International Journal of Scientific & Engineering Research Volume 12, Issue 6, June-2021 12

ISSN 2229-5518

Image Encryption Using Chaos Maps

B. Sai Kumar1, L. Vikhyath2, R. Geetha Krishna Pavansai3
1,2,3
Student, Sreenidhi Institute of Science and Technology, Ghatkesar, Telangana

Abstract— As the exchange of data over the open networks and Internet is rapidly growing, security of the data becomes a major concern.
The solution is to encrypt the data. The data can be text, image, audio, video etc. Images are one of the largest media in this generation.
Earlier image encryption techniques like Advanced Encryption Standard (AES), Data Encryption Standard (DES), Rivest–Shamir–Adleman
(RSA) etc. exhibit low levels of security and weak anti attack ability. This problem was overcome by using chaos-based cryptography. The
chaotic systems are very sensitive to initial conditions and control parameters which make them suitable for image encryption. The main aim
of this project is to examine the efficiency of different chaotic maps, Chaos technology encrypts images such that the possibility of deciphering
is reduced greatly as the cipher text presents a randomness. Chaos based encryption is a very important means of modern digital encryption.
The statistical characteristic if the original image is transformed, thereby increasing the difficulty of unauthorized breaking of the encryption.
Various performance metrics like peak signal to noise ratio (PSNR), mean-square error (MSE) and execution time evaluation are performed
on the obtained decrypted image.

Index Terms— Chaos Maps, Henon Map, Arnold Cat Map, Logistic Map, Image Encryption, PSNR, MSE, Autocorrelation.

——————————  ——————————

1 INTRODUCTION
Over the past two decades, the world has witnessed of strong keys makes it difficult for the cryptanalyst to guess
tremendous advancements in the area of multimedia-based the key.
applications like medical imaging, and multimedia
image/video database services. The rapid growth of digital Chaotic systems are highly sensitive. So, if the starting
media and multimedia-based applications has increased the circumstances or control parameters are chosen as "Keys"
requirement of transmitting them over public networks. The and "Pathways" for encryption/decryption, chaotic maps can
key obstacle in such applications for scientific and research be beneficial in encryption/decryption schemes. Chaos-
community has been the efficient and secure transmission based encryption systems are symmetric since they employ
over the network. A digital image is either a gray image the same parameters for encrypting and decrypting. In
represented as two-dimensional matrix having pixel chaotic maps, ergodicity and sensitivity to beginning
intensities or a colored image represented as three- circumstances show excellent confusion and diffusion
dimensional matrix or RGB matrix where the three planes of qualities, which are critical for an efficient encryption
matrix correspond to red, green and blue respectively. Since strategy. Further, the control parameters and initial
images are bulky in data, therefore encryption algorithms conditions of chaotic maps form a very large key space
should be designed in such a way that are faster, involve enhancing the security against brute force attack.
fewer complex computations and secure to various possible
attacks.
2 LITERATURE SURVEY
One of the most essential ways for ensuring the security of
The search for new encryption schemes suitable to image/
digital photographs and videos transmitted over the internet
video data, improvements to existing cryptographic
is encryption. Because of the inherent characteristics of
techniques and proofs of security continues at a rapid pace.
image/video, such as mass data volume, strong correlation
A number of encryption algorithms suitable for image data
between neighboring pixels, and high data redundancy,
has been suggested in literature. Chaos is one such
encryption methods that can handle the underlying main
dimension widely used in such algorithms. The
challenges of data storage, speed, and security are required.
phenomenon of chaos theory was first introduced by
Various real-time image/video encryption approaches have
Edward Lorenz in 1972 with conceptualization of “Butterfly
been proposed in the literature, which are faster and more
Effect". There is a link between chaos features and traditional
secure than traditional encryption algorithms. One of the
cryptography, according to many academics.
paradigms that appears to be useful for cryptography is
According to Q.V. Lawande(2005), the strength of chaos. Various encryption strategies have been proposed in
cryptography is found in the selection of strong keys, which the literature that use chaos to increase the security of an
are confidential parameters utilized in encryption. Selection

IJSER © 2021
http://www.ijser.org
International Journal of Scientific & Engineering Research Volume 12, Issue 6, June-2021 13
ISSN 2229-5518

encryption scheme due to its properties. the same point of origin. X1 and Y1. Therefore, the
encryption and key sensitivity algorithms work together to
2.3 Motivations of the Proposed Research Work prevent all types of code-breaking attacks.
Traditional encryption algorithms such as DES (Data
Encryption Standard), RSA (named after Rivest, Shamir, Xn+1 = 1+Yn – 1.4Xn 2
Adleman) & AES (Advanced Encryption Standard) are
focused on number theory and insist spatial information, Yn+1 = 0.3Xn (3.2)
and are based on image/intrinsic video's characteristics, e.g.
massive data capacity, strong correlating between adjacent 3.1.1 Image Encryption by Henon Chaotic System
pixels and high data redundancy. For multimedia-based
The shuffled image is encrypted using pseudo-random
encryption techniques, data storage, speed and security are
binary sequence generated by taking key values for Henon
the key issues.
map.
As seen from literature survey, a lot of research work has
been carried out in these directions but there are still Step 1: Choose the starting value of (X1, Y1) on the Henon
possibilities of making image encryption process more card. This value serves as the initial secret symmetric key for
efficient and robust. Motivations for this research work are the Hénon card.
listed below-
Step 2: The Hénon card acts as a key stream generator for
Chaotic values are useful in generating keys, hash functions, cryptographic systems. The dimensions of the sequence is
digital signatures etc. which are useful in cryptography. Use the size of the image and the number of henon sequence is
of chaos has made the encryption process comparatively obtained through equation (3.2) when the size of the image
faster because of ease to generate long chaotic is m/n.
pseudorandom sequences whose values seems to be
uncorrelated if initial value and related parameters for Step 3: The empirical analysis concluded that the threshold
generation of sequence are not known but still security is a of 0.3992 was determined to balance the series. Then, as
big concern and an efficient encryption technique robust to
shown in equation (3), the decimal value is converted to a
attacks remains an ongoing challenge in research
community. The work proposes an efficient chaos-based binary value based on this threshold value. Where Z is a
encryption technique that is analyzed in conjunction with binary sequence.
different security parameters like statistical analysis, key-
space analysis, analytics of key sensitivity, entropy analysis, Zi = (3.3)
sound and block loss and other encryption schemes.
Step 4: The Henon pattern will be reduced to a decimal value
by integrating all eight straight bits.
3 METHODOLOGY
Step 5: Encryption is performed using an OR operation that
3.1 Henon Map Image Encryption Algorithm
excludes every bit between the composite image and the
The chaotic map of Henon is used as a symmetric chipboard sequence generated in step 4.
system. The Henon chaotic map creates pseudo-random
3.1.2 Decryption of Encrypted Image
binary code, which was described below, two dimensional
non-linear dynamic time The chaotic behavior of the system is deterministic, so it
Xn+1 = 1+Yn – aXn2 finally reconstructs the image using the same keys (X1, Y1).
Yn+1 = bXn n=0, 1, 2… (3.1) How to decode a mixed image. This composite image is
The parameters a and b are of parameters because the placed in the exact reverse order of encoding, and the
system's dynamic behavior depends on these values. The original image is captured by the receiver.
system can still be unpredictable only if the values of a and
b are 1.4 and 0.3. Turbulent, discontinuous, or periodic 3.2 Arnold Cat Map Image Algorithm
orbits. The starting points X1 and Y1 serve as symmetric keys
of the chaotic cryptosystem used for encryption on the Chaos is a common technology used in the generator of
sender side and decryption on the receiver side. Since the random numbers, because this algorithm is more simple and
Hénon map is deterministic, the decoding of the encoded can be used both for storage and process objects in the
image reconstructs the original image on the receiver side at process stream bound objects, only a few functions (chaotic

IJSER © 2021
http://www.ijser.org
International Journal of Scientific & Engineering Research Volume 12, Issue 6, June-2021 14
ISSN 2229-5518

maps) and some (first conditions) are used quite well if the 2. if d = (5^i) for i >=1, n = 2*d
process takes a considerable length of time.
3. if d = 6(5^i) for i>=1, n = 2*d
Arnold's Cat Map is a chaotic two-dimensional image that
can be applied to alter the image pixels position without 4. else n <= 12*d / 7 (4.3)
discarding any information from the image; S = { (x, y)| x, y
Where d is the decrypted image matrix
= 0, 1, 2... N-1}. N. The following equation can be used to
write the 2-D image of the Arnold Cat Map:

(4.1)

When p and q are positive, the determinant (A) = 1. (X' y'), Figure 4.1 Decryption an image flow diagram
when Arnold's cat mapping algorithm is performed, this is
The below is an example to illustrate Arnold cat map
the new location of the original pixel (x, y). From the cat map
encryption:
of Arnold to the number of iterations, R represents a random
drawing containing in the original image all values for the
same pixel. The number of iterations for R to carry out varies
according to p, q and N factors. The cat mapping algorithm
of Arnold however has the p, q and R iterations, all of which
can be used as a secret key. Arnold's algorithm for cat
mapping:

Step 1: Get image matrix and gray scaled image matrix.

