Received 19 May 2022, accepted 13 June 2022, date of publication 22 June 2022, date of current version 30 June 2022.

Digital Object Identifier 10.1109/ACCESS.2022.3185229

Color Image Encryption Using 2D

Sine-Cosine Coupling Map
ZEZONG ZHANG 1 , JIANENG TANG 1,2 , FENG ZHANG3 , HUI NI3 ,
JINYUAN CHEN2 , AND ZHONGMING HUANG2
1 College of Information Science and Engineering, Huaqiao University, Xiamen 361021, China
2 College of Engineering, Huaqiao University, Quanzhou 362021, China
3 Fujian MM Electronics Company Ltd., Quanzhou 362000, China

Corresponding author: Jianeng Tang ([email protected])

This work was supported in part by the National Natural Science Foundation of China under Grant 61573004, in part by the Foundation of
Quanzhou under Grant 2018C106R, and in part by the Pilot Project of Fujian Province under Grant 2022H0017.

ABSTRACT In this paper, a new two-dimensional sine-cosine coupling chaos map(2D-SCCM) is proposed.
The performance of the proposed two-dimensional chaotic system is analysed by using trajectory distri-
bution maps, Lyapunov exponents, sample entropy and sequence sensitivity, etc. The results show that the
2D-SCCM has better randomness and ergodicity, as well as a wider hyperchaotic range than the existing
partial two-dimensional chaos maps. To verify its application in practice, a 2D-SCCM-based color image
encryption algorithm is proposed. First, the plain image is used to combine with hash function to generate
the key. Then, we construct S-Boxes using the random sequence generated by 2D-SCCM combined with
Arnold map. Finally, a color image encryption algorithm is proposed by using the constructed S-Boxes
and the chaotic map combined with hash function. Experimental simulations and security tests show that
the proposed encryption algorithm has high encryption efficiency and strong security, and can effectively
protect images from various attacks.

INDEX TERMS Image encryption, chaos, color image encryption, S-Box, Arnold map.

I. INTRODUCTION contain image filtering [3], DNA coding [5], frequency

With the rapid development of modern network technol- domain transform [6], elliptic curves [7], compressive sens-
ogy, the use of images for information transmission has ing [8], S-Box [9], fractional Fourier transform [10], neu-
become more and more frequent. The security of images ral networks [11], chaotic systems [12]–[14], etc. However,
has received extensive attention from researchers because of among these technologies, because of the special nature of
the widespread use of images in the Internet. In order to chaotic systems, cryptography based on chaotic systems has
protect the information of images, many schemes have been become one of the most popular cryptographic schemes.
proposed by researchers, including data hiding [1], water- There are many unique characteristics of chaotic systems,
marking [2], and image encryption [3]–[5]. Among all the including high sensitivity to initial conditions and parame-
schemes, image encryption is one of the most straightforward ters, internal randomness and ergodicity, which make them
and effective schemes [4]. suitable for image encryption. Since J. Fridrich first applied
Due to the large redundancy and strong correlation chaotic system to image encryption [14], the image encryp-
between image pixels, image encryption is different from tion schemes based on chaotic systems have been developed
text encryption [5]. Traditional data encryption schemes rapidly, and various chaos-based image encryption schemes
mainly include Advanced Encryption Standard (AES) and have been proposed. For example, Wu et al. proposed an
Data Encryption Standard (DES). When they are applied image encryption algorithm based on 2D-HSM and DNA
to image encryption, it takes a lot of time resulting in low coding techniques, where they performed a diffusion oper-
encryption efficiency [4]. Therefore, researchers have pro- ation on the image by DNA and scrambled the image by
posed many different image encryption techniques, which chaotic sequences [5]. Mansouri et al. proposed an image
encryption algorithm based on a one-dimensional chaotic
The associate editor coordinating the review of this manuscript and system, where they added chaotic sequences to the original
approving it for publication was Jun Wang . image during the diffusion process to improve the security of

This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
VOLUME 10, 2022 67669
 Z. Zhang et al.: Color Image Encryption Using 2D Sine-Cosine Coupling Map

the algorithm [15]. J. Arif et al. proposed an image encryption II. 2D SINE-COSINE COUPLING MAP
algorithm based on Logistic map and S-Box [9]. They used This section first reviews two traditional one-dimensional
S-Box to perform pixel substitution in the pixel diffusion discrete chaotic maps, and then introduces the mathematical
stage to enhance the security of the algorithm, but this algo- model of the two-dimensional sine-cosine coupling chaotic
rithm can only be used for grayscale images. Similarly, there map proposed in this paper.
are many other image encryption schemes based on chaotic
systems that have been proposed by researchers [16]–[19]. A. LOGISTIC MAP
Although many image encryption algorithms based on Logistic map is a one-dimensional discrete chaotic map with
chaotic systems have been proposed, some of them have the mathematical definition of Eq(1).
been proven to be insecure [5]. Because the one-dimensional
chaotic map has the characteristics of small key space and xi+1 = 4µxi (1 − xi ) (1)
simple chaotic orbit, the attacker can easily deduce the ini-
tial conditions and chaotic orbit of the system [12]. While where its control parameter µ ∈ (0, 1]. The system is in a
high-dimensional chaotic systems have relatively complex chaotic state when the parameter µ ∈ [0.89, 1].
chaotic behavior and larger key space, the computational
complexity and space complexity of high-dimensional chaos B. SINE MAP
are much larger than those of one-dimensional chaotic sys- Sine map is also a one-dimensional discrete chaotic map with
tems. Thus, it is relatively difficult to implement in practical the mathematical definition of Eq(2), which is derived from
hardware, and it takes a long time. Therefore, it becomes par- the sine function.
ticularly important to propose a chaotic system with complex
chaotic behavior and low computational complexity [19]. xi+1 = ρ sin(πxi ) (2)
In this paper, a new two-dimensional discrete chaotic map
is proposed, called 2D-SCCM. On the basis of studying the where its control parameter ρ ∈ (0, 1]. The system is in a
classical one-dimensional discrete chaotic map, a new two- chaotic state when the parameter ρ ∈ [0.87, 1].
dimensional discrete chaotic map is constructed by coupling
the output variables of multiple one-dimensional discrete C. 2D-SCCM
systems. At the same time, because the cosine function is The traditional one-dimensional discrete chaotic map has
introduced in the system, the output of the entire chaotic sys- many disadvantages such as narrow chaotic interval and sim-
tem is in the range of [−1,1]. To verify the chaotic properties ple system structure. To solve these problems, a new two-
of the proposed system, first the system is analyzed by using dimensional sine-cosine coupling chaotic system is proposed
trajectory distribution diagrams, Lyapunov exponents(LE), with the mathematical definition of Eq(3).
sample entropy, and sequence sensitivity. Then the chaotic (
properties are compared with those of other 2D chaotic maps. xi+1 = sin(4απyi (1 − xi ))
(3)
Finally, the randomness of the system-generated sequences yi+1 = cos(8βπ sin(π + yi ) + xi+1 )
were tested by using the National Institute of Standards and
Technology (NIST) statistical test tool. The results show that where α, β are both system control parameters with α 6 = 0,
the proposed two-dimensional chaotic map has more complex β 6= 0. It can be seen from the mathematical definition of
chaotic behavior and wider hyperchaotic interval. Based on Eq(3) that the 2D-SCCM is mainly derived from Logistic
the proposed chaotic map, we further propose a color image and Sine maps. Its output signals are finally modulated with
encryption algorithm. The algorithm can be divided into two the nonlinear cosine function, which further improves the
major parts: scrambling and diffusion. In the image scram- randomness of the output signals. Meanwhile, because of the
bling stage, the chaotic sequences are used to perform overall boundedness of the cosine function, the final output signals x
scrambling. At the same time, pixels are replaced one by one and y of the system are in the range of [−1,1].
by introducing S-Boxes. In the diffusion stage, XOR and the
modular diffusion are used for pixel diffusion, so that the III. BEHAVIOR ANALYSIS
three components of the color image interact with each other In this section, the dynamical behavior of the proposed
to achieve better encryption performance. The experimental chaotic system is analyzed by chaos trajectory, Lyapunov
simulation results and security analysis show that the pro- exponent, sample entropy, and sequence sensitivity. Mean-
posed algorithm can encrypt color images of arbitrary size while, the randomness of the system output sequences are
and has strong resistance to attacks. tested by using the NIST statistical test tool. In addition,
This paper is organized as follows. Section II introduces we compare the relevant properties with three existing 2D
the mathematical model of 2D-SCCM. Section III analyzes discrete chaotic systems, including 2D Sine Logistic modula-
dynamic behavior of the 2D-SCCM. Section IV details the tion map (2D-SLMM) [19], 2D Logistic-Sine-coupling map
proposed color image encryption algorithm. Section V intro- (2D-LSCM) [12], 2D hyperchaotic map (2D-HM) [17].
duces the simulation and safety analysis. And section VI is a The 2D-SLMM has two parameters γ and η. When the
brief summary about the paper. control parameter η = 3, it has a better chaotic performance.

