslavskii chaotic
values. and can be considered as constants.
The backward diffusion is performed using following equation to make the influence of
every pixel equal.
((((( ) ) ) ) )

.
AL
( )
Where is current pixel, and are previously encrypted pixels, is forward

G
diffused image pixels, and are the 2-D Zaslavskii chaotic values and and

IL .
is nly
LE
can be considered as constants. Finally, the encrypted image is obtained after the
diffusions using Eq. (11) and Eq. (12) in two directions.

fil O
4.3. Encryption Algorithm
is n
The encryption algorithm is composed of thirteen steps.
th rsio

Step 1. Read the original image and store the pixel values in the matrix
e
.
Step 2. Generate chaotic values of sequence ( ) and chaotic values of
sequence ( ) using Eq. (1, 2)
by e

Step 3. Copy chaotic values to and chaotic values to .

e eV

Step 4. Sort and , find the position of sorted chaotic values in the original chaotic
sequence and store in and .
Step 5. Scramble all the rows by using .
ad in

Step 6. Scramble all the columns by using .

m nl

Step 7. Apply 2-PHT to the permuted image using Eq. (6, 7) according to Figure 1.
Step 8. Transform 2-D processed image to 1-D array i.e. dimension transform from
O

Step 9. Generate chaotic values ( ) ( )} using Eq. (1, 2).

Step 10. Transform real chaotic values to integers using Eq.(10).
Step 11. Perform forward diffusion using Eq. (11).
ok

Step 12. Perform backward diffusion using Eq. (12).

Step 13.Transform the 1-D encrypted array to 2-D array i.e., dimension transform from
.
Bo

4.4. Decryption
Decryption involves reconstructing gray levels of the original image from the encrypted
image. It is a simple inverse process of the proposed encryption algorithm.

Copyright ⓒ 2014 SERSC 189

International Journal of Hybrid Information Technology
Vol.7, No.4 (2014)

5. Experiments and Security Analysis

The proposed scheme is implemented on Linux platform using C language using a personal
computer with an intel (R) Core(TM) i3-2120 CPU at 3.30 GHz with 2.91 GB of RAM. The
initial parameters of the 2-D Zaslavskii map are randomly set to = 0.65, = 0.79, =
12.6995, = 9.1, = 3.0}. The Test images are gray-scale images of size chosen
from USC-SIPI image database (sipi.usc.edu/database/). An image encryption scheme should
resist the attacks such as brute-force attacks, statistical attacks, differential attacks and so on.
This section analyzes the properties of the proposed encryption scheme to show its effectiveness
in resisting these attacks.
The proposed encryption algorithm has been tested with several test images of differing
content. Figure 2 shows the visual inspection of the original, encrypted and decrypted images of

.
different images after applied only one round of encryption algorithm. The first row shows the

AL
original images, second row shows the encrypted images and the last row shows the decrypted
images. The encrypted images are non-recognizable in appearance, unintelligible,

G
incomprehensible, random and noise-like images without any leakage of the original

IL .
is nly
information. This demonstrates that the proposed algorithm can be used to protect various

LE
images for diverse protection. The decrypted images are exactly same as the original images.

fil O
5.1. Histogram Analysis
The histograms present the statistical characteristics of images. An image histogram plots the
is n
frequency of occurrences of each gray level. An encrypted image is expected to have no
th rsio

statistical similarity with the original image to prevent the leakage of information. The
e
histogram of original image consists of spikes with some shape. These spikes correspond to
gray values that appear more often in the image. The histogram of encrypted image is expected
by e

to be sufficiently uniform to resist statistical attack. The histogram of several plain images are
e eV

computed and analyzed. The histogram for Lena image is shown in Figure 3. The histogram of
the encrypted image is uniformly distributed and is completely different from that of the
original image, and bear no statistical resemblance to the original image. Hence the proposed
ad in

algorithm is resistant to statistical attacks.

m nl

5.2. Key-space Analysis

O

The key-space of an encryption system should be sufficiently large enough to resist brute-
force attacks. Brute-force attack is an attack where an opponent tries to break the cryptosystem
by exhaustive search with all possible keys. The proposed encryption scheme has two initial
values and three parameters, hence the key consists of totally five real values { , , , , }.
ok
Bo

