With the rapid development and application of Internet technology in recent years, the issue of information security has received more and more attention. Digital steganography is used as a means of secure communication to hide information by modifying the carrier. However, steganography can also be used for illegal acts, so it is of great significance to study steganalysis techniques. The steganalysis technology can be used to solve the illegal steganography problem of computer vision and engineering applications technology. Most of the images in the Internet are color images, and steganalysis for color images is a very critical problem in the field of steganalysis at this stage. Currently proposed algorithms for steganalysis of color images mainly rely on the manual design of steganographic features, and the steganographic features do not fully consider the internal connection between the three channels of color images. In recent years, advanced steganography techniques for color images have been proposed, which brings more serious challenges to color image steganalysis. Quaternions are a good tool to represent color images, and the transformation of quaternions can fully exploit the correlation among color image channels. In this paper, we propose a color image steganalysis algorithm based on quaternion discrete cosine transform, firstly, the image is represented by quaternion, then the quaternion discrete cosine transform is applied to it, and the coefficients obtained from the transformation are extracted to design features of the coeval matrix. The experimental results show that the proposed algorithm works better than the typical color image steganalysis algorithm.
Citation: Meng Xu, Xiangyang Luo, Jinwei Wang, Hao Wang. Color image steganalysis based on quaternion discrete cosine transform[J]. Electronic Research Archive, 2023, 31(7): 4102-4118. doi: 10.3934/era.2023209
[1] | Harman Kaur, Meenakshi Rana . Congruences for sixth order mock theta functions λ(q) and ρ(q). Electronic Research Archive, 2021, 29(6): 4257-4268. doi: 10.3934/era.2021084 |
[2] | Meenakshi Rana, Shruti Sharma . Combinatorics of some fifth and sixth order mock theta functions. Electronic Research Archive, 2021, 29(1): 1803-1818. doi: 10.3934/era.2020092 |
[3] | Changjian Wang, Jiayue Zhu . Global dynamics to a quasilinear chemotaxis system under some critical parameter conditions. Electronic Research Archive, 2024, 32(3): 2180-2202. doi: 10.3934/era.2024099 |
[4] | Chang-Jian Wang, Yu-Tao Yang . Boundedness criteria for the quasilinear attraction-repulsion chemotaxis system with nonlinear signal production and logistic source. Electronic Research Archive, 2023, 31(1): 299-318. doi: 10.3934/era.2023015 |
[5] | Maoji Ri, Shuibo Huang, Canyun Huang . Non-existence of solutions to some degenerate coercivity elliptic equations involving measures data. Electronic Research Archive, 2020, 28(1): 165-182. doi: 10.3934/era.2020011 |
[6] | Nan Li . Summability in anisotropic mixed-norm Hardy spaces. Electronic Research Archive, 2022, 30(9): 3362-3376. doi: 10.3934/era.2022171 |
[7] | Lili Li, Boya Zhou, Huiqin Wei, Fengyan Wu . Analysis of a fourth-order compact θ-method for delay parabolic equations. Electronic Research Archive, 2024, 32(4): 2805-2823. doi: 10.3934/era.2024127 |
[8] | Jianxing Du, Xifeng Su . On the existence of solutions for the Frenkel-Kontorova models on quasi-crystals. Electronic Research Archive, 2021, 29(6): 4177-4198. doi: 10.3934/era.2021078 |
[9] | Zihan Zheng, Juan Wang, Liming Cai . Global boundedness in a Keller-Segel system with nonlinear indirect signal consumption mechanism. Electronic Research Archive, 2024, 32(8): 4796-4808. doi: 10.3934/era.2024219 |
[10] | Ying Hou, Liangyun Chen . Constructions of three kinds of Bihom-superalgebras. Electronic Research Archive, 2021, 29(6): 3741-3760. doi: 10.3934/era.2021059 |
With the rapid development and application of Internet technology in recent years, the issue of information security has received more and more attention. Digital steganography is used as a means of secure communication to hide information by modifying the carrier. However, steganography can also be used for illegal acts, so it is of great significance to study steganalysis techniques. The steganalysis technology can be used to solve the illegal steganography problem of computer vision and engineering applications technology. Most of the images in the Internet are color images, and steganalysis for color images is a very critical problem in the field of steganalysis at this stage. Currently proposed algorithms for steganalysis of color images mainly rely on the manual design of steganographic features, and the steganographic features do not fully consider the internal connection between the three channels of color images. In recent years, advanced steganography techniques for color images have been proposed, which brings more serious challenges to color image steganalysis. Quaternions are a good tool to represent color images, and the transformation of quaternions can fully exploit the correlation among color image channels. In this paper, we propose a color image steganalysis algorithm based on quaternion discrete cosine transform, firstly, the image is represented by quaternion, then the quaternion discrete cosine transform is applied to it, and the coefficients obtained from the transformation are extracted to design features of the coeval matrix. The experimental results show that the proposed algorithm works better than the typical color image steganalysis algorithm.