67670 VOLUME 10, 2022

Z. Zhang et al.: Color Image Encryption Using 2D Sine-Cosine Coupling Map

FIGURE 1. Trajectories of four 2D chaotic maps: (a) the 2D-SLMM with parameter γ = 1; (b) the 2D-LSCM with parameter θ = 0.99;
(c) the 2D-HM with parameter κ = 2 and λ = 1; (d) the 2D-SCCM with parameter α = 4 and β = 4.

Thus, it is simply defined as follow [19]: (c) 2D-HM, (d) 2D-SCCM. When plotting these trajectory
( distributions, the parameters are set that can make all corre-
xi+1 = γ (sin(πyi ) + 3)xi (1 − xi )
(4) sponding chaotic maps attain a completely chaotic state [12].
yi+1 = γ (sin(πxi+1 ) + 3)yi (1 − yi ) At the same time, the first 5000 terms of the system output
The 2D-LSCM only has one parameter θ and the definition are discarded in order to make the system reach a stable
is written as follow [12]: chaotic state. It can be seen from the Fig.1 that the trajectory
( distribution diagram of 2D-SLMM only occupies a small part
xi+1 = sin(π(4θxi (1 − xi ) + (1 − θ) sin(πyi ))) of the phase plane, and the trajectory distribution diagram
(5)
yi+1 = sin(π(4θyi (1 − yi ) + (1 − θ) sin(πxi+1 ))) of 2D-LSCM does not completely occupy the entire phase
plane, while the trajectory distribution diagrams of 2D-HM
The 2D-HM has two parameters κ and λ. its mathematical
and 2D-SCCM completely occupies the entire phase plane.
expression is defined as follow [17]:
In addition, the trajectory distribution of 2D-SCCM is more
 xi+1 = sin( κ )

uniform than that of 2D-HM, indicating that the sequences
sin(yi ) (6) generated by 2D-SCCM are more random than the other three
y
i+1 = λ sin(π(x i + yi )) chaotic systems.
A. CHAOS TRAJECTORY
The trajectory distribution diagram of a chaotic system B. LYAPUNOV EXPONENT
describes the variation of the output of the system over time. The initial state sensitivity is one of the main features to
Systems with complex chaotic behavior usually have com- determine whether a dynamic system exists chaotic behavior.
plex trajectory distribution maps. And it also occupies a larger A good chaotic system should gradually output two distinct
phase space [17]. sequences of iterations after inputting two initial iterations
Fig.1 shows the trajectory distributions of four different with small differences. We usually use the Lyapunov Expo-
2D chaotic systems, including (a) 2D-SLMM, (b) 2D-LSCM, nent (LE) [12] to characterize the sensitivity of a chaotic