190 Copyright ⓒ 2014 SERSC

 International Journal of Hybrid Information Technology
Vol.7, No.4 (2014)

.
AL
G
IL .
is nly
LE
fil O
is n
th rsio
e
Figure 2. Original Images, Encrypted Images and Decrypted Images with
Proposed Algorithm (a) Lena (b) Mandrill (c) Aerial (d) Earth from Space (e)
Pepper
by e
e eV
ad in
m nl
O
ok

Figure 3. Histograms of Original Image and Encrypted Image. (a) Original Image
Bo

(b) Histogram of Original Image (c) Encrypted Image (d) Histogram of Encrypted
Image

With 64 bits for each parameter and there are five real values, so the key-length is 320 bits
and the key-space is 2320. Hence the proposed algorithm has adequate key-space and resistant to
brute-force attacks. Table 1 shows the key-space size of the proposed algorithm and other
algorithms.

Copyright ⓒ 2014 SERSC 191

International Journal of Hybrid Information Technology
Vol.7, No.4 (2014)

Table 1. Key-space of the Proposed Method and Some of the other Methods in
the Literature
Encryption Proposed approach Ref.[13] Ref.[9]
scheme
Key-space size 2320 2128 2256

5.3. Information Entropy Analysis

Information entropy is used to measure the amount of unpredictability in information
content. Entropy tests whether an image is random kind of image with random distribution of

.
pixel values. Entropy is defined as,

AL
( ) ∑ ( ) ( )

G
( )

IL .
is nly
LE
Where represents the pixel values, ( ) is the probability of the symbol and r is the
number of symbols and is 256 for gray level image. Suppose the gray level image has gray
fil O
values with equal probabilities,
entropy value ( )
( ), according to Eq. (13), we obtain its
. In the original plain image there exist correlation and the pixel values
is n
are seldom random, hence the entropy value is generally smaller than the ideal value 8. The
th rsio

entropy reaches the maximum ideal value of 8 when all pixel values are randomly distributed.
e
Table 2 lists the entropy values for the original plain images and the encrypted images. From
the results it is observed that the entropy of encrypted images are very close to the ideal value of
8. The information leakage in the proposed encryption scheme is negligible and is secure
by e

against the entropy based attacks. The comparison of entropy values with other approaches are
e eV

listed in Table 3.

Table 2. Entropy Values for Original and Encrypted Images for Different Images
ad in

Entropy
m nl

Image original image encrypted image

O

Lena 7.426985 7.997620

Mandrill 7.242483 7.997126
Aerial 7.313656 7.997519
Earth 7.044457 7.997273
Pepper 7.577819 7.997282
ok

Table 3. The Entropy Analysis of the Proposed Scheme with other Methods for
Lena Image
Bo

Method Proposed Ref. [24] RC5 RC6 Ref. [23]

Approach
Entropy values 7.997620 7.9884 7.9812 7.9829 7.9923
5.4. Correlation Analysis
Generally, for any plain-image having visual content, each pixel is highly correlated with its
adjacent pixels in all the three directions: horizontal, vertical and diagonal. A good encryption

192 Copyright ⓒ 2014 SERSC

 International Journal of Hybrid Information Technology
Vol.7, No.4 (2014)

scheme should produce encrypted images with no such correlations in the neighboring pixels.
The correlation coefficient of adjacent pixels is calculated according to Eq. (14-17).

( ) ∑ ( )

( ) ∑( ( )) ( )

( ) ∑( ( )) ( ( )) ( )

.
AL
( )
( )
√ ( )√ ( )

G
IL .
Where and are adjacent pixels of original or encrypted images, ( ) is the mean value,

is nly
LE
( ) is the deviation with respect to mean, ( ) is the covariance between adjacent pixels,
and is the correlation coefficient. To test the correlation in the original and encrypted

fil O
images, 2048 pairs of adjacent pixels are randomly chosen in horizontal, vertical and diagonal
directions, and their correlation coefficients are calculated using Eq. (17). The Table 4 lists the
is n
computed correlation coefficients of original and encrypted images for different images. From
th rsio