Ramanujan's last letter to Hardy is one of the most mysterious and important mathematical letters in the history of mathematics. He introduced a class of functions that he called mock theta functions in his letter. For nearly a century, properties of these functions have been widely studied by different mathematicians. The important direction involves the arithmetic properties (see [1,2]), combinatorics (see [3,4]), identities between these functions, and generalized Lambert series (see [5,6]). For the interested reader, regarding the history and new developments in the study of mock theta functions, we refer to [7].
In 2007, McIntosh studied two second order mock theta functions in reference [8]; more details are given in reference [9]. These mock theta functions are:
A(q)=∞∑n=0q(n+1)2(−q;q2)n(q;q2)2(n+1)=∞∑n=0qn+1(−q2;q2)n(q;q2)n+1, | (1.1) |
B(q)=∞∑n=0qn(−q;q2)n(q;q2)n+1=∞∑n=0qn(n+1)(−q2;q2)n(q;q2)2n+1, | (1.2) |
where
(a;q)n=n−1∏i=0(1−aqi),(a;q)∞=∞∏i=0(1−aqi), |
(a1,a2,⋯,am;q)∞=(a1;q)∞(a2;q)∞⋯(am;q)∞, |
for |q|<1.
The functions A(q) and B(q) have been combinatorially interpreted in terms of overpartitions in [3] using the odd Ferrers diagram. In this paper, we study some arithmetic properties of one of the second order mock theta functions B(q). We start by noting, Bringmann, Ono and Rhoades [10] obtained the following identity:
B(q)+B(−q)2=f54f42, | (1.3) |
where
fkm:=(qm;qm)k∞, |
for positive integers m and k. We consider the function
B(q):=∞∑n=0b(n)qn. | (1.4) |
Followed by Eq (1.3), the even part of B(q) is given by:
∞∑n=0b(2n)qn=f52f41. | (1.5) |
In 2012, applying the theory of (mock) modular forms and Zwegers' results, Chan and Mao [5] established two identities for b(n), shown as:
∞∑n=0b(4n+1)qn=2f82f71, | (1.6) |
∞∑n=0b(4n+2)qn=4f22f44f51. | (1.7) |
In a sequel, Qu, Wang and Yao [6] found that all the coefficients for odd powers of q in B(q) are even. Recently, Mao [11] gave analogues of Eqs (1.6) and (1.7) modulo 6
∞∑n=0b(6n+2)qn=4f102f23f101f6, | (1.8) |
∞∑n=0b(6n+4)qn=9f42f43f6f81, | (1.9) |
and proved several congruences for the coefficients of B(q). Motivated from this, we prove similar results for b(n) by applying identities on the coefficients in arithmetic progressions. We present some congruence relations for the coefficients of B(q) modulo certain numbers of the form 2α⋅3β,2α⋅5β,2α⋅7β where α,β≥0. Our main theorems are given below:
Theorem 1.1. For n≥0, we have
∞∑n=0b(12n+9)qn=18[f92f123f171f36+2f52f43f6f91+28f62f33f66f141], | (1.10) |
∞∑n=0b(12n+10)qn=36[2f162f106f201f3f412−qf282f33f212f241f84f26−16q2f22f33f84f212f161f26]. | (1.11) |
In particular, b(12n+9)≡0(mod18),b(12n+10)≡0(mod36).
Theorem 1.2. For n≥0, we have
∞∑n=0b(18n+10)qn=72[f162f213f271f96+38qf132f123f241+64q2f102f33f96f211], | (1.12) |
∞∑n=0b(18n+16)qn=72[5f152f183f261f66+64qf122f93f36f231+32q2f92f126f201]. | (1.13) |
In particular, b(18n+10)≡0(mod72),b(18n+16)≡0(mod72).
Apart from these congruences, we find some relations between b(n) and restricted partition functions. Here we recall, Partition of a positive integer ν, is a representation of ν as a sum of non-increasing sequence of positive integers μ1,μ2,⋯,μn. The number of partitions of ν is denoted by p(ν) which is called the partition function. If certain conditions are imposed on parts of the partition, is called the restricted partition and corresponding partition function is named as restricted partition function. Euler proved the following recurrence for p(n) [12] [p. 12, Cor. 1.8]:
(n)−p(n−1)−p(n−2)+p(n−5)+p(n−7)−p(n−12)−p(n−15)+⋯+(−1)kp(n−k(3k−1)/2)+(−1)kp(n−k(3k+1)/2)+⋯={1, if n=0,0, otherwise. |
The numbers k(3k±1)/2 are pentagonal numbers. Following the same idea, different recurrence relations have been found by some researchers for restricted partition functions. For instance, Ewell [13] presented the recurrence for p(n) involving the triangular numbers. For more study of recurrences, see [14,15,16]. Under the influence of these efforts, we express the coefficients of mock theta function B(q) which are in arithmetic progression in terms of recurrence of some restricted partition functions.
