Open Access
Add to BibSonomy
Abstract
In this paper, we propose a visual-cryptographic image hiding method based on visual cryptography (VC) and volume grating optics. The secret image is converted to the encrypted visual keys (VKs) according to the normal VC algorithm. Then the VKs are further embellished to the QR-code-like appearance and hidden as the holographic optical elements (HOEs), which are fabricated by the holographic exposure of photopolymer. In the decryption process, the fabricated HOE-VKs are illuminated with the laser beam. The reconstructed VKs are overlaid to extract the hidden information directly in the optical facility, without additional computation. Optical experiments verify that the VKs in HOE mode improves the system with high security and good robustness on the random noise attack. Besides, the volume grating nature also enlarges the system bandwidth by using the multiplexing technique. The proposed method may provide a promising potential for the practical image hiding systems.
© 2018 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
1. Introduction
In recent years, the security issue has become more and more important in our daily life with the information technology development. Various encryption methods have been reported in the literature to secure information [1–13]. Among them, visual cryptography (VC) is known as an effective and simple way to fulfil the encryption process. The basic VC algorithm was firstly proposed by Kafri and Keren [14], and then formalized and extended by Naor and Shamir [15]. In typical VC, the target image is randomly expanded to several binary visual keys (VKs) according to the pixel expansion rule and shared to the users [16,17]. Decryption can be achieved simply by superimposing the shared images, which requires no extra computation, resulting in the convenient extraction. No information about the secret image leaks from each individual shared-key, so the VC algorithm possesses a very high security level.
The VC algorithm has gained a lot development in the past two decades. Many extensions have been proposed and investigated, such as optimization of the contrast [18,19], encryption of gray scale images [20–22] and colorful images [23–25], rotation-multiplexing of the VKs to reveal multi-secret-images [26–28], and improvement of the resolution ratio [29,30]. The VC is also combined with other encryption methods, such as the QR codes. Some researchers proposed QR-code-based VC systems [31–34], in which the picture of the QR code is further split into double VKs and then embedded into the background image. The hidden QR code is extracted by stacking the shared-keys together, and then further information can be obtained from the revealed QR code. However, since the traditional VC ciphertexts are pure-amplitude and random-like images, the VKs are easy to attract the attention of the attackers who then may attack the VKs and make the encryption fail. Although the VKs can also be designed into meaningful images to improve the problem, the decrypted image suffers low quality and poor contrast [35–37].
In our previous work, we have proposed the invisible VC method and developed corresponding optical hiding systems [38,39]. The indirect VC takes advantage of the diffractive optics. The pure-amplitude VKs are substituted with the phase-only VKs and fabricated into diffractive optical elements (DOEs). In this way, the vulnerable weakness of VC can be improved, and the security is enhanced for practical applications.
In this paper, we still inherit the invisible VC concept and propose a low-cost VC image hiding system which utilizes the classical VC and the volume grating optics. The secret image is transformed to the shared VKs according to the conventional VC algorithm and then embellished to QR-code-like appearance for alignment convenience and security improvement. The VKs are further hidden as holographic optical elements (HOEs), which are fabricated by the holographic exposure of photopolymer. In the decryption process, the HOE-VKs are illuminated with laser beam to reconstruct the original VKs and overlaid in the optical path. The hidden information can be extracted directly by human vision, without complex optical facility and additional computation. The HOE-VKs improve the image hiding system with high security and good robustness to random noise attack. Besides, the exposed VKs are essentially volume gratings, thus their propagation complies the Bragg diffraction law, which distinguishes from the DOE-VKs and improves the security further. The holographic fabrication process is also beneficial to decrease the system cost for practical applications.
2. Method
2.1 Principle of the classical visual cryptography
As an effective encryption technique, VC is widely used in the image encoding situations. Its decryption only relies on the human visual system, so even people with no professional background can operate the decryption and extract the secret messages [40,41]. With different security requirements, the VC can also be used as an optical authentication or identification manner [42–44]. Since the digital certification has high-risk security vulnerabilities, the optical certification becomes to a good candidate for the anti-counterfeiting, which makes the VC being a quite hot research spot in recent years.
In the classical VC, the image is converted into binary distributions by pixel expansion. Figure 1(a) gives an example of such expansion, where 2 × 2 condition is considered, and it has six possible choices of P1 to P6. In the encryption, each pixel array of Fig. 1(a) is randomly chosen to conduct the expansion. The white pixels of the secret image are transformed into two identical shared VKs, while the black pixels take the complementary shared VKs, as shown in Fig. 1(b). Such random selection is independent for each pixel, which ensures the high security. The encoded VKs look like messy code, and no information can be read from single VK, as shown in Figs. 1(c) and 1(d). Decryption is achieved by overlaying the shared VKs. Figure 1(e) shows the revealed message of “surprise”.