Table 4 it can be seen that the two adjacent pixels in the original image are highly correlated to
each other, whereas the correlation coefficients for encrypted images are very close to zero.
e
Hence the proposed approach is resistant to statistical attacks. The comparison of correlation
results with other methods are given in Table 5.
by e
e eV

Table 4. Correlation Coefficients of Adjacent Pixels in Different Directions for

Original and Encrypted Images
Correlation coefficients for original image Correlation coefficients for encrypted images
ad in
m nl

Image Horizontal Vertical Diagonal Horizontal Vertical Diagonal

O

Lena 0.968683 0.943269 0.933408 -0.002689 -0.011126 0.001347

Mandrill 0.772042 0.845529 0.740922 -0.000542 -0.007360 -0.001562
Aerial 0.842869 0.901867 0.778739 -0.004581 -0.003564 0.005904
Earth 0.944232 0.936486 0.901125 -0.000688 0.006161 0.003926
Pepper 0.966785 0.960691 0.940792 0.001079 0.004488 0.004205
ok

Table 5. The Correlation Analysis of the Proposed Scheme with other Methods
for Lena Image
Bo

Method Plain-image Proposed scheme AES Chen’s Arnold’s

Horizontal 0.977352 -0.002689 -0.0160 0.0442 0.0787
Direction Vertical 0.851794 -0.011126 0.8018 0.9728 -0.0793
Diagonal 0.761644 0.001347 -0.0140 0.0469 -0.0633

5.5. Gray Value Degree ( ) Analysis

The gray difference of a pixel with its four neighbors in an image can be calculated as
follows,

Copyright ⓒ 2014 SERSC 193

International Journal of Hybrid Information Technology
Vol.7, No.4 (2014)

( )
∑ ( ) ( ) ( )
( ) { ( )
( )
( )
Where ( ) represents the pixel value at location( ), and ( ) denotes the pixel
values of four neighborhood pixels. The average neighborhood gray difference for the complete
image can be calculated by Eq. (19).
∑ ∑ ( )
( ( )) ( )
( ) ( )
Where and are the number of rows and columns of the image. By using (18) and (19)

.
the gray value degree is defined as,

AL
( ( )) ( ( ))
( )
( ( )) ( ( ))

G
IL .
is nly
LE
Where and denote the average neighborhood gray difference of original plain image
and encrypted image. Table 6 lists computed gray value degree values for different images by
the proposed approach. From Table 6 it is observed that the gray value degrees computed by the

other approaches.
fil O
proposed method are close to the ideal value of 1. Table 7 shows the comparison of with
is n
th rsio

Table 6. Gray Value Degree Values for Different Test Images

e
Image value
Lena 0.962362
by e

Mandrill 0.903770
Aerial 0.917014
e eV

Earth 0.958430
Pepper 0.963746
ad in

Table 7. The Analysis of the Proposed Scheme with other Approaches

m nl

value
Image Proposed Arnold’s Ref.[13]
O

approach
Lena 0.962362 0.89 0.954

5.6. Peak Signal to Noise Ratio ( ) Analysis

The can be used to perform objective evaluation of the encryption methods. The
ok

is computed by considering the original plain image as signal and the encrypted image as
a noise. The can be calculated by using the following formula,
Bo

( ) ( )
√
Where is mean square error and is computed according to Eq. (22)

∑ ∑( | ( ) ( )| ) ( )

194 Copyright ⓒ 2014 SERSC

 International Journal of Hybrid Information Technology
Vol.7, No.4 (2014)

Where ( ) is pixel value of original plain image and ( ) is pixel values of encrypted
image at location ( ). The calculated values for different test images are listed in Table
8. From Table 8 it is observed that the values are low, which indicates the difficulty in
getting original plain image from the encrypted image for attackers.

Table 8. The Values for Different Test Images

Image (dB)

Lena 9.227133
Mandrill 9.707395
Aerial 9.282019
Earth 9.402777

.
AL
Pepper 8.890136

5.7. Key Sensitivity Analysis

G
IL .
Key sensitivity implies that the small change in the secret key should produce entirely

is nly
LE
different encrypted image. The key sensitivity test is conducted by using following steps.
Step 1. The original plain image is encrypted by using a test key to produce cipher image
fil O
Step 2. The original plain image is encrypted again with a small change in the test key
to produce cipher image .
, i.e.,
is n
Step 3. The two cipher images and with slightly different keys are compared pixel by
th rsio

pixel to observe the number of differing pixels.

e
The two parameters and are used to assess the key sensitivity and are discussed
below.
by e

(Number of Pixels Change Rate) is used to compute the total number of different
e eV

pixels in two images and is calculated by using Eq. (23).