This paper is organized as follows: Section 2, here we recall some preliminary lemmas and present the proof of Theorems 1.1 and 1.2. Section 3 includes some more congruences based on the above results. Section 4 depicts the links between b(n) and some of the restricted partition functions.
Before proving the results, we recall Ramanujan's theta function:
j(a,b)=∞∑n=−∞an(n+1)2bn(n−1)2, for|ab|<1. |
Some special cases of j(a,b) are:
ϕ(q):=j(q,q)=∞∑n=−∞qn2=f52f21f24,ψ(q):=j(q,q3)=∞∑n=0qn(n+1)/2=f22f1. |
Also,
ϕ(−q)=f21f2. |
The above function satisfy the following properties (see Entries 19, 20 in [17]).
j(a,b)=(−a,−b,ab;ab)∞,(Jacobi's triple product identity), |
j(−q,−q2)=(q;q)∞,(Euler's pentagonal number theorem). |
We note the following identities which will be used below.
Lemma 2.1. [[18], Eq (3.1)] We have
f32f31=f6f3+3qf46f59f83f18+6q2f36f29f218f73+12q3f26f518f63f9. | (2.1) |
Lemma 2.2. We have
f22f1=f6f29f3f18+qf218f9, | (2.2) |
f2f21=f46f69f83f318+2qf36f39f73+4q2f26f318f63. | (2.3) |
Proof. The first identity follows from [[19] Eq (14.3.3)]. The proof of second identity can be seen from [20].
Lemma 2.3. We have
1f41=f144f142f48+4qf24f48f102, | (2.4) |
f41=f104f22f48−4qf22f48f24. | (2.5) |
Proof. Identity (2.4) is Eq (1.10.1) from [19]. To obtain (2.5), replacing q by −q and then using
(−q;−q)∞=f32f1f4. |
Now, we present the proof of Theorems 1.1 and 1.2.
Proof of Theorems 1.1 and 1.2. From Eq (1.6), we have
∞∑n=0b(4n+1)qn=2(f32f31)3⋅f22f1. |
Substituting the values from Eqs (2.1) and (2.2) in above, we get
∞∑n=0b(4n+1)qn=2f36f29f33f18+2qf26f218f23f9+12qf66f79f103f218+18q2f96f129f173f318+36q2f56f49f18f93+90q3f86f99f163+72q3f46f9f418f83+48q4f36f718f73f29+288q4f76f69f318f153+504q5f66f39f618f143+576q6f56f918f133. | (2.6) |
Bringing out the terms involving q3n+2, dividing by q2 and replacing q3 by q, we get (1.10). Considering Eq (1.5), we have
∞∑n=0b(2n)qn=f32f31⋅f22f1. |
Substituting the values from Eqs (2.1) and (2.2), we obtain
∞∑n=0b(2n)qn=(f6f3+3qf46f59f83f18+6q2f36f29f218f73+12q3f26f518f63f9)(f6f29f3f18+qf218f9). |
Extracting the terms involving q3n,q3n+1,q3n+2 from the above equation, we have
∞∑n=0b(6n)qn=f22f23f21f6+18qf32f3f46f71, | (2.7) |
∞∑n=0b(6n+2)qn=f2f26f1f3+3f52f73f91f26+12qf22f76f61f23, | (2.8) |
∞∑n=0b(6n+4)qn=9f42f43f6f81. | (2.9) |
Using Eqs (2.4) and (2.5) in Eq (2.9), we get
∞∑n=0b(6n+4)qn=9f42f6(f144f142f48+4qf24f48f102)2(f1012f26f424−4q3f26f424f212). |
Extracting the terms involving q2n,q2n+1 from above, we arrive at
∞∑n=0b(12n+4)qn=9(f282f106f241f3f84f412+16qf42f84f106f161f3f412−32q2f162f33f412f201f26), | (2.10) |
∞∑n=0b(12n+10)qn=9(8f162f106f201f3f412−4qf282f33f412f241f84f26−16q2f42f33f84f412f161f26). | (2.11) |
From Eq (2.11), we ultimately arrive at Eq (1.11). To prove Theorem 1.2, consider Eq (2.9) as:
∞∑n=0b(6n+4)qn=9f43f6(f2f21)4. |
Using Eq (2.3) in above, we get
∞∑n=0b(6n+4)qn=9f176f249f283f1218+72qf166f219f273f918+360q2f156f189f263f618+288q3f146f159f253f318+864q3f126f159f193f618+2736q4f136f129f243+4608q5f126f99f318f233+5760q6f116f69f618f223+4608q7f106f39f918f213+2304q8f96f1218f203. | (2.12) |
Bringing out the terms involving q3n+1 and q3n+2 from Eq (2.12), we get Eqs (1.12) and (1.13), respectively.