VOLUME 10, 2022 67671

 Z. Zhang et al.: Color Image Encryption Using 2D Sine-Cosine Coupling Map

system. Assuming that a one-dimensional discrete chaotic of the system and the more complex the chaotic behavior
system is defined by xi+1 = F(xi ), where F(x) is piecewise of the system [22]. Fig.4 shows the comparison of SE tests
differentiable [12], the LE of the system is defined as for Sine, Logistic and four 2D chaotic systems under different
control parameter conditions. It can be seen that the SE val-
n−1
1 X  0  ues of the proposed 2D-SCCM system are larger than those
LE = lim ln F (xi ) (7)
n→∞ n of the other five chaotic systems, indicating that the proposed
i=0
chaotic system has more complex chaotic characteristics.
LE is a numerical feature that describes the average sepa-
ration rate and divergence degree of adjacent phase space D. SEQUENCE SENSITIVITY ANALYSIS
trajectories [12]. A positive LE means that the system A good chaotic system should be extremely sensitive to
exist chaotic behavior. And the larger the value of LE is, changes in the initial values. In other words, if the system’s
the more complex behavior the chaotic system has. For inputs are slightly different, its output sequence should be
high-dimensional chaotic systems, there are generally sev- significantly different. In order to test the sensitivity of the
eral different LE values in the system. When the system system, we input initial values with a difference of 10−15 , and
has two or more positive LE values, we call this system the changes of the output sequences are shown in Fig.5. They
a hyperchaotic system. Compared with the general chaotic can be seen that the system’s outputs are significantly differ-
system, the hyperchaotic system has more complex chaotic ent after 14 iterations, indicating that our proposed system is
behavior. extremely sensitive to the initial values.
A two-dimensional chaotic system usually has two LE
values, and Fig.2 shows the LE of different two-dimensional E. NIST
chaotic systems under their respective control parameters.
To further investigate the randomness of the system output
When the parameter range is set to [0,1], the LE values
sequences, we used the National Institute of Standards and
of each system behave as follows: the 2D-SLMM shows
Technology (NIST) SP800-22 test tool [24]. The NIST sta-
hyperchaotic behavior when γ ∈ [0.85, 1], the 2D-LSCM
tistical test tool includes 15 statistical tests. It is used for
shows hyperchaotic behavior when θ ∈ [0.57, 1], the
evaluating the randomness of the sequences generated by
2D-HM shows hyperchaotic behavior when λ ∈ [0.2, 0.34]
2D-SCCM [34]. After the NIST statistical test tool finishes
or λ ∈ [0.36, 0.58], and the 2D-SCCM shows hyperchaotic
testing, each test item will generate a corresponding test
behavior when α ∈ [0.28, 1] or β ∈ [0.15, 1]. The
value P ∈ [0, 1]. And when the corresponding generated
result shows that the 2D-SCCM has a wider range of hyper-
P-value is greater than 0.01, it indicates passing the test item.
chaotic intervals compared with the other three chaotic maps.
When all test items pass, it indicates that the sequence has
In order to better compare the magnitude of LE, we com-
strong randomness, while a larger P-value indicates stronger
pare the two LE values with the four chaotic systems under
randomness [12]. The computational statistics of the NIST
certain parameters in Fig.3. The results show that the LE
test are shown in Table 1. From the results, the sequences gen-
values of 2D-SCCM are larger than the other three exist-
erated by the system can pass all tests of 15 items, indicating
ing chaotic systems in a large interval, which further indi-
that they have strong pseudo-randomness.
cates that the proposed system have more complex chaotic
behavior.
TABLE 1. P-value of NIST statistical test.
C. SAMPLE ENTROPY
Sample Entropy (SE) is a new evaluation method for test-
ing the complexity of time series based on approximate
entropy [22], which is used to describe the output simi-
larity of dynamic systems. The SE of a given time series
{z1 , z2 , · · · , zN } of m dimension is defined as [23]
K1
SE(m, r, N ) = − log (8)
K2
where m is the reconstruction dimension and r is the calcu-
lated threshold, which we usually set m = 2, r = 0.2 × std
(the std is the standard deviation of the tested time series).
The K1 represents the number of vectors whose Chebyshev
distance is less than or equal to r between ZmC1 (i) and
ZmC1 (j), and the K2 represents the number of vectors whose
Chebyshev distance is less than or equal to r between Zm (i)
and Zm (j), where Zm (i) = {zi , zi+1 , · · · , zi+m−1 }. The larger
the calculated SE value, the lower the level of regularity

67672 VOLUME 10, 2022

Z. Zhang et al.: Color Image Encryption Using 2D Sine-Cosine Coupling Map

FIGURE 2. The two LEs of four 2D chaotic maps: (a) the 2D-SLMM; (b) the 2D-LSCM; (c) the 2D-HM with fixed parameter λ = 1;
(d) the 2D-HM with fixed parameter κ = 2; (e) the 2D-SCCM with fixed parameter α = 4; (f) the 2D-SCCM with fixed parameter
β = 4.

IV. ENCRYPTION ALGORITHM used to diffuse the three color components of the original
This section introduces a color image encryption algorithm image to change the pixel values. Then, the sequences are
based on 2D-SCCM combined with S-Box scrambling, and used to generate S-Boxes combined with Arnold mapping
the overall encryption structure is shown in Fig.6. The hash to confuse the diffused image, and the confusing method
value is generated by the hash function and plain image, and can change the position and value of the image pixels at the
the key is updated by the hash value to generate the finally same time. Finally, combined with the hash value and chaotic
key. First, the chaotic sequences generated by 2D-SCCM are sequences, the scrambled image is diffused again to improve

VOLUME 10, 2022 67673

 Z. Zhang et al.: Color Image Encryption Using 2D Sine-Cosine Coupling Map

FIGURE 3. The comparison of LE: (a) the comparison of 2D-HM(λ), 2D-SLMM(γ ) and 2D-SCCM(α) for LE1; (b) the
comparison of 2D-HM(κ), 2D-SLMM(γ ) and 2D-SCCM(α) for LE2; (c) the comparison of 2D-HM(λ), 2D-LSCM(θ) and
2D-SCCM(β) for LE1; (d) the comparison of 2D-HM(λ), 2D-LSCM(θ) and 2D-SCCM(β) for LE2.

FIGURE 4. The SE comparison of different chaotic systems: (a) the Logistic, Sine, 2D-LSCM, 2D-SLMM(η = 3), 2D-HM(λ = 1)
and 2D-SCCM(β=4); (b) 2D-HM(κ = 1) and 2D-SCCM(α = 4).

the security performance of image encryption. The specific obtained from the plain image is used to update the initial key,
encryption steps are described as follows. which reduces the possibility of the system being attacked
by the chosen plaintext. Here we choose the SHA-256 func-
A. KEY GENERATION tion, the generated hash value is divide into 64 blocks, and
A separate system key increases the possibility of the system converted each block to a decimal number, denoted as k1 ,
being attacked to a chosen plaintext attack. The hash value k2 ,· · · ,k64 . Then, the initial key is set to K = {x0 , y0 , a0 , b0 },

67674 VOLUME 10, 2022

Z. Zhang et al.: Color Image Encryption Using 2D Sine-Cosine Coupling Map

FIGURE 5. Sequence sensitivity analysis: (a)x (b)y.

FIGURE 6. Block diagram of the overall encryption algorithm.

and the set initial key is updated by using Eq(9). iter is the number of iterations of the Arnold map. The SQ
is the number of iteration terms discarded in the encryption
(k1 + k2 + · · · + k16 )

x =( ) mod 1 + x0 process to bring the chaotic system to full chaos. The CZ is the
 1



 (k17 + k18 + · · · + k32 ) preset value used in taking the modal diffusion. The floor(Q)
 α = ((k1 + k2 + · · · + k32 )/64) mod 1 + a0



 means taking the largest integer not larger than Q. The mean
 β = ((k32 + k33 + · · · + k64 )/16) mod 1 + b0


 represents taking the average, and the mod means the modulo
iter = floor(mean(k1 , k2 , · · · , k64 )) + 40 (9) operation.