∑ ( )
( )
ad in

( ) ( )
( ) { ( )
m nl
O

Where and are two encrypted images with slightly different keys and . ( )
and ( ) are the pixel values of and at location ( ). is a bipolar array with the
same size as and and its contents are either or based on Eq. (24).
(Unified Average Changing Intensity) is used to compute the average intensity
difference between two encrypted images and is given by,
ok

( ) ( )
[∑ ] ( )
Bo

The key sensitivity is analyzed by testing one parameter at a time with a very small change in
the key. The proposed scheme has five parameters { , }. Table 9 lists the
and values for five different parameters. From Table 9 it can be seen that the and
values are close to their ideal values of 99.6% and 33.4%. Hence the proposed scheme
has high key sensitivity.

Copyright ⓒ 2014 SERSC 195

International Journal of Hybrid Information Technology
Vol.7, No.4 (2014)

Table 9. Key Sensitivity Results for Lena Image with Different Parameters of the
Chaotic Map
Parameter (%) (%)
changed
X 99.645996 33.375423
Y 99.615479 33.457809
99.603271 33.551277
99.630737 33.457211
99.617004 33.406235

Key sensitivity is also tested pictorially with the following approach. The original key is

.
altered with a small change and a different key is generated. The keys can be expressed as,

AL
original key = (0.65, 0.79, 12.6695, 9.1, 3.0), and the slightly modified key =
(0.6500000000001, 0.79, 12.6695, 9.1, 3.0). The two encrypted images ( and ) with slightly

G
different keys ( and ) are shown in figure 4b-c respectively. Even though both look similar,

IL .
they are almost different from each other. This can be observed by finding the difference image

is nly
LE
between and . Figure 4d shows the difference image between and . It is observed that
most of the pixels in difference image are nonzero, hence the difference is big enough.
fil O
is n
th rsio
e
by e
e eV

Figure 4. Key Sensitivity Analysis for Encryption Process for Lena Image (a)
ad in

Original Image (b) Encrypted Image with Correct Key (c) Encrypted Image
m nl

with Slightly Different Key (d) Difference Image between and

O

The key sensitivity test is also analyzed for decryption process. The decryption is performed
with correct key and slightly different keys. Figure 5a shows the decrypted image with correct
key = (0.65, 0.79, 12.6695, 9.1, 3.0) and Figure 5b-c are decrypted images with slightly
altered keys = (0.6500000000001, 0.79, 12.6695, 9.1, 3.0) and = (0.65, 0.79, 12.6695,
9.100000000001, 3.0). Hence, if there is a small change in the key, the correct decryption
cannot be achieved.
ok

5.8. Plain-image Sensitivity Analysis

Bo

To resist differential attacks, the encryption algorithms should satisfy the plain image
sensitivity condition, i.e., a small change in the original plain-image should cause a significant
change in the encrypted image. The and parameters are used to test the plain-
image sensitivity as given in Eq. (23-25). The and values are computed for
different randomly chosen locations by changing one pixel at a time and the results are listed in
Table 10. From Table 10, it is observed that the and values are close to their ideal
values of 99.6% and 33.4% irrespective of the pixel position. Hence the proposed approach has
good sensitivity to plain-images.