This segment of the paper contains some more interesting congruence relations for b(n).
Theorem 3.1. For n≥0, we have
b(12n+1)≡{2(−1)k(mod6),ifn=3k(3k+1)/2,0(mod6),otherwise. | (3.1) |
Theorem 3.2. For n≥0, we have
b(2n)≡{(−1)k(2k+1)(mod4),ifn=k(k+1),0(mod4),otherwise. | (3.2) |
Theorem 3.3. For n≥0, we have
b(36n+10)≡0(mod72), | (3.3) |
b(36n+13)≡0(mod6), | (3.4) |
b(36n+25)≡0(mod12), | (3.5) |
b(36n+34)≡0(mod144), | (3.6) |
b(108n+t)≡0(mod18),for t∈{49,85}. | (3.7) |
Theorem 3.4. For n≥0, we have
b(20n+t)≡0(mod5),for t∈{8,16} | (3.8) |
b(20n+t)≡0(mod20),for t∈{6,18} | (3.9) |
b(20n+17)≡0(mod10), | (3.10) |
b(28n+t)≡0(mod14),for t∈{5,21,25}. | (3.11) |
Proof of Theorem 3.1. From Eq (2.6), picking out the terms involving q3n and replacing q3 by q, we have
∞∑n=0b(12n+1)qn=2f32f23f31f6+90qf82f93f161+72qf42f3f46f81+576q2f52f96f31. | (3.12) |
Reducing modulo 6, we obtain
∞∑n=0b(12n+1)qn≡2f3(mod6). | (3.13) |
With the help of Euler's pentagonal number theorem,
∞∑n=0b(12n+1)qn≡2∞∑k=−∞(−1)kq3k(3k+1)2(mod6), |
which completes the proof of Theorem 3.1.
Proof of Theorem 3.2. Reducing Eq (1.5) modulo 4, we get
∞∑n=0b(2n)qn≡f32(mod4). | (3.14) |
From Jacobi's triple product identity, we obtain
∞∑n=0b(2n)qn≡∞∑k=0(−1)k(2k+1)qk(k+1)(mod4), |
which completes the proof of Theorem 3.2.
Proof of Theorem 3.3. Consider Eq (1.11), reducing modulo 72
∞∑n=0b(12n+10)qn≡36qf282f33f412f241f84f26(mod72), |
∞∑n=0b(12n+10)qn≡36qf282f33f412f122f84f12=36qf162f33f312f84(mod72) |
or
∞∑n=0b(12n+10)qn≡36qf33f312(mod72). | (3.15) |
Extracting the terms involving q3n, replacing q3 by q in Eq (3.15), we arrive at Eq (3.3). Similarly, consider Eq (1.13) and reducing modulo 144, we have
∞∑n=0b(18n+16)qn≡72⋅5f152f183f261f66(mod144),≡72f152f96f132f66=72f22f36(mod144). |
Extracting the terms involving q2n+1, dividing both sides by q and replacing q2 by q, we get Eq (3.6).
From Eq (3.20), we get
∞∑n=0b(12n+1)qn≡2f3(mod6). |
Bringing out the terms containing q3n+1, dividing both sides by q and replacing q3 by q, we have b(36n+13)≡0(mod6). Reducing Eq (3.12) modulo 12, we have
∞∑n=0b(12n+1)qn≡2f32f23f31f6+90qf82f93f161(mod12), |
∞∑n=0b(12n+1)qn≡2f23f6(f6f3+3qf46f59f83f18+6q2f36f29f218f73+12q3f26f518f63f9)+6qf82f93f82. |
Extracting the terms containing q3n+2, dividing by q2 and replacing q3 by q, we obtain Eq (3.5). Reducing Eq (3.12) modulo 18,
∞∑n=0b(12n+1)qn≡2f32f23f31f6(mod18),=2f23f6(f6f3+3qf46f59f83f18+6q2f36f29f218f73+12q3f26f518f63f9). |
Extracting the terms involving q3n+1, dividing both sides by q and replacing q3 by q, we have
∞∑n=0b(36n+13)qn≡6f32f53f61f6≡6f6f53f23f6(mod18) |
or
∞∑n=0b(36n+13)qn≡6f33(mod18). |
Extracting the terms containing q3n+1,q3n+2 from above to get Eq (3.7).