SQ = (k1 + k2 + · · · + k64 ) × 2



B. XOR-DIFFUSION



 CZ = (k1 + k2 + · · · + k64 ) mod 256
Step 1: Input a color plain image P with a size of M ×N , and

 y = ( mean(k33 , k34 , · · · , k48 ) ) mod 1 + y



1 0 decompose it into three components of R, G and B with a size
mean(k49 , k50 , · · · , k64 )

of M × N , which are denoted as Pr, Pg and Pb respectively.
where {x1 , y1 } is the updated initial condition and {α, β} is Step 2: Substitute {x1 , y1 , α, β} into 2D-SCCM, and iterate
the updated control parameters of the chaotic system. The the system n = SQ+M ×N times. Then the first SQ items are

VOLUME 10, 2022 67675

 Z. Zhang et al.: Color Image Encryption Using 2D Sine-Cosine Coupling Map

(
discard. Finally we can get the sequence X1 and Y1 according a1 = (floor(abs(Y 2(iter) × 1012 ))) mod 20 + 1
to Eq(10). b1 = (floor(abs(Y 2(iter + 1) × 1012 ))) mod 20 + 1
(
X1 = {xSQ+1 , xSQ+2 , · · · , xSQ+MN } (15)
(10)
(
Y 1 = {ySQ+1 , ySQ+2 , · · · , ySQ+MN } a2 = (floor(abs(Y 2(iter + 2) × 1012 ))) mod 20 + 1
0 0 b2 = (floor(abs(Y 2(iter + 3) × 1012 ))) mod 20 + 1
Step 3: Let
 y0 1 = , x1 = ySQ+MN , then substitute
xSQ+MN
the updated x1 , y1 , α, β into the chaotic system. Iterate the
0 (16)
system n = SQ + M × N times with discarding the first SQ
Step 2: Scramble S-Box1 according to the initial parame-
items. Finally we can get the sequence X2 and Y2 according
ters obtained in the Step1. Two new S-Boxes are generated,
to Eq(11).
( namely S-Box2 and S-Box3.
X2 = {x 0 SQ+1 , x 0 SQ+2 , · · · , x 0 SQ+MN } (
(11) S-Box2 = Arnold(S-Box1, iter, a1, b1)
Y 2 = {y0 SQ+1 , y0 SQ+2 , · · · , y0 SQ+MN } (17)
S-Box3 = Arnold(S-Box2, iter, a2, b2)
Step 4: Convert the obtained sequences X1 into matrices
A1 with a size of M × N according to Eq(12). Similarly, Step 3: Sort the chaotic sequences X1, X2, Y1, Y2 to get
we can get B1, A2. At this time, the element values of A1, the index sequences SC1, SC2, SC3, SC4.
B1, and A2 are all in the range of [0, 255], where the abs 
represents taking the absolute value. 
 [−, SC1] = sort(X1)

 [−, SC2] = sort(X2)
X11(i) = (abs(floor(X1(i) × 1012 ))) mod 256

 (18)
 [−, SC3] = sort(Y 1)
A1 = reshape(X11, [M , N ])

 


 


 12 [−, SC4] = sort(Y 2)
 Y 11(i) = (abs(floor(Y 1(i) × 10 ))) mod 256


B1 = reshape(Y 11, [M , N ]) (12) Step 4: Perform bit plane decomposition on the encrypted
12



 X22(i) = (abs(floor(X2(i) × 10 ))) mod 256 image components obtained by the diffusion in the subsection



 A2 = reshape(X22, [M , N ]) B. Then recombine the upper four bits and the lower four bits


 respectively. Finally obtain the decomposed image compo-
i = 1, 2 · · · MN
nents PR1, PR2, PG1, PG2, PB1, PB2.
Step 5: The generated A1, A2 and B1 are used to perform 
bitwise XOR with the three image components of the color 
 PR1 = PR mod 16 + 1

image. The image components PR, PG and PB are obtained PR2 = floor(PR/16) + 1




respectively after the first diffusion encryption.

 PG1 = PG mod 16 + 1
 (19)
 PR(i, j) = Pr(i, j) ⊕ A1(i, j)
 PG2 = floor(PG/16) + 1
i = 1, 2, · · · , M
 


PG(i, j) = Pg(i, j) ⊕ A2(i, j) (13)
 PB1 = PB mod 16 + 1
= 1, 2, · · · , N

j


PB2 = floor(PB/16) + 1
 
PB(i, j) = Pb(i, j) ⊕ B1(i, j)


C. CONFUSION Step 5: Use the chaotic sequence Y2 to generate the param-

eter H for controlling the scrambling combination according
Generally, each pixel of natural image has strong correlation,
to Eq(20), in which Fig.7 shows the six ways of the set
and a good encryption algorithm is to eliminate the correla-
scrambling combination. The final scrambling image compo-
tion between elements [15]. Since we only change the value
nents WR, WG, WB are obtained according to the selected
of each pixel in the diffusion stage without modifying the
scrambling combination using Eq(21). In this paper, Eq(21)
position of the pixel, the correlation between the original pix-
represents scrambling combination , 1 and other combina-
els cannot be changed. In this section, we use a combination
tions are similarly represented.
of chaotic sequences and S-Boxes to perform the scrambling
operation on the diffused image, and the specific scrambling H = (floor(abs(Y 2(iter) × 1012 ))) mod 6 + 1 (20)
steps are as follows.
Step 1: Using the chaotic sequences X1, Y1, X2, Y2 from
the subsection B in section IV, generate S-Box1 according
to Eq(14). And we can obtain the initial parameters of the
Arnold map according to Eq(15) and Eq(16), respectively.

 [−, S] = sort(X1(1 : 256) + Y 1(1 : 256)

+ X2(1 : 256) + Y 2(1 : 256))

S-Box1 = reshape(S, [16, 16]) − ones(16, 16)

(14) FIGURE 7. The scrambled combination.

67676 VOLUME 10, 2022

Z. Zhang et al.: Color Image Encryption Using 2D Sine-Cosine Coupling Map

1
FIGURE 8. A scrambling example of a 4 × 4 size image, with the PR component of the color image taking the scrambling of combination .



 WR(k) = S-Box1(PR1(SC4(k)), PR2(SC1(k))) Step 3: The obtained encrypted components CR, CG and
WG(k) = S-Box2(PG1(SC4(k)), PG2(SC2(k)))

CB are recombined to form the final color encrypted image C.
 WB(k) = S-Box3(PB1(SC4(k)), PB2(SC3(k)))
 The decryption algorithm is the inverse of the encryption
k = 1, 2, · · · , MN

algorithm, which is not described in detail here.
(21)
V. EXPERIMENTAL SIMULATION AND SAFETY ANALYSIS
In order to better describe the whole process of scrambling, In this section, the simulation test of the proposed algorithm is
we present a simple 4 × 4 image scrambling example in carried out, and the encryption performance of the algorithm
Fig.8, which takes the WR generation process in scrambling is analyzed at the same time. Our selected test images include
combination 1 as an example. The methods of other compo- 512*512*3 Lena, all black image, all white image and some
nents and scrambling combinations are similar and will not color images from USC-SIPI [4]. Simulations are carried out
be described in detail here. on Windows10 environment using MATLAB 2020b [34]. The
processor of the computer used is Intel(R) Core(TM) i5-7200
D. MODULAR DIFFUSION CPU @ 2.50 GHz, and the running memory of the computer
To further improve the security of the cryptosystem, we intro- is 12 GB.
duce the sequential mod-taking diffusion of elements after the
XOR diffusion and the S-Box-based permutation. In the mod- A. EXPERIMENTAL SIMULATION
taking diffusion, each newly generated element is related to The essence of the image encryption algorithm is to transform
the previous element, and the elements of each color com- the natural image to be transmitted into an unrecognizable
ponent are also related to the other components. Therefore, noise-like image. Only the recipient can recover information
the mod-taking diffusion can effectively improve the security of the original image by applying the correct key and the
of the encryption algorithm, and the detailed steps are as cipher image. Others cannot recover any usable information
follows. of the original image without the correct key even if they get
Step 1: Use the chaotic sequence Y2 from the subsection the cipher image.
B to generate the required sequence B2 in the mod-taking The simulation test results are shown Fig.9. It can be seen
diffusion according to Eq(22). from Fig.9(a) that the original image has rich information.
And from Fig.9(c), it can be seen that the encrypted image
B2 = (abs(floor(Y 2 × 1012 ))) mod 256 (22)
has no valid information. When the key is used, the decrypted
Step 2: Perform modular diffusion operation for WR, WG image consistent with the original image can be obtained
and WB according to Eq(23). as shown in Fig.9(e). The simulation results show that our
 proposed algorithm can convert a rich color image into an