Step 2: Arnold Cat encryption

 Read image
As a first test, we consider the RGB values of 3x3 pixels for
 Iterate until it reaches key the values shown in the Table, and its outcomes.

 Call Arnold Cat transform method

Step 3: Arnold Cat transform

 R([x,y]) = [ (x + y) mod n, ( x + 2y ) mod n ]

(4.2)
The next encryption and decryption algorithms are done
Obtained R matrix is the required map that is used in
after the pixel value has been obtained. The next procedure
encryption and decryption.
is Arnold's Cat Map The process of encryption and
Step 4: Arnold Cat Map decryption decryption involves pixel wiggle and pixel hop as follows:

 The number of iterations 'n' at which the original

image will reappear is given by these rules of
thumb:

 Here 'd' is the dimension of the square image:

1. if d = 2(5^i) for i >=1, n = 3*d

IJSER © 2021
http://www.ijser.org
International Journal of Scientific & Engineering Research Volume 12, Issue 6, June-2021 15
ISSN 2229-5518

The alphabetic characters (0–9 & A–F) are the ki's here, and
the session key is represented in each group of two
alphanumeric characters. The secret key can also be
displayed as an ASCII mode

K = K1K2.K10 (ASCII); (5.2)

Every Ki here is an 8-bit secret key block, i.e. session key.

2. Two chaotic logistic maps in the proposed

algorithm have been used to achieve the following
objective of image encryption:
Xn+1 =3:9999Xn(1-Xn); (5.3)

Yn+1 =3:9999Yn(1-Yn): (5.4)

Consistent maintaining the system parameter values in the

algorithm of the two logistic maps (3.999). That is a very
chaos case, but for those maps the original conditions (X0
and Y0) are calculated with a certain amount of data. Session
key operation.

3. To calculate the initial condition X0 for the first

logistics map, select three blocks of the session key.
Convert to binary string like, K4K5K6:

B1 = K41K42…K48K51K52…K58K61K62…..K68; (5.5)
This process represents an iterative procedure – for the
process of iteration, first, the number of pixels depending on Here, the kijs are the session key's binary numbers (0 or 1).
the user's wishes is versatile, the author uses one iteration We then use the above binary representation to calculate a
only to demonstrate the Cats map algorithm and, after this real number X01:
iteration, the results are the following pixels:
X01 = (K41x20 +K42x21 +……+ K48x27 + K51x28 + K52X29 +…+
K58X215 + K61X216+K62X217+…..+K68X223)/224. (5.6)

Further, we compute another real number X02 as follows:

(5.7)
3.3 Logistic Map Image Encryption Algorithm
Here, as explained in Eq, ki's are secret key elements in
Process of step-by-step image encryption and decryption hexadecimal mode. [1]. We are now calculating the initial X0
using two chaotic logistic maps. condition for the first X01 and X02 logistic map.

1. An 80-bit exterior private key is used in the X0 = (X01 + X02) mod 1 (5.8)
preferred image encryption procedure. The private
key is also split into 8-bit blocks, each of which is 4. Repeating the first logistic diagram using the initial
referred to as a session key. conditions obtained in step 3, we generate a
sequence of 24 real numbers f1, f2. .., f24. Other
K = k1k2.k20 (hexadecimal); (5.1) values will be removed from the string. A sequence
of real numbers is converted to a sequence of whole

IJSER © 2021
http://www.ijser.org
International Journal of Scientific & Engineering Research Volume 12, Issue 6, June-2021 16
ISSN 2229-5518

numbers according to the following formula: For the next 15 pixel image file repeat the steps 6 and 7

8. After encrypting a 16-pixel block of the image file,

session keys K1 through K9 are modified as follows:

5. We choose three blocks of session keys, i.e. K1K2K3, (5.11)

and convert it into a binary string in order to
calculate initial condition Y0 for a second logistic The first logistic map (i.e., X0 = f24) after changing the private
map: key as above and getting the final value of X in step 4 as the
conditional initial value is also explained in step 4. Generate
a sequence of 24 real numbers as described above, and repeat
Here kij’s are the session key i th block binary numbers (0 or steps 5 to 7 until all the image files are exhausted.
1). We then use the above binary representation to calculate
This is very similar to the encryption method described
the true Y01 number:
above. The only difference is at step 7. That is, for the
cryptographic process, the second logistics label (K10) is
repeated 10 times, after which the corresponding
(5.9) cryptographic operation (depending on the result of the
logistics card) is performed. The second logistic map (K10) is
In addition, we calculate the following real number Y02: iterated 10 times initially for the decryption process, then the
appropriate decryption operation is performed (in reverse
order again, depending on the outcomes of the following
maps).
(5.10)
4 ANALYSIS
B2[Pk] in this case denotes that the Pkth bit value is either 0
or 1. In the B2 binary string, i.e. Now we compute the The analysis of these algorithms is done by plotting Intensity
primary state Y0 for the second logistic map using Y01 and histograms and adjacent pixel autocorrelation graphs.
Y02 as:
4.1 Intensity Histogram
Y0 = (Y01 + Y02) mod 1: (13)
The histograph of the picture generally refers to the pixel
6. Read three bytes consistently from the image file. intensity value histogram in the context of the image
These three bytes are red, green, and blue (RGB) processing. This histogram is a histogram showing the
values and form one pixel of the picture together. number of pixels for each of the distinct picture intensity
values. Grayscale pictures may be 256 distinct intensities,
7. Range [0,1,0.9] Place in 24 intersecting spaces and in such that the histogram depicts 256 pixel allocation numbers
8 various groups. Then allocate to each of these graphically between these values of the grayscale. The
groups different types of activities. Table 1 presents histogram of a colored picture is also available. You can get
every operation's groups and time range. separate histograms for the red, green, and blue channels, or
you can create a 3D histogram showing the red, blue, and
green channels and their respective brightnesses on three
Once again, use the first condition Y0 of step 5, reiterate the axes. Score represents the number of pixels. The operation is
second logistic map. The result of the second logistic map subject to implementation. This might be an image of the
determines what the red, green and blue (RGB) bytes are to chart in a suitable picture format, or a data file describing the
be encrypted / decrypted. This step is (K10) 10 (i. Finally, the chart's stats.
red, green and blue bytes are entered in the file (decimal
number corresponding to the key of the 10th session).That is, It's really simple to operate. The image was examined in a
a single pixel encoding is performed. single scan, retaining and generating the suitable histogram
the amount of consecutive pixels that was identified at each

IJSER © 2021
http://www.ijser.org
International Journal of Scientific & Engineering Research Volume 12, Issue 6, June-2021 17
ISSN 2229-5518

intensity value. Three algorithms show changes in pixel pixel correlation. The image encoding/encoding algorithms
intensity on a histogram. will produce an output image that is significantly (i.e.,
statistically) different from the pure version. In this context,
The Intensity histograms of original, encrypted images for the main limitation of traditional assessment methods is that
all 3 algorithms shows the variation of pixel intensity they provide qualitative rather than qualitative
through a graph. measurements. As a result, many researchers have come
together in recent years to improve traditional assessment
The image treatment describes a digital image's tone
methods and develop new statistical tests for the
distribution using an intensity histogram. The observer can
randomness of images.
evaluate the full tonal distributed on a glance by glancing at
the histogram of the given image. The horizontal axis Modified during encoding and uses histogram analysis to
indicates the change in the colour and the vertical axis shows describe the distribution of pixels of a single image relative
the number of pixels. to the encoded image. Histogram analysis is generally
limited to visual assessment, i.e., whether a uniform
distribution of the histogram of the encoded picture is
considerably different from the picture of the origin.
However, a more detailed analysis evaluates the relevance
of the histogram. That is, we use the chi-square test to
evaluate it as a qualitative measure of the distribution of
histogram values in a coded image that approaches the
characteristics of a uniform distribution.

The digital image coding scheme, for coded images, makes

these correlations as close as possible to the zero values of
the individual images. The same set of adjacent pixels is well
distributed in the histogram.

Intensity Histogram of Encrypted image (Henon Map)

Intensity Histogram for Encrypted image (Logistics Map)

4.2 Adjacent Pixel Autocorrelation

As new image encryption and encryption algorithms are

developed, researchers are using a variety of methods to
evaluate the effectiveness and security of the algorithms. For Adjacent pixel Autocorrelation of Encrypted Image
example, global Shannon entropy measurement, histogram
analysis, differential cryptographic analysis, and neighbor

IJSER © 2021
http://www.ijser.org
International Journal of Scientific & Engineering Research Volume 12, Issue 6, June-2021 18
ISSN 2229-5518

5 CONCLUSION 7 REFERENCES
In this project we saw execution and working of three main [1] Monjul Saikia, Vikash Beruah (2017) Chaotic Map Based
Chaos Algorithms namely Arnold map, logistic and henon Image Encryption in Spatial Domain: A Brief Survey. In:
map. The conclusion that we came to after this project is that: Mandal J., Satapathy S., Sanyal M., Proceedings of the
First International Conference on Intelligent Computing
 Arnold cat map takes more time to compute when and Communication. Advances in Intelligent Systems
compared to other two algorithms i.e., the run time and Computing, vol 458. Springer.
of this map is comparatively very high.
 Logistic map generates key based on the map [2] K. Sakthidasan, B. Krishna, “A New Chaotic Algorithm for
function and this key is used for further encryption. Image Encryption and Decryption of Digital Color Images”,
 Henon map has less execution time than that of International Journal of Information and Education
both the other algorithms i.e., it is faster. Technology, vol. 1, no. 2, June 2011

After doing performance measures we came to a conclusion [3] Nitu Hazarika, Manjul Saikia, “A Novel Partial Image
that: Encryption Using Logistic Map”, 2014 International
Conference on Signal Processing & Integrated Networks
 Logistic map has the least PSNR value among them. (SPIN), IEEE, 2014
 Arnold cat map encryption has the highest PSNR.
[4] S. Kromodimoeljo, Teori & Aplikasi kriptografi, Medan,
The conclusion here is that depending on the application in Indonesia: SPK Consulting, 2009.
which we need the encryption we need to choose our map.
If we need highest PSNR and Highest Encryption we need [5] R. Munir, Pengolahan Citra Digital, Jakarta, Indonesia:
to go for Arnold cat map but the catch here is we need to Informatika , 2004.
compromise our run time.
[6] N. A. Abbas, "Image encryption based on Independent
If execution time is more important than that of level of Component Analysis and Arnold’s Cat Map", Egyptian
encryption, we can choose Logistic map Encryption. Informatics Journal, pp. 139-146, 2016.

If we need to use encryption in an application which needs [7] R. Purba, A. Halim and I. Syahputra, "Enkripsi Citra
moderate level of encryption and faster execution time than Digital Menggunakan Arnold’s Cat Map Dan Nonlinear
that of Arnold cat map encryption we can use Henon map Chaotic Algorithm", JSM STMIK Mikroskil, vol. XV, no. 2,
encryption. pp. 61-71, 2014.

[8] E. AVAROĞLU, "Pseudo Random Number Generator Based

6 FUTURE SCOPE
on Arnold's Cat Map and Statistical Analysis", Turkey,
As we are now aware of how to encrypt images, we can 2011.
move forward to encrypt videos i.e., by encrypting each
frame of the video. We can store the data to the cloud directly [9] Devaramapati Amarnath Gupta, Singh and I. Mangal,
that is encrypted. Implementing encryption for cloud data "Image Encryption using Arnold Cat Map and S-Box",
increases security. We can move forward to encrypting using IJARCSSE, vol. IV, no. 8, pp. 807-812, 2014
neural networking techniques which will help us reduce
[10] Chen, Jun-xin & Zhu, Zhi-liang & Fu, Chong & Yu, Hai
time complexity of the code and also increase the encryption
& Zhang, Li-bo. (2015). A fast chaos-based image
rate. Using parallel computing technology to see the
encryption scheme with a dynamic state variables
execution of our algorithm i.e., encryption in one computer selection mechanism. Communications in Nonlinear
and decryption in another computer. Science and Numerical Simulation. 20. 846–860.
10.1016/j.cnsns.2014.06.032.

[11] Somaya Al-Maadeed, Afnan Al-Ali, Turki Abdalla, "A

New Chaos-Based Image-Encryption and Compression
Algorithm", Journal of Electrical and Computer
Engineering, vol. 2012, Article ID 179693, 11 pages, 2012.

IJSER © 2021
http://www.ijser.org
International Journal of Scientific & Engineering Research Volume 12, Issue 6, June-2021 19
ISSN 2229-5518

[12] Shenyong Xiao, YaShuang Deng, "Design and Analysis

of Chaos-Based Image Encryption Algorithm via Switch
Control Mechanism", Security and Communication
Networks, vol. 2020, Article ID 7913061, 12 pages, 2020.

[13] Abhinav Sure, Tanuj Kalgutkar and Supeksha

Waghmare. “Chaotic maps Image Encryption and
Decryption.” (2012).

[14] Chaos Cryptography: Theory, Algorithms and

Applications. Germany: Springer Berlin, 2011.

[15] Gao, Tiegan & Chun, Zenqiang. (2008). Image

encryption based on shuffling algorithm. Chaos,
Solitons & Fractals. 38.213-220.
10.1016/j.chaos.2006.11.009.

IJSER © 2021
http://www.ijser.org