Proof of Theorem 3.4. From Eqs (1.5) and (2.4), we have
∞∑n=0b(2n)qn=f52(f144f142f48+4qf24f48f102). |
Bringing out the terms containing even powers of q, we obtain
∞∑n=0b(4n)qn=f142f91f44, |
which can be written as:
∞∑n=0b(4n)qn=f152f101f54.f1f4f2≡f310f25f20.f1f4f2(mod5). |
Here
f1f4f2=(q;q)∞(q4;q4)∞(q2;q2)∞,=(q;q2)∞(q2;q2)∞(q4;q4)∞(q2;q2)∞, |
f1f4f2=(q,q3,q4;q4)∞=∞∑n=−∞(−1)nq2n2−n, | (3.16) |
where the last equality follows from Jacobi's triple product identity. Using the above identity, we have
∞∑n=0b(4n)qn≡f310f25f20∞∑n=−∞(−1)nq2n2−n(mod5). | (3.17) |
Since 2n2−n≢2,4(mod5), it follows that the coefficients of q5n+2,q5n+4 in ∑∞n=0b(4n)qn are congruent to 0(mod5), which proves that b(20n+t)≡0(mod5), for t∈{8,16}.
Consider Eq (1.7)
∞∑n=0b(4n+2)qn=4f54f51f22f4≡4f20f5f22f4(mod20). |
Now
f22f4=(q2;q2)2∞(q4;q4)∞,=(q2;q2)∞(q2;q4)∞(q4;q4)∞(q4;q4)∞, |
f22f4=(q2,q2,q4;q4)∞=∞∑n=−∞(−1)nq2n2. |
Using the above identity, we get
∞∑n=0b(4n+2)qn≡4f20f5∞∑n=−∞(−1)nq2n2(mod20). | (3.18) |
Since 2n2≢1,4(mod5), it follows that the coefficients of q5n+1,q5n+4 in ∑∞n=0b(4n+2)qn are congruent to 0(mod20), which proves Eq (3.9). For the proof of next part, consider Eq (1.6) as:
∞∑n=0b(4n+1)qn=2f52f101f31f32≡2f10f25f31f32(mod10), |
∞∑n=0b(4n+1)qn≡2f10f25∞∑k=0(−1)k(2k+1)qk(k+1)2∞∑m=0(−1)m(2m+1)qm(m+1)(mod10). | (3.19) |
Therefore, to contribute the coefficient of q5n+4, (k,m)≡(2,2)(mod5) and thus the contribution towards the coefficient of q5n+4 is a multiple of 5.
Consider Eq (1.6) as:
∞∑n=0b(4n+1)qn=2f72f71f2≡2f14f7f2(mod14). |
With the help of Euler's pentagonal number theorem,
∞∑n=0b(4n+1)qn≡2f14f7∞∑n=−∞(−1)nqn(3n+1)(mod14). | (3.20) |
As n(3n+1)≢1,5,6(mod7), it readily proves Eq (3.11).
In this section, we find some recurrence relations connecting b(n) and restricted partition functions. First we define some notations. Let ¯pl(n) denotes the number of overpartitions of n with l copies. Then
∞∑n=0¯pl(n)qn=(f2f21)l. |
Let pld(n) denotes the number of partitions of n into distinct parts with l copies. Then
∞∑n=0pld(n)qn=(f2f1)l. |
Theorem 4.1. We have
b(2n)=¯p2(n)−3¯p2(n)+5¯p2(n)+⋯+(−1)k(2k+1)¯p2(n−k(k+1))+⋯, | (4.1) |
(2n)=p4d(n)−p4d(n−2)−p4d(n−4)+p4d(n−10)+p4d(n−14)+⋯+(−1)kp4d(n−k(3k−1))+(−1)kp4d(n−k(3k+1))+⋯. | (4.2) |
Theorem 4.2.
(4n+1)=2p8d(n)−2p8d(n−1)−2p8d(n−2)+2p8d(n−5)+2p8d(n−7)+⋯+(−1)k2p8d(n−k(3k−1)2)+(−1)k2p8d(n−k(3k+1)2)+⋯, | (4.3) |
b(4n+1)=2n∑c=0b(2c)p3d(n−c). | (4.4) |
Theorem 4.3.