 CR(1) = (WR(1) + B2(1) + CZ ) mod 256 unrecognizable ciphertext image.
CR(i) = (CR(i − 1) + B2(i) + WR(i)) mod 256




 CG(1) = (WG(1) + B2(1) + CR(MN )) mod 256 B. KEY ANALYSIS


CG(i) = (CG(i − 1) + B2(i) + WG(i)) mod 256 (23) 1) KEY SPACE ANALYSIS
CB(1) (WB(1) B2(1) CG(MN )) 256



 = + + mod An ideal image encryption algorithm should have as large a



 CB(i) = (CB(i − 1) + B2(i) + WB(i)) mod 256 key space as possible so that it can resist possible brute-force
i = 2, 3, · · · , MN

exhaustive attacks. Reference [25] gives a standard for an

VOLUME 10, 2022 67677

 Z. Zhang et al.: Color Image Encryption Using 2D Sine-Cosine Coupling Map

FIGURE 9. Simulation test of encryption algorithm: (a) plain images; (b) histogram of the plain images; (c) cipher Images;
(d) histogram of the cipher images; (e) decrypted images.

ideal algorithm key space, which should be larger than 2100 . Key3, Key4, and Key5. Here we use Lena as a test image.
The key of the encryption algorithm proposed in this paper is Hash=’6eca336ff63320b496048f0bc37edd1a8a5d764eefa
divided into two parts: (1) initial conditions {x0 , y0 } and con- 85e54862984db7f301258’;
trol parameters {a0 , b0 }; (2) the number of chaotic sequences Hash+1= ’6eca336ff63320b496048f0bc37edd1a8a5d764e
discards SQ. In this paper, we assume that the computational efa85e54862984db7f301259’;
precision of the computer is 10−15 . Thus, the key space of Key = {H ash, x0 , y0 , a0 , b0 },
this paper is 1060 × 1921 ≈ 2209 , which is much larger than Key1 = {H ash + 1, x0 , y0 , a0 , b0 },
the key space standard mentioned in the reference [25]. Key2 = {H ash, x0 + 10−15 , y0 , a0 , b0 },
Key3 = {H ash, x0 , y0 + 10−15 , a0 , b0 },
2) KEY SENSITIVITY ANALYSIS Key4 = {H ash, x0 , y0 , a0 + 10−15 , b0 },
For a superior encryption algorithm, the key should not Key5 = {H ash, x0 , y0 , a0 , b0 + 10−15 }.
only have a large key space but also be extremely sensi- In order to test the encryption sensitivity of the key, first
tive. In order to test the key sensitivity of our proposed we encrypt the original image with Key2 to get the cipher
algorithm, we encrypt the image with a key to obtain the image Fig.10(c), then we subtract it with the cipher image
ciphertext image Fig.10(b), and then decrypt to obtain the obtained with Key to get Fig.10(d). The result shows that
decrypted image Fig.10(e). Because the Key is divided into the two cipher images have a significant difference. And
two parts, where the SQ is determined by the hash value. the mathematical analysis of the two cipher images shows
Therefore, we first change Hash to Hash+1 to test the that there is 99.6105% difference between the two cipher
sensitivity of the SQ in the key. Then, for another part, images, indicating that the key is extremely sensitive in the
we add a tiny perturbation of 10−15 to each parameter of encryption algorithm. Meanwhile, in order to test the decryp-
the key respectively to form four new keys, named Key2, tion sensitivity of the key, we decrypt the ciphertext image

67678 VOLUME 10, 2022

Z. Zhang et al.: Color Image Encryption Using 2D Sine-Cosine Coupling Map

FIGURE 10. Key sensitivity analyses: (a) plain image; (b) the cipher image with Key; (c) the cipher image with Key2;
(d) |(b)-(c)|; (e)the decrypted image with Key to (b); (f) the decrypted image with Key1 to (b); (g) the decrypted image
with Key2 to (b); (h) the decrypted image with Key3 to (b); (i) the decrypted image with Key4 to (b); (j) the decrypted
image with Key5 to (b).

TABLE 2. The calculation results of χ 2 .

Fig.10(b) using the different keys containing Key1, Key2, where fk is the frequency of occurrence of pixel value k,
Key3, Key4 and Key5 respectively. The results show that fe = MN /256, and M, N are the width and height of the image
only the correct key can recover the original image, and other pixels respectively. When the test degree is 0.05, the stan-
slightly altered key cannot recover the image information. dard value χ0.052 = 293.2478 can be obtained. When the
It shows that the key has extremely strong decryption sen- calculated χ result is less than the standard value, it means
2

sitivity. In summary, the proposed algorithm key not only has that the chi-square test passes [26]. The calculation results of
a large key space, but also has an extremely strong sensitivity. χ 2 for each component of different color images are shown in
Table 2. The results show that the χ 2 of all encrypted image
C. STATISTICAL ANALYSIS components is smaller than the standard value, which means
1) HISTOGRAM ANALYSIS that the pixel value distribution of the encrypted image is
The histogram can well show the distribution of pixels in uniform.
an image. Generally, the histogram of natural images has
significant pixel distribution characteristics. An hacker can be 2) CORRELATION ANALYSIS
able to get related information of the original image by ana- A good image encryption algorithm must be able to break the
lyzing the histogram information. Therefore, the pixels of the correlation about the original images, as there is generally a
encrypted ciphertext image should be as evenly distributed strong correlation between pixels in natural images. The cor-
as possible. The histogram of the original image is shown in relation between image pixels is mainly manifested in three
Fig.9(b). It can be seen that the histogram distribution of the directions including horizontal, vertical and diagonal. In this
original image has obvious variability. The histogram of the paper, we use the method of literature [21] to measure the
ciphertext image is shown in Fig.9(d), and the results show magnitude of correlation by calculating the autocorrelation
that the histogram distribution of the encrypted image is uni- coefficient (AC), which is defined as
form without obvious features. At the same time, in order to
quantitatively analyze the distribution of image pixels, we use E [(Xt − E[Xt ])] (Xt+1 − E[Xt ])
AC = (25)
the chi-square test (χ 2 ) mentioned in the literature [26] to E (Xt − E[Xt ])2
 
analyze the images, which is defined as
255 where X t and X t+1 are two adjacent pixel sequences, and
X (fk − fe )2
χ =2
(24) E[•] denotes the mathematical expectation. The calculated
fe value of the autocorrelation coefficient will fall into the range
k=0

VOLUME 10, 2022 67679

 Z. Zhang et al.: Color Image Encryption Using 2D Sine-Cosine Coupling Map

TABLE 3. Correlation statistics of plain image and cipher image.