Fig. 1 (a) Six possible expansions P1-P6 for 2 × 2 case, (b) An example of the encoding process for white and black pixel of the secret image, (c) and (d) The encoded VK1 and VK2, respectively, (e) The revealed message of “surprise”.
Although the classical VC has high security, the feature of the ciphertexts is easy to recognize (please see Figs. 1(c) and 1(d)) and be attacked [38,39]. This vulnerable weakness drags VC for practical application. Thus, how to improve this problem is one of the key research highlights in the VC domain.
2.2 Visual-cryptographic image hiding system
To improve the weakness of VC, we propose a visual-cryptographic image hiding (VCIH) system in this paper, which takes advantage of the classical VC and the volume grating optics. Figure 2 shows its hiding procedure with three steps. At beginning, the message “OK” is implanted into a binary image and used as the secret image to be hidden, and the next three hiding steps are illustrated in detail as follows:
Step 1: Visual cryptography. In this step, the encoding principle of VC is employed to generate a set of VKs, and such process can be achieved by using a normal VC algorithm, as mentioned in the former section. Here we omit the details of the algorithm for the sake of simplicity. If each pixel is expanded with 2 × 2 array, double VKs with mosaic-like distributions will be generated as shown in Fig. 2. According to the VC concept, these VKs are completely unrelated with each other, which ensures the proposed VCIH with firm security footstone.
Step 2: QR code ornament. The VKs are then embellished into QR-code-like appearance. It is just a disguise strategy which takes advantage of the mosaic-like property of the VKs in classical VC algorithm, as can be seen from the blue column and green column of Fig. 2. We add three rectangle marks to the related corners of the VKs and pretend them into QR-code-like appearance. On the one hand, this disguise gives the mosaic-like VKs a certain meaning, so they may lose the attacker’s attention and subsequently improve the system security. On the other hand, it is also helpful for the alignment of VKs in the decryption process. The rectangle marks can offer a function of sight bead to benefit the superposition of the revealed VKs. This operation also improves the efficiency of the VKs optical extraction. Such QR-code ornament is just a kind of cover strategy, which restrains the distinct feature of the VKs. Except for the QR code, some other cover methods can also be used to achieve the disguise purpose. Here we adopt the QR-code because of its simple and easy implementation.
Step 3: Image hiding. We finally conduct the image hiding operation in this step by using the optical holography technique. The embellished VKs are treated as input objects in holography and then fabricated into HOEs of shared-keys. The fabrication is realized by the exposure of photopolymer on glass substrate. Since the recording material usually has a certain thickness, the fabricated HOE-VKs possess volume grating nature. As shown in the red column of Fig. 2, no clues can be seen from the final hidden shared-keys. Owing to the volume optical property, the hidden VKs complies the Bragg diffraction law in the optical extraction, therefore only when the illumination satisfies the preconcerted conditions, the perdue VKs are revealed for stacking decryption. It means that the interference angle, wavelength and polarization of the exposure light can serve as new system keys, so the security is enhanced further.
After the three steps above, the secret information “OK” can be hidden in a group of the fabricated HOE-VKs, which are completely-mutually-unrelated, invisible and system-keys-extended. As a result, the hiding procedure of the proposed VCIH is sufficiently rigorous, leading to a high security for concealing purpose.
2.3 Fabrication of visual keys
The VKs are fabricated in the holographic exposure facility, and Fig. 3 shows its optical schematic. The laser is controlled by an electronic shutter, which serves as a time manager for exposure. Double half-wave plates HWP1 and HWP2 modulate the polarization of the light beam to obtain the best interference result. M1, M2 and M3 stand for the mirror. The object lens (OL), filter and lens constitute the collimation and beam expanding system. Then the beam passes through a polarization beam splitter (PBS) to produce two interference light portions of reference and object. The VK is placed on a translation stage (TS) located at distance of d, so its position is precisely adjusted to fit the illumination zone. In some cases, the TS can also be an electronic intensity modulator, such as the spatial light modulator (SLM). In this way, the VKs are uploaded to the SLM in real-time, so the fabrication efficiency is improved. M3 is fixed on a rotation stage (RS) and controls the incident angle. The VKs are fabricated into HOE mode by exposure of photopolymer on the glass substrate.
The reference beam and object beam in Fig. 3 are at the same side of the photopolymer, thus the fabricated HOE-VKs work in the transmission type [45,46]. If the two beams are at the different side, the case changes to the reflective type. Note that both the two types can be used for fabrication with different optical arrangements, and in this paper, we just employ the transmission type to illustrate the proposed system for the sake of convenience.