(6n+2)=4p10d(n)−8p10d(n−3)+8p10d(n−12)+8p10d(n−27)+⋯+8(−1)kp10d(n−3k2)+⋯. | (4.5) |
Proof of Theorem 4.1. Consider (1.5) as:
∞∑n=0b(2n)qn=(f2f21)2⋅f32. |
Then
∞∑n=0b(2n)qn=(∞∑n=0¯p2(n)qn)(∞∑k=0(−1)k(2k+1)qk(k+1)),=∞∑n=0∞∑k=0(−1)k(2k+1)¯p2(n)qn+k(k+1),=∞∑n=0(∞∑k=0(−1)k(2k+1)¯p2(n−k(k+1)))qn. |
From the last equality, we readily arrive at (4.1). To prove (4.2), consider (1.5) as:
∞∑n=0b(2n)qn=(f2f1)4⋅f2,=(∞∑n=0p4d(n)qn)(∞∑k=−∞(−1)kqk(3k+1)),=(∞∑n=0p4d(n)qn)(1+∞∑k=1(−1)kqk(3k−1)+∞∑k=1(−1)kqk(3k+1)), |
∞∑n=0b(2n)qn=∞∑n=0p4d(n)qn+∞∑n=0(∞∑k=1(−1)kp4d(n)qk(3k−1)+n)+∞∑n=0(∞∑k=1(−1)kp4d(n)qk(3k+1)+n), |
∞∑n=0b(2n)qn=∞∑n=0p4d(n)qn+∞∑n=0(∞∑k=1(−1)kp4d(n−k(3k−1))qn)+∞∑n=0(∞∑k=1(−1)kp4d(n−k(3k+1))qn), |
which proves Eq (4.2).
Proof of Theorem 4.2. Consider Eq (1.6) as:
∞∑n=0b(4n+1)qn=2(f2f1)8f1,=2(∞∑n=0p8d(n)qn)(∞∑k=−∞(−1)kqk(3k+1)2),=2(∞∑n=0p8d(n)qn)(1+∞∑k=1(−1)kqk(3k−1)/2+∞∑k=1(−1)kqk(3k+1)/2), |
∞∑n=0b(4n+1)qn=∞∑n=0p8d(n)qn+∞∑n=0∞∑k=1(−1)kp8d(n)qk(3k−1)/2+n+∞∑n=0∞∑k=1(−1)kp8d(n)qk(3k+1)/2+n, |
∞∑n=0b(4n+1)qn=∞∑n=0p8d(n)qn+∞∑n=0(∞∑k=1(−1)kp8d(n−k(3k−1)2))qn+∞∑n=0(∞∑k=1(−1)kp8d(n−k(3k+1)2))qn, |
which proves Eq (4.3). To prove Eq (4.4), consider Eq (1.6) as:
∞∑n=0b(4n+1)qn=2(f52f41)f32f31,=2(∞∑n=0b(2n)qn)(∞∑k=0p3d(k)qk),=2∞∑n=0(n∑c=0b(2c)p3d(n−c))qn. |
Comparing the coefficients of qn, we arrive at Eq (4.4).
Proof of Theorem 4.3. Consider Eq (1.8) as:
∞∑n=0b(6n+2)qn=4(f2f1)10⋅f23f6,=4(∞∑n=0p10d(n)qn)(∞∑k=−∞(−1)kq3k2),=4(∞∑n=0p10d(n)qn)(1+2∞∑k=1(−1)kq3k2),=4∞∑n=0p10d(n)qn+8∞∑n=0(∞∑k=1(−1)kp10d(n)q3k2+n),=4∞∑n=0p10d(n)qn+8∞∑n=0(∞∑k=1(−1)kp10d(n−3k2))qn. |
Comparing the coefficients of qn to obtain Eq (4.5).
In this paper, we have provided the arithmetic properties of second order mock theta function B(q), introduced by McIntosh. Some congruences are proved for the coefficients of B(q) modulo specific numbers. The questions which arise from this work are:
(i) Are there exist congruences modulo higher primes for B(q)?
(ii) Is there exist any other technique (like modular forms) that helps to look for some more arithmetic properties of B(q)?
(iii) How can we explore the other second order mock theta function A(q)?
The first author is supported by University Grants Commission (UGC), under grant Ref No. 971/(CSIR-UGC NET JUNE 2018) and the the second author is supported by Science and Engineering Research Research Board (SERB-MATRICS) grant MTR/2019/000123. The authors of this paper are thankful to Dr. Rupam Barman, IIT Guwahati, for his valuable insight during establishing Theorems 3.1 and 3.2. We would like to thank the referee for carefully reading our paper and offering corrections and helpful suggestions.
The authors declare there is no conflicts of interest.