TABLE 4. Comparison of correlation coefficients under different encryption algorithms (Lena).

of [−1,1]. When the absolute value of AC is smaller, it indi- average uncertainty of all pixel value [34]. For an
cates that the correlation between pixels is smaller. image whose pixel value is in the range of [0, 255],
The statistical results of the correlation coefficients of each the calculation expression of the information entropy(IE)
image component are shown in Table 3. The results show that is as
the correlation coefficients about the encrypted images tend 255
X
to be close to 0, which indicates that the proposed encryption IE = − f (xi )log2 f (xi ) (26)
algorithm can effectively reduce the correlation between pix- i=0
els. Meanwhile, the comparison data of correlation under dif-
where f (xi ) denotes the frequency of occurrence of pixel
ferent encryption algorithms are shown in Table 4. The results
value xi . From Eq(26), it can be concluded that IE can obtain
show that the proposed algorithm is more destructive to the
the theoretical maximum value of 8 when each pixel value
original image correlation compared to other encryption algo-
has the same probability of occurrence. And the larger the
rithms. In order to further visualize the change of correlation
calculated value is, the more uniform the pixel distribution
before and after encryption, the correlation statistics of Lena
is. The calculation results of IE values for different sizes
as an example are shown in Fig.11. The figures show that
of images are shown in Table 5. The results show that all
there are an obvious linear correlation between adjacent pix-
plain images have relatively small information entropy. How-
els of the original image, while the pixels of the encrypted
ever, the IE values of encrypted images are all close to 8,
image are uniformly distributed and almost no correlation.
indicating that the encrypted image pixels are distributed
In summary, the proposed encryption algorithm can break the
uniformly [18]. In addition, Table 6 shows the compari-
correlation between adjacent pixels of the original image.
son results of the IE value for different encryption algo-
rithms. The results show that the information entropy of
D. INFORMATION ENTROPY the proposed algorithm is closer to the theoretical value 8,
The information entropy is used to describe the uncer- which indicates that the proposed algorithm has stronger
tainty of the signal, and it can also be used to reflect security.

67680 VOLUME 10, 2022

Z. Zhang et al.: Color Image Encryption Using 2D Sine-Cosine Coupling Map

FIGURE 11. Visualize the correlation between pixels(Lena): (a) the correlation of the R component of the plain image in each direction;
(b) the correlation of the G component of the plain image in each direction; (c) the correlation of the B component of the plain image in
each direction; (d) the correlation of the R component of the ciphertext image in each direction; (e)the correlation of the G component of
the ciphertext image in each direction; (f) the correlation of the B component of the ciphertext image in each direction.

TABLE 5. Information entropy of images.

TABLE 6. Comparison of information entropy of different encryption algorithms (Lena).

E. DIFFERENTIAL ATTACK ANALYSIS Finally, the purpose of decryption is achieved in this way [4].
Differential attack is a common cryptanalytic attack. Firstly, Therefore, a good encryption algorithm should have excellent
two plain images with minor differences are encrypted sepa- resistance to differential attacks.
rately by an attacker. Then the attacker establishes an intrinsic In order to better analyze the ability of encryption algo-
connection between the ciphertext and the plain image by rithms to resist differential attacks, the Number of Pixels
comparing the differences between the encrypted images. Change Rate (NPCR) and the Unified Average Changing

VOLUME 10, 2022 67681

 Z. Zhang et al.: Color Image Encryption Using 2D Sine-Cosine Coupling Map

TABLE 7. Test passing criteria for NPCR and UACI.

TABLE 8. Calculation results of NPCR and UACI.

TABLE 9. Comparison of NPCR and UACI for different encryption algorithms (Lena).

Intensity (UACI) are often used as evaluation indicators. They as an example is shown in Table 9. And the results show that
are defined as the proposed encryption algorithm is closer to the theoreti-
 M X N cal value, so our algorithm is more resistant to differential
X W (i, j)
, attacks.


 NPCR(C 1 C 2 ) = × 100%
MN


( i=1 j=1 (27)

 0 if C1 (i, j) = C2 (i, j) F. NOISE ATTACK ANALYSIS
 W (i, j) =


1 if C1 (i, j) 6 = C2 (i, j) Because all transmission channels exist noise, images are sus-
ceptible to noise contamination during transmission. A good
N
M X
X |C1 (i, j) − C2 (i, j)| encryption algorithm should be able to recover as much of the
UACI (C1 , C2 ) = × 100%
MN × 255 plain image information as possible in a noise-contaminated
i=1 j=1
ciphertext image. In order to test the ability of our proposed
(28) algorithm to resist noise attack, we first add different degrees
where C1 and C2 denote two different encrypted images. And of noise to the ciphertext image, then use the correct key
they are generated by encrypting two original images whose to decrypt the ciphertext image, finally observe the recovery
pixels differ by only one bit. MN is the total number of image degree of the image under different noise pollution. The salt
pixels. The literature [32] gives the test passing criteria in and pepper noise with density of 1%, 5%, 10%, 20% is added
Table 7, where the ideal expectation values of NPCR and to the cipher image respectively, and the decryption results
UACI are 99.6094% and 33.4635% respectively [32]. When are shown in Fig.12. The results show that even if the added
the calculated values of both NPCR and UACI pass the test noise density reaches 20%, the proposed algorithm can still
standard, the closer the calculation result is to the theoret- recover a large amount of image information, indicating that
ical value, the better the encryption effect and the stronger our algorithm has a strong ability to resist noise attack.
the ability to resist differential attacks. In order to calculate
without loss of generality, in this paper, we randomly perform G. DATA LOSS ATTACK ANALYSIS
100 times NPCR and UACI analyses on the test images, and In the process of information transmission, there is a high pos-
then we have the average value of the 100 times data. The sibility of data loss. At the same time, there are also attackers
final calculation results are shown in Table 8, which shows who want to make the correct image information unavailable
that the tested images can pass the test and are close to the to the receiver by maliciously interfering with the normal
theoretical value. In addition, the comparison of NPCR and transmission of the information. It is desirable for the receiver
UACI for different encryption algorithms using Lena images to recover the original image as much as possible in case of

67682 VOLUME 10, 2022

Z. Zhang et al.: Color Image Encryption Using 2D Sine-Cosine Coupling Map

FIGURE 12. Salt and pepper noise test: (a) decrypted image with noise density of 1%; (b) decrypted image with noise density
of 5%; (c) decrypted image with noise density of 10%; (d) decrypted image with noise density of 20%.