Figure 4 shows the detailed illustration of the holographic exposure with transmission type. In the exposure, as shown in Fig. 4(a), the beam with VK is the object light, and the other plane wave light performs as reference function. ko and kr represent the wave vector of the object beam and reference beam, respectively, and the angle between them is θ. The object beam propagates d distance, then intervenes at the photopolymer layer on the glass substrate. Assuming the object distribution is , its propagation can be expressed with Fresnel diffraction [47]:
where is the object wavefront, is the imaginary unit, λ is the wavelength, and is the wave number. In the extraction, we use a light beam with the same parameters of kr to illuminate the HOE, as shown in Fig. 4(b). The reconstructed light ko will be produced at the other side of the HOE, which fits the transmission type, and the VK is revealed at the distance of d. Figure 4(c) illustrates the vector diagram of the fabricated HOE, in which the wave vector of the HOE is written by:where kh stands for the wave vector of the HOE. Since the photopolymer has a certain thickness, the fabricated HOE performs as volume grating and the extraction must satisfy the Bragg condition [45,46]. It means that the extracted illumination should obey the preconcerted parameters, otherwise the reconstruction will fail and no information of the VKs can be obtained. That is to say: the angle θ, wavelength λ, and polarization of the beam serve as system keys. Therefore, the hiding method possesses very high security.Fig. 4 Illustration of the holographic exposure with transmission type, (a) exposure, (b) extraction and (c) vector diagram of the HOE.
The shared VKs are fabricated by the optical holographic exposure of the photopolymer on glass substrate. Such exposure is simple and easy to realize. The designed VKs can be developed into master masks, and then series shared-keys are easily copied for actual delivery. Thus, the system cost may be decreased lower through the repetitive fabrication ability of the master masks. Besides, due to the volume grating nature of the VKs, the system capacity is expanded by using the multiplexing technique, which means that we can fabricate multiple VKs on just single recording plate. In this way, the system cost is reduced once again. The fabrication price of the HOE is lower and the time-consuming is less, which makes the proposed VCIH system be more appropriate for the practical applications. Besides, the size of the HOE-VKs can be extended by use of the holographic print technique.
3. Experimental results and discussions
3.1 Optical fabrication and extraction of the visual keys
Optical experiments are performed to demonstrate the proposed VCIH method. The fabrication is implemented in the optical facility as shown in Fig. 3. The VKs are generated by normal VC algorithm and embellished with QR code appearance. These procedures are illustrated in the step 1 and step 2 of Fig. 2. For the convenience of implementation, we employ the 2 × 2 expansion way, so double VKs are produced in the encryption process. The laser wavelength is 532nm. The photopolymer is sensitive to the illuminating laser, and its tested sensitivity threshold is about 22mJ/cm2. After beam expansion, the beam covers almost 1.5cm2 zone with intensity of 655mW. Therefore, the related exposure time is about 50.4ms. Considering some system errors, we finally set the exposure time to 55ms managed by the electronic shutter.
The two generated VKs are developed by aid of the chrome plated printing technique, and Fig. 5(a) shows a printed example. We use it as objects in the holographic exposure. In fact, the VKs can be also expressed by uploading to an electronic intensity modulator, such as the amplitude type SLM. Note that the given VKs above are all pure-amplitude distributions, thus their diffraction patterns only spread in small angles and this phenomenon influences the resistance for occlusion attack of VKs. If we add random phase element to the object path, this problem can be improved. Figures 5(b) and 5(c) show two fabricated HOE-VKs. From them, we can find no information of the VKs, which matches the invisible VC concept. This hiding operation removes the vulnerable weakness of the VC and enhances the system security degree.
Fig. 5 (a) Printed VKs as objects in the holographic exposure, (b) and (c) The fabricated two HOE-VKs.
Decryption of the proposed method can be achieved by the optical extraction device, and Fig. 6 gives the optical extraction facility map. The laser wavelength is the same with the recording light (532nm). The polarization and the intensity of the light are controlled by the HWP and the attenuator respectively. The illumination beam is obtained by a beam expander (BE) and a lens. An aperture stop regulates the size of the light beam. Double beam splitters (BS) are used to supply the illumination for double VKs (VK1 and VK2) and then overlay them to accomplish the decryption. A camera is employed to record the revealed image. M1, M2, M3 and M4 are the mirrors to modulate the light path.
One of the good properties of VC is that the decryption only needs to stack the shared-VKs, and the secret message can be read directly by human visual system. Figures 7(a) and 7(b) are the double VKs extracted from the fabricated HOE-VKs, and they have good contrast quality. As is discussed before, the QR-code-like ornament has the function of alignment. The rectangles at the three corners of the VK are beneficial to the superposition in the decryption. After carefully adjustion, the revealed image of “OK” is captured by the camera and shown in Fig. 7(c). From the experimental results, we can see that