[1] | I. Cox, M. Miller, J. Bloom, J. Fridrich, T. Kalker, Digital Watermarking and Steganography, Morgan Kaufmann, 2007. https://doi.org/10.1016/B978-0-12-372585-1.X5001-3 |
[2] |
W. Luo, F. Huang, J. Huang, Edge adaptive image steganography based on LSB matching revisited, IEEE Trans. Inf. Forensics Secur., 5 (2010), 201-214. https://doi.org/10.1109/TIFS.2010.2041812 doi: 10.1109/TIFS.2010.2041812
![]() |
[3] | E. Kawaguchi, R. O. Eason, Principles and applications of BPCS steganography, in Proceedings of the Photonics East (ISAM, VVDC, IEMB), International Society for Optics and Photonics, 3528 (1999), 464-473. https://doi.org/10.1117/12.337436 |
[4] | V. K. Sharma, V. Shrivastava, A steganography algorithm for hiding image in image by improved LSB substitution by minimize detection, J. Theor. Appl. Inf. Technol., 36 (2012), 1-8. Available from: http://www.jatit.org/volumes/Vol36No1/1Vol36No1.pdf. |
[5] | B. C. Nguyen, S. M. Yoon, H. K. Lee, Multi bit plane image steganography, in Digital Watermarking, Springer, (2006), 61-70. https://doi.org/10.1007/11922841_6 |
[6] |
M. Omoomi, S. Samavi, S. Dumitrescu, An efficient high payload ±1 data embedding scheme, Multimedia Tools Appl., 54 (2011), 201-218. https://doi.org/10.1007/s11042-010-0517-z doi: 10.1007/s11042-010-0517-z
![]() |
[7] | D. Upham, Jpeg-jsteg-v4, 2008. Available from: http://www.funet.fi/pub/crypt/steganography/jpeg-jsteg-v4.diff.gz. |
[8] |
F. A. Petitcolas, R. J. Anderson, M. G. Kuhn, Information hiding-a survey, Proc. IEEE, 87 (1999), 1062-1078. https://doi.org/10.1109/5.771065 doi: 10.1109/5.771065
![]() |
[9] | N. Provos, Defending against statistical steganalysis, in Proceedings of the Usenix Security Symposium, 10 (2001), 323-336. https://dl.acm.org/doi/10.5555/1251327.1251351 |
[10] | P. Sallee, Model-based steganography, in Digital Watermarking, Springer, (2004), 154-167. https://doi.org/10.1007/978-3-540-24624-4_12 |
[11] | T. Pevný, T. Filler, P. Bas, Using high-dimensional image models to perform highly undetectable steganography, in International Workshop on Information Hiding, 6387 (2010), 161-177. https://doi.org/10.1007/978-3-642-16435-4_13 |
[12] |
J. Fridrich, M. Goljan, D. Soukal, Perturbed quantization steganography, Multimedia Syst., 11 (2005), 98-107. https://doi.org/10.1007/s00530-005-0194-3 doi: 10.1007/s00530-005-0194-3
![]() |
[13] | X. Song, F. Liu, C. Yang, X. Luo, Y. Zhang, Steganalysis of adaptive JPEG steganography using 2D gabor filters, in Proceedings of the 3rd ACM Workshop on Information Hiding and Multimedia Security, (2015), 15-23. https://doi.org/10.1145/2756601.2756608 |
[14] | A. Westfeld, A. Pfitzmann, Attacks on steganographic systems, in International Workshop on Information Hiding, Springer, Berlin, Heidelberg, (2000), 61-76. https://doi.org/10.1007/10719724_5 |
[15] | J. Fridrich, M. Goljan, D. Rui, Reliable detection of LSB steganography in color and grayscale images, in Proceedings of the 2001 Workshop on Multimedia and Security: New Challenges, 2002. https://doi.org/10.1145/1232454.1232466 |
[16] |
S. Dumitrescu, X. Wu, W. Zhe, Detection of LSB steganography via sample pair analysis, IEEE Trans. Signal Process., 51 (2002), 1995-2007. https://doi.org/10.1109/TSP.2003.812753 doi: 10.1109/TSP.2003.812753
![]() |
[17] | J. Fridrich, M. Goljan, On estimation of secret message length in LSB steganography in spatial domain, in Security, Steganography, and Watermarking of Multimedia Contents VI, San Jose, California, USA, 2004. https://doi.org/10.1117/12.521350 |
[18] | J. Fridrich, M. Long, Steganalysis of LSB encoding in color images, in Proceedings of IEEE International Conference on Multimedia and Expo, (2000), 1279-1282. https://doi.org/10.1109/ICME.2000.871000 |
[19] |
W. You, H. Zhang, X. Zhao, A siamese CNN for image steganalysis, IEEE Trans. Inf. Forensics Secur., 16 (2020), 291-306. https://doi.org/10.1109/TIFS.2020.3013204 doi: 10.1109/TIFS.2020.3013204
![]() |
[20] |
K. Wei, W. Luo, S. Tan, J. Huang, Universal deep network for steganalysis of color image based on channel representation, IEEE Trans. Inf. Forensics Secur., 17 (2022), 3022-3036. https://doi.org/10.1109/TIFS.2022.3196265 doi: 10.1109/TIFS.2022.3196265