FIGURE 13. Analysis of information loss: (a) 1/16 loss of ciphertext information; (b) 1/8 loss of ciphertext information; (c) 1/4 loss
of ciphertext information; (d) 1/2 loss of ciphertext information; (e) decrypted image of (a); (f) decrypted image of (b);
(g) decrypted image of (c); (h) decrypted image of (d).

TABLE 10. Comparative analysis of encryption/decryption speed.

partial data loss. Therefore, in order to test the ability of our algorithms is shown in Table 10. The results show that the
proposed algorithm to cope with information loss, we first encryption and decryption speed of the proposed algorithm
perform the following four processes on the cipher image: is better than these algorithms. Therefore, the proposed algo-
(a) 1/16 of the data is lost in all components of the cipher rithm has high encryption efficiency.
image; (b) 1/8 of the data is lost in all components of the
cipher image; (c) 1/4 of the data is lost in all components of VI. CONCLUSION
the cipher image; (d) 1/2 of the data is lost in all components In this paper, we propose a new two-dimensional discrete
of the cipher image. Then, the cipher images with data loss are chaotic system, called 2D-SCCM, which is designed based on
decrypted using the correct key. Finally the results are shown the study of the Logistic map and the cosine function. In order
in Fig.13. The results show that most of the information of to better evaluate the chaotic behavior of the proposed system,
the image can still be recovered when 1/2 of the cipher image we use various testing methods, including trajectory distri-
is lost. Thus, the proposed algorithm has the ability to resist bution map, Lyapunov exponent, sample entropy, sequence
information loss. sensitivity. And the randomness of the generated time series
was examined by using the NIST test tool. Experimental
H. SPEED ANALYSIS results and comparative analyses show that the 2D-SCCM
In order to better analyze the encryption and decryption speed has a wider hyperchaotic interval, more complex chaotic
of the proposed algorithm, the speed of different encryption behavior, and better ergodicity compared to some existing