196 Copyright ⓒ 2014 SERSC

 International Journal of Hybrid Information Technology
Vol.7, No.4 (2014)

Figure 5. Key Sensitivity Analysis for Decryption Process for Lena Image. (a)
Decryption with Correct Key (b-c) Decryption with Slightly Changed Keys

.
AL
Table 10. Plain-image Sensitivity Test at Different Positions for Lena Image
Position (%) (%)

G
(0,0) 99.752808 33.499668

IL .
is nly
LE
(35,80) 99.673462 33.456978
(100,150) 99.633789 33.421177
(128,128) 99.591064 33.421783
fil O (255,255) 99.586296 33.377789
is n
5.9. Computational Speed Analysis
th rsio

The proposed approach makes use of simple operations such as addition, subtraction and xor,
e
and these are not computationally intensive. Hence the approach can offer a fast and efficient
way for image encryption. The complexity of the proposed method is ( ), where and
by e

are the number of rows and columns of the image. The time needed to encrypt 256 256 gray
e eV

scale image is 3.43 micro-seconds and for decryption it is the same. So the proposed scheme
can encrypt 19106 Mb data per second. The comparisons of the encryption time with other
approaches are listed in Table 11.
ad in

Table 11. Execution Time Analysis using a 256x256 Image

m nl

Methods Encryption time

O

Guodong Ye 0.150 seconds

Gao. T.G 0.633 seconds
Ye.R.S >10 seconds
Huang X.L. 0.547 seconds
Proposed scheme 0.00000343 seconds
6. Conclusions
ok

In this paper, a novel approach for image encryption based on 2-D Zaslavskii map and
Bo

pseudo Hadmard transform is proposed. The proposed method offer high security and high
speed. It is implemented under Linux platform with C language, and the achieved speed is 3.43
µs, hence it is computationally efficient. The proposed approach has key space of 2320 , which is
large enough to prevent brute-force attacks. The average entropy achieved is 7.997364, which is
close to the ideal value of 8, and hence the information leakage is negligible. The and
values are close to their ideal values of 99.6% and 33.4% for key sensitivity and plain
image sensitivity tests and hence resistant to differential attacks. The correlations are close to
zero and the histogram is almost uniformly distributed, so statistical attacks are resisted. The
is near to 1 and the is lower. The results shown in section 5 are the obtained results

Copyright ⓒ 2014 SERSC 197

International Journal of Hybrid Information Technology
Vol.7, No.4 (2014)

after one round of encryption/decryption. So it is observed that even in the first round all the
security parameters are already high. The proposed approach can be extended for color images.

References
[1] A. A. Abd El-Latif, L. Li, T. Zhang, N. Wang, X. Song and X. Niu, “Digital image encryption scheme based on
multiple chaotic systems”, Sensing and Imaging, An international journal on continuing subsurface sensing
technologies and applications, Springer, vol. 56, no. 2, (2012), pp. 67-88.
[2] I. Shantheesh Sam, P. Devaraj and R. S. Bhuvaneswaran, “An intertwining chaotic maps based image
encryption scheme”, Nonlinear Dynamics, Springer, (2012), pp. 1995-2007.
[3] A. A. Abd El-Latif, L. Li and X. Niu, “A new image encryption scheme based on cyclic elliptic curve and
chaotic system”, Multimedia tools and applications, Springer, vol. 70, (2014), pp. 1559-1584.
[4] G. Ye and K.-W. Wong, “An efficient chaotic image encryption algorithm based on a generalized Arnold map”,
Nonlinear Dynamics,Springer, (2012), pp. 2079-2087.

.
AL
[5] C. K. Huang, C. W. Liao, S. L. Hsu and Y. C. Jeng, “Implementation of gray image encryption with pixel
shuffling and gray-level encryption by single chaotic system”, Telecommunication systems: Springer, vol. 52,
no. 2, (2013), pp. 563-571.

G
[6] N. Taneja, B. Raman and I. Gupta, “Combinational domain encryption for still visual data”, Multimedia tools

IL .
and applications An international journal, Springer, vol. 159, no. 3, (2012), pp. 775-793.

is nly
LE
[7] M. Francois, T. Grosges, D. Barchiesi and R. Erra, “A new image encryption scheme based on a chaotic
function”, Signal Processing; Image Communications: Elsevier, vol. 27, no. 3, (2012), pp. 249-259.
[8] X. Huang, “Image encryption algorithm using chaotic chebyshev generator”, Nonlinear dynamics, Springer,

[9] fil O
(2011), pp. 2411-2417.
D. Chattopadhyay, M. K. Mandal and D. Nandi, “symmetric key chaotic image encryption using circle map”,
Indian journal of science and technology, vol. 4, (2011), pp. 593-599.
is n
[10] S. Sam, P. Devaraj and R. S. Bhuvaneswaran, “A novel image cipher based on a mixed transformed logistic
th rsio

maps”, Multimedia tools and applications, An international Journal, Springer, vol. 56, no. 2, (2012), pp. 315-
e
330.
[11] J. won Yoon and H. Kim, “An image encryption scheme with a pseudorandom permutation based on chaotic
maps”, Communications in nonlinear science and numerical simulations, Elsevier, vol. 15, no. 12, (2010), pp.
by e

3998-4006.
G. Ye, “Image scrambling encryption algorithm of pixel bit based on chao map”, Pattern recognition letters,
e eV

[12]
Elsevier, vol. 31, no. 5, (2010), pp. 347-354.
[13] R. Rhouma, E. Solak and S. Belghith, “Cryptanalysis of a new substitution-diffusion based image cipher”,
Communications in nonlinear science and numerical simulations, Elsevier, vol. 15, no. 7, (2010), pp. 1887-
ad in

1892.
[14] S. Behnia, A Akhshani, S. Ahadpour, H. Mahmodi and A. Akhavan, “A fast chaotic encryption scheme based
m nl

on piecewise nonlinear chaotic maps”, Physics letters, Elsevier, (2007), pp. 391-396.
[15] S. Lian, J. Sun and Z. Wang, “A block cipher based on a suitable use of the chaotic standard map”, Chaos,
O

Solitons and fractals, Elsevier, vol. 26, no. 1, (2005), pp. 117-129.
[16] G. Chen, Y. Mao and C. K. Chui, “A symmetric encryption scheme based on 3D chaotic maps”, Chaos,
Solitons and fractals, Elsevier, vol. 21, no. 3, (2004), pp. 749-761.
[17] I. M. T. Al-shara’a and D. mohomod kreem Al-Ftlawy, “The dynamics of the 2-D piecewise Tinkerbell map”,
Mathematical theory and modeling, vol.3, no.8, (2013), pp. 121-132.
[18] G. Alvarez and S. Li, “Some basic cryptographic requirements for chaos-based cryptosystems”, International
journal of bifurcation and chaos, vol. 16, no.8, (2006), pp. 2129-2151.
ok

[19] L. Zhao, A. Adhikari, D. Xiao and K. Sakurai, “On the security analysis of an image scrambling encryption of
pixel bit and its improved scheme based on self-correlation encryption”, Communications in nonlinear science
Bo

and numerical simulations, Elsevier, vol. 17, no. 8, (2012), pp. 3303-3327.
[20] V. Patidar, N. K. Pareek and K. K. Sud, “A new substitution diffusion based image cipher using chaotic
standard and logistic maps”, Communications in nonlinear science and numerical simulations, Elsevier, vol. 14,
(2009), pp. 3056-3075.
[21] N. K. Pareek, V. Patidar and K. K. Sud, “Image encryption using chaotic logistic map”, Image Vision and
computing, vol. 24, (2006), pp. 926-934.

198 Copyright ⓒ 2014 SERSC

 International Journal of Hybrid Information Technology
Vol.7, No.4 (2014)

Authors
Linganagouda Kulakarni, he received PhD degree in pattern
recognition from Mysore university, India. His research interests are
image processing, computer networks and information security. He is
currently working as professor at computer science department, BVB
college of engineering and technology, Hubli, India.

Gururaj Hanchinamani, he received ME degree in computer science

.
AL
and engineering from Walchand college of engineering Sangli, India. He
is pursuing PhD degree at Visvesvaraiah technological university
Belgaum. His research interests are information security and computer

G
IL .
architectures. He is currently working as associate professor at computer

is nly
LE
science department, BVB college of engineering and technology, Hubli,
India.

fil O
is n
th rsio
e
by e
e eV
ad in
m nl
O
ok
Bo

Copyright ⓒ 2014 SERSC 199

 International Journal of Hybrid Information Technology
Vol.7, No.4 (2014)

.
AL
G
IL .
is nly
LE
fil O
is n
th rsio
e
by e
e eV
ad in
m nl
O
ok
Bo

200 Copyright ⓒ 2014 SERSC

View publication stats