![]() |
[21] | G. Su, H. Lu, H. Fang, B. Yang, A steganalysis algorithm based on statistic characteristics of the color images, in Proceedings of International Conference on Computer Science and Network Technology, (2011), 2294-2297. https://doi.org/10.1109/ICCSNT.2011.6182432 |
[22] |
H. Abdulrahman, M. Chaumont, P. Montesinos, B. Magnier, Color images steganalysis using RGB channel geometric transformation measures, Secur. Commun. Netw., 9 (2016), 2945-2956. https://doi.org/10.1002/sec.1427 doi: 10.1002/sec.1427
![]() |
[23] | M. Goljan, J. Fridrich, R. Cogranne, Rich model for steganalysis of color images, in Proceedings of IEEE International Workshop on Information Forensics and Security, (2014), 185-190. https://doi.org/10.1109/WIFS.2014.7084325 |
[24] | M. Goljan, J. Fridrich, CFA-aware features for steganalysis of color images, in Proceedings of SPIE 9409, Media Watermarking, Security, and Forensics, (2015), 1-13. https://doi.org/10.1117/12.2078399 |
[25] |
X. Liao, G. Chen, J. Yin, Content-adaptive steganalysis for color images, Secur. Commun. Netw., 9 (2016), 5756-5763. https://doi.org/10.1002/sec.1734 doi: 10.1002/sec.1734
![]() |
[26] | S. Lyu, H. Farid, Steganalysis using color wavelet statistics and one-class support vector machines, in Proceedings of SPIE 5306, Security, Steganography, and Watermarking of Multimedia Contents VI, (2004), 35-45. https://doi.org/10.1117/12.526012 |
[27] | Q. Liu, A. H. Sung, J. Xu, B. M. Ribeiro, Image complexity and feature extraction for steganalysis of LSB matching steganography, in Proceedings of the 18th International Conference on Pattern Recognition, (2006), 267-270. https://doi.org/10.1109/ICPR.2006.684 |
[28] |
F. Y. Li, X. P. Zhang, J. Yu, Steganalysis for color JPEG images based on ensemble proportion training, J. Electron. Inf. Technol., 36 (2014), 114-120. https://doi.org/10.3724/SP.J.1146.2013.00443 doi: 10.3724/SP.J.1146.2013.00443
![]() |
[29] | H. Abdulrahman, M. Chaumont, P. Montesinos, B. Magnier, Color image steganalysis based on steerable gaussian filters bank, in Proceedings of the 4th ACM Workshop on Information Hiding and Multimedia Security, (2016), 109-114. https://doi.org/10.1145/2909827.2930799 |
[30] | Y. Kang, F. Liu, C. Yang, L. Xiang, X. Luo, P. Wang, Color image steganalysis based on channel gradient correlation, Int. J. Distrib. Sens. Netw., 15 (2019). https://doi.org/10.1177/1550147719852031 |
[31] |
C. Qin, Y. Chen, K. Chen, X. Dong, W. Zhang, Y. He, et al., Feature fusion based adversarial example detection against second-round adversarial attacks, IEEE Trans. Artif. Intell., 2022 (2022), 1-12. https://doi.org/10.1109/TAI.2022.3190816 doi: 10.1109/TAI.2022.3190816
![]() |
[32] |
J. Zeng, S. Tan, G. Liu, B. Li, J. Huang, WISERNet: wider separate-then-reunion network for steganalysis of color images, IEEE Trans. Inf. Forensics Secur., 14 (2019), 2735-2748. https://doi.org/10.1109/TIFS.2019.2904413 doi: 10.1109/TIFS.2019.2904413
![]() |
[33] |
T. Wu, W. Ren, D. Li, L. Wang, J. Jia, JPEG steganalysis based on denoising network and attention module, Int. J. Intell. Syst., 37 (2021), 5011-5030. https://doi.org/10.1002/int.22749 doi: 10.1002/int.22749
![]() |
[34] | J. Kodovsky, J. Fridrich, Steganalysis of JPEG images using rich models, in Proceedings of SPIE - Media Watermarking, Security, and Forensics, 2012. https://doi.org/10.1117/12.907495 |
[35] | W. Feng, B. Hu, Quaternion discrete cosine transform and its application in color template matching, in Congress on Image and Signal Processing, Sanya, (2008), 252-256. https://doi.org/10.1109/CISP.2008.61 |
[36] | V. Holub, J. Fridrich, Designing steganographic distortion using directional filters, in 2012 IEEE International Workshop on Information Forensics and Security (WIFS), (2012), 234-239. https://doi.org/10.1109/WIFS.2012.6412655 |
1. | Olivia X.M. Yao, New congruences modulo 9 for the coefficients of Gordon-McIntosh's mock theta function ξ ( q ) , 2024, 47, 1607-3606, 239, 10.2989/16073606.2023.2205604 | |
2. |
Yueya Hu, Eric H. Liu, Olivia X. M. Yao,
Congruences modulo 4 and 8 for Ramanujan’s sixth-order mock theta function ρ(q) ,
2025,
66,
1382-4090,
10.1007/s11139-024-01018-x
|