VOLUME 10, 2022 67683

 Z. Zhang et al.: Color Image Encryption Using 2D Sine-Cosine Coupling Map

chaotic systems. Based on the 2D-SCCM, we further propose [17] L. Teng, X. Wang, F. Yang, and Y. Xian, ‘‘Color image encryption based
a new color image encryption algorithm by combining the on cross 2D hyperchaotic map using combined cycle shift scrambling
and selecting diffusion,’’ Nonlinear Dyn., vol. 105, no. 2, pp. 1859–1876,
hash function. The algorithm consists of four main parts: Jul. 2021.
key update, XOR-diffusion, pixel scrambling and substitu- [18] Z. Hua, Z. Zhu, S. Yi, Z. Zhang, and H. Huang, ‘‘Cross-plane colour image
tion combined with S-Box, and Modular diffusion. Firstly, encryption using a two-dimensional logistic tent modular map,’’ Inf. Sci.,
vol. 546, pp. 1063–1083, Feb. 2021.
the updated key associates with the plain image, which [19] Z. Hua, Y. Zhou, C.-M. Pun, and C. L. P. Chen, ‘‘2D Sine Logistic
makes the whole algorithm effective against the selective modulation map for image encryption,’’ Inf. Sci., vol. 297, pp. 80–94,
plaintext attack. Then, the XOR-diffusion and the Modu- Mar. 2015.
[20] R. M. May, ‘‘Simple mathematical models with very complicated dynam-
lar diffusion can effectively change the value of the pixels. ics,’’ Nature, vol. 261, pp. 459–467, Jun. 1976.
Finally, in the pixel scrambling and substitution combined [21] Z. Hua and Y. Zhou, ‘‘Image encryption using 2D logistic-adjusted-sine
with S-Box stage, not only the value of the pixels can be map,’’ Inf. Sci., vol. 339, pp. 237–253, Apr. 2016.
[22] J. S. Richman and J. R. Moorman, ‘‘Physiological time-series analysis
changed, but also the position of the pixels, thereby breaking using approximate entropy and sample entropy,’’ Amer. J. Physiol.-Heart
the correlation between the plain image pixels. Experimental Circulatory Physiol., vol. 278, no. 6, pp. H2039–H2049, Jun. 2000.
simulations show that the proposed encryption algorithm can [23] Z. Hua, Y. Zhou, and H. Huang, ‘‘Cosine-transform-based chaotic system
for image encryption,’’ Inf. Sci., vol. 480, no. 1, pp. 403–419, Apr. 2019.
effectively convert natural images into unrecognizable noise- [24] L. E. Bassham. (Sep. 2010). A Statistical Test Suite for Random and Pseu-
like images. The security analyses show that the algorithm is dorandom Number Generators for Cryptographic Applications. Accessed:
able to resist various common cryptanalysis attacks. Mar. 16, 2022. [Online]. Available: https://www.nist.gov/publications/
statistical-test-suite-random-and-pseudorandom-number-generators-
cryptographics
REFERENCES [25] G. Alvarez and L. I. Shujun, ‘‘Some basic cryptographic requirements
for chaos-based cryptosystems,’’ Int. J. Bifurcation Chaos, vol. 16, no. 8,
[1] M. Fallahpour, ‘‘Reversible image data hiding based on gradient
pp. 2129–2151, Aug. 2006.
adjusted prediction,’’ IEICE Electron. Exp., vol. 5, no. 20, pp. 76–870,
[26] G. Hu and B. Li, ‘‘Coupling chaotic system based on unit transform and
Oct. 2008.
its applications in image encryption,’’ Signal Process., vol. 178, Jan. 2021,
[2] C. Wang, X. Wang, Z. Xia, and C. Zhang, ‘‘Ternary radial harmonic Fourier
Art. no. 107790.
moments based robust stereo image zero-watermarking algorithm,’’ Inf.
[27] K. Xuejing and G. Zihui, ‘‘A new color image encryption scheme based
Sci., vol. 470, pp. 109–120, Jan. 2019.
on DNA encoding and spatiotemporal chaotic system,’’ Signal Process.,
[3] Z. Hua and Y. Zhou, ‘‘Design of image cipher using block-based scram- Image Commun., vol. 80, Feb. 2020, Art. no. 115670.
bling and image filtering,’’ Inf. Sci., vol. 396, pp. 97–113, Aug. 2017. [28] X. Wu, K. Wang, X. Wang, H. Kan, and J. Kurths, ‘‘Color image DNA
[4] Z. Hua, Z. Zhu, Y. Chen, and Y. Li, ‘‘Color image encryption using encryption using NCA map-based CML and one-time keys,’’ Signal Pro-
orthogonal Latin squares and a new 2D chaotic system,’’ Nonlinear Dyn., cess., vol. 148, pp. 272–287, Jul. 2018.
vol. 104, pp. 4505–4522, May 2021. [29] J. Chen, J. Tang, F. Zhang, H. Ni, and Y. Tang, ‘‘A novel digital color
[5] J. Wu, X. Liao, and B. Yang, ‘‘Image encryption using 2D henon-sine map image encryption algorithm based on a new 4-D hyper-chaotic system and
and DNA approach,’’ Signal Process., vol. 153, pp. 11–23, Dec. 2018. an improved S-box,’’ Int. J. Innov. Comput., Inf. Control, vol. 18, no. 1,
[6] Y. Luo, M. Du, and J. Liu, ‘‘A symmetrical image encryption scheme pp. 73–92, Feb. 2022.
in wavelet and time domain,’’ Commun. Nonlinear Sci., vol. 20, no. 2, [30] X. Wu, J. Kurths, and H. Kan, ‘‘A robust and lossless DNA encryp-
pp. 447–460, Feb. 2015. tion scheme for color images,’’ Multimed Tools Appl., vol. 77, no. 10,
[7] M. Ramzan, T. Shah, M. M. Hazzazi, A. Aljaedi, and A. R. Alharbi, pp. 12349–12376, 2018.
‘‘Construction of S-boxes using different maps over elliptic curves for [31] A. U. Rehman, X. Liao, R. Ashraf, S. Ullah, and H. Wang, ‘‘A color
image encryption,’’ IEEE Access, vol. 9, pp. 157106–157123, 2021. image encryption technique using exclusive-OR with DNA complemen-
[8] Z. Hua, K. Zhang, Y. Li, and Y. Zhou, ‘‘Visually secure image encryption tary rules based on chaos theory and SHA-2,’’ Optik, vol. 159, pp. 348–367,
using adaptive-thresholding sparsification and parallel compressive sens- Apr. 2018.
ing,’’ Signal Process., vol. 183, Jun. 2021, Art. no. 107998. [32] Y. Wu, J. P. Noonan, and S. Agaian, ‘‘NPCR and UACI randomness
[9] J. Arif, M. A. Khan, B. Ghaleb, J. Ahmad, A. Munir, U. Rashid, and tests for image encryption,’’ Cyber J., Multidisciplinary J. Sci. Technol.,
A. Y. Al-Dubai, ‘‘A novel chaotic permutation-substitution image encryp- J. Select. Areas Telecommu., vol. 1, pp. 31–38, Apr. 2011. [Online]. Avail-
tion scheme based on logistic map and random substitution,’’ IEEE Access, able: http://www.cyberjournals.com/Papers/Apr2011/05.pdf
vol. 10, pp. 12966–12982, 2022. [33] G. Kaur, R. Agarwal, and V. Patidar, ‘‘Color image encryption
[10] M. A. B. Farah, R. Guesmi, A. Kachouri, and M. Samet, ‘‘A novel scheme based on fractional Hartley transform and chaotic substitution–
chaos based optical image encryption using fractional Fourier transform permutation,’’ Vis. Comput., vol. 38, no. 3, pp. 1027–1050, Mar. 2022.
and DNA sequence operation,’’ Opt. Laser Technol., vol. 121, Jan. 2020, [34] F. Zhang, X. Zhang, M. Cao, F. Ma, and Z. Li, ‘‘Characteristic analysis
Art. no. 105777. of 2D lag-complex logistic map and its application in image encryption,’’
[11] P. Mani, R. Rajan, L. Shanmugam, and Y. H. Joo, ‘‘Adaptive con- IEEE MultimediaMag., vol. 28, no. 4, pp. 96–106, Oct. 2021.
trol for fractional order induced chaotic fuzzy cellular neural networks
and its application to image encryption,’’ Inf. Sci., vol. 491, pp. 74–89,
Jul. 2019.
[12] Z. Hua, F. Jin, B. Xu, and H. Huang, ‘‘2D logistic-sine-coupling map for
image encryption,’’ Signal Process., vol. 149, pp. 148–161, Aug. 2018.
[13] Y. Zhou, L. Bao, and C. L. P. Chen, ‘‘Image encryption using a new
ZEZONG ZHANG received the B.S. degree in
parametric switching chaotic system,’’ Signal Process., vol. 93, no. 11,
pp. 3039–3052, 2013.
electronic information science and technology
[14] J. Fridrich, ‘‘Image encryption based on chaotic maps,’’ in Proc. IEEE
from the College of Electronic Information Engi-
Int. Conf. Syst., Man, Cybern. Comput. Simulation, vol. 2, Oct. 1997, neering, China West Normal University, China,
pp. 1105–1110. in 2021. He is currently pursuing the degree with
[15] A. Mansouri and X. Wang, ‘‘A novel one-dimensional sine powered chaotic the College of Information Science and Engi-
map and its application in a new image encryption scheme,’’ Inf. Sci., neering, Huaqiao University, China. His research
vol. 520, pp. 46–62, May 2020. interests include nonlinear dynamics and con-
[16] Z. Hua, J. Li, Y. Chen, and S. Yi, ‘‘Design and application of an S-box using trol, information security, cryptanalysis, and the
complete Latin square,’’ Nonlinear Dyn., vol. 104, no. 1, pp. 807–825, application of chaotic cryptography in image
Mar. 2021. processing.

67684 VOLUME 10, 2022

Z. Zhang et al.: Color Image Encryption Using 2D Sine-Cosine Coupling Map

JIANENG TANG received the B.Sc. degree in JINYUAN CHEN is currently pursuing the M.S.
electronic information science and technology degree with the College of Engineering, Huaqiao
from Xijiang Normal University, China, in 2006, University, Quanzhou, China. His research inter-
the M.Sc. degree in circuits and systems from ests include information security, chaotic synchro-
Ningxia University, China, in 2009, and the Ph.D. nization and control, and the application of chaotic
degree in information and communication engi- cryptography in image processing.
neering from Southeast University, China, in 2012.
He is currently an Associate Professor at the Col-
lege of Engineering, Huaqiao University, China.
He has published over 30 papers in journals and
conferences. His research interests include image encryption, RF circuit
design, complex network synchronization, and chaos synchronization and
control.

FENG ZHANG received the B.Sc. degree in

applied physics from the University of Elec-
tronic Science and Technology of China, China,
in 2007. He is currently a Deputy General Man-
ager of Fujian MM Electronics Company Ltd. His
research interests include image encryption and
RF circuit design.

HUI NI received the B.Sc. degree in project ZHONGMING HUANG is currently pursuing
management from Fuzhou University, China, the M.S. degree with the College of Engineer-
in 2014. He is currently a Deputy General Man- ing, Huaqiao University, Quanzhou, China. His
ager of Fujian MM Electronics Company Ltd. His research interests include information security and
research interests include image encryption and image privacy protection.
RF circuit design.

VOLUME 10, 2022 67685