Image Encryption Using 2D Chaotic
Map
WHY?
Introduction
Image Encryption
Image encryption techniques are broadly
classified in to four categories:
• Optical
• Spatial Domain
• Transform Domain
• Compressive Sensing
Chaos a Spatial Domain Based Image
Encryption Techniques
Chaos-Bases systems frame work
Chaotic Map
Some of the maps along use in the field are:
• 2D Rational Maps
• Arnold’s cat map
• Circle map Logistic map
• Exponential map
Image encryption based on a new 2D logistic
adjusted logistic map
Objective of the proposed research
• Design and demonstration of 2D chaos based
pseudorandom number generator
• Proposed Scheme comparison to the existing techniques
and proof of competence
Chaotic maps
1-D chaotic maps
• The famous 1-D chaotic maps used in the literature are logistic and
sine map. In the both maps each iterate t+1 is dependent on the t
instant condition and control parameter µ.
• For the visualization of the behavior of the 1-D chaotic maps two
plots are used:
1. Bifurcation diagram
2. Iteration diagram
Bifurcation diagram
• This diagram shows the distribution of the iterates verse
the control parameter. In case of logistic map, the
iterate i+1 is represented as:
• xi+1=µxi(1-xi)
Iteration diagram
• This diagram is used to show how the iterate varies by
the variation of the iterates for constant value of µ
across initial state xt or xi.
• The diagram is for the µ value of 0.9 (when the system
is in chaos state).
2-D Chaotic map
• In the 2-D chaotic map in spite of one output iterate xi+1
we have only two output iterate (xi+1 ,yi+1).
Reference Map Name Equation
[9] 2D Logistic Map (2D- xi+1 = μ(3yi + 1)xi(1 - xi)
LM) yi+1 = μ(3xi+1 + 1)yi(1 - yi)
μ ∈ (0, 1]
[10]
2D Sine Logistic xi+1 = μ[sin(πyi) + 3]xi(1 - xi)
Modulation Map (2D- yi+1 = μ[sin(πxi+1) + 3]yi(1 - yi)
SLMM) μ ∈ (0, 1]
[11] 2D-LASM xi+1 = sin[πμ(yi + 3)xi(1 - xi)]
yi+1 = sin[πμ(xi+1 + 3)yi(1 - yi)]
μ ∈ (0, 1]
2D Logistic Adjusted Logistic Map (2D-LALM)2-
D map
The Proposed Map
• The proposed map is evolved from the above discussed
referred 2-D map. The equation for the proposed map
are:
• Where μ ∈ (0, 1].
• The proposed model is more chaotic with non-uniform
trajectory plots as compared to the previous.
Lyapunov Stability analysis
• The proposed map is compared with the maps
mentioned above using Lyapunov stability analysis.
Where the analysis provides the intuition about the
sensitivity of the system on the initial condition
• The Lyapunov exponents are determined using the
initial conditions.
Example
•
• Let’s consider we have two initial conditions xt and yt .
• So, the absolute difference D(t) between the two condition is:
• D(t)=
• For most of the systems the D(t) is approximated as:
• D(t)=D02λt
• Where Do is separation between x0 and y0
• And from here the term λ is referred to as Lyapunov exponent so if λ>0 system is sensitive to
initial condition otherwise not. So λ positive means D(t) is exponential growth the initial orbits
will get more separated as compared to be negative they will come closer. [12], [13], [11]
• Larger the value of λ larger is the sensitivity.
• For the proposed scheme and previous schemes, the Lyapunov constants are provided in the
Figure 5.
• It can be seen that in Figure 5 (d) that Lyapunov exponents for the proposed map are positive
for a wide range of µ (0.5 ≤ μ ≤ 0.9). So the system is better than the existing map system.
Result
Proposed Scheme Steps
• The proposed algorithm working steps are presented as
below:
• Step-1: Plain Input Image of size
• Step-2: Adding a randomized border to the image
A border of a definite size and having randomized pixel-
values is added around the original image. This
additional step ensures that, for every run of the encrypt
process (even with the same key), the resulting cipher-
image is significantly different.
• Step-3: Initializing the two 2-D chaotic maps
232 Key bit is used in the proposed model. The key can
be visualizing as a 232-K bits array
• Step-3: Generating the two pseudo-image matrices (S(1),
S(2))
• Using the two pairs of initial states (x0(i), y0(i)) and the
system parameters (μi),two pseudo-image matrices
(S(1), S(2)) are generated.
• Step-4: Encryption/Decryption of the enhanced image
matrix P-dash
• Using each of the two pseudo-image matrices (S(k); k =
1, 2), the following two steps are carried out in proper
sequence to encrypt the enhanced image matrix P- dash
• Shuffling of the pixels
• D = [dij], i = 1, 2, ...M-dash, j = 1, 2, ...N-dash
• dij = (i - 1)N-dash + j
• Let lD be the number of bits needed to represent dij. A new
matrix R(k) is evaluated as,
R(k) = S(k) + D + P –dash
• Consequently, each element of R(k) will have corresponding
elements of the three matrices on the right-hand side
• The elements of R(k) are then shuffled using row-wise
and column-wise sorting as ,
• R(k-dash) = column-sort(row-sort(R(k)))
• The final output of pixel-shuffling operation is obtained
as the image matrix T (k):
T (k) = [R(k-dash) ]AND [28 – 1]
• Hence, each element of T (k) just has an integer represented
by the lowest eight bits of the corresponding elements of
R(k). In other words, the elements of the image matrix T (k)
are the shuffled elements of the image matrix P .
• Xoring and diffusion of the randomization among pixels
• The final output is:
O(k) = oij (k) ; i = 1, 2, ...M-dash, j = 1, 2, ...N -dash
Where
Result discussion
Security analysis
232 bit keys
K1 Results
K2 Results
Differential analysis
References
[
1
N. Masuda and K. Aihara, "Cryptosystems with discretized chaotic maps," Ieee transactions on circuits
] and systems i: fundamental theory and applications, vol. 49, pp. 28-40, 2002.
[
2
G. Gu and J. Ling, "A fast image encryption method by using chaotic 3D cat maps," Optik, vol. 125, pp. 4700-4705,
] 2014.
[
3
X. Huang and G. Ye, "An efficient self-adaptive model for chaotic image encryption algorithm," Communications in
] Nonlinear Science and Numerical Simulation, vol. 19, pp. 4094-4104, 2014.
[
4
A. Ono and T. Kohda, "Solvable three-dimensional rational chaotic map defined by Jacobian elliptic functions,"
] International Journal of Bifurcation and Chaos, vol. 17, pp. 3645-3650, 2007.
[
5
A. Umamageswari and G. R. Suresh, "Security in medical image communication with arnold's cat map method and
] reversible watermarking," in 2013 International Conference on Circuits, Power and Computing Technologies (ICCPCT),
2013.
[ D. Chattopadhyay, M. K. Mandal and D. Nandi, "Symmetric key chaotic image encryption using circle map," Indian
6
] Journal of Science and Technology, vol. 4, pp. 593-599, 2011.
[ N. K. Pareek, V. Patidar and K. K. Sud, "Image encryption using chaotic logistic map," Image and vision computing, vol.
7
] 24, pp. 926-934, 2006.
[ M. C. Shastry, N. Nagaraj and P. G. Vaidya, "The B-exponential map: A generalization of the logistic
8map, and its applications in generating pseudo-random numbers," arXiv preprint cs/0607069, 2006.
]
[ M. Sharma, "Image encryption based on a new 2D logistic adjusted logistic map," Multimedia Tools and Applications,
9vol. 79, pp. 355-374, 2020.
]
[ Y. Wu, J. P. Noonan, G. Yang and H. Jin, "Image encryption using the two-dimensional logistic chaotic map," Journal
1 of Electronic Imaging, vol. 21, p. 013014, 2012.
0
]
[ Z. Hua, Y. Zhou, C.-M. Pun and C. L. P. Chen, "2D Sine Logistic modulation map for image encryption," Information
1 Sciences, vol. 297, pp. 80-94, 2015.
1
]
[ Z. Hua and Y. Zhou, "Image encryption using 2D Logistic-adjusted-Sine map," Information Sciences, vol. 339, pp. 237-
1 253, 2016.
2
]
[ G. Benettin, L. Galgani and J.-M. Strelcyn, "Kolmogorov entropy and numerical experiments," Physical Review A, vol.
1 14, p. 2338, 1976.
3
]
[ R. Brown, P. Bryant and H. D. I. Abarbanel, "Computing the Lyapunov spectrum of a dynamical system from an
1 observed time series," Physical Review A, vol. 43, p. 2787, 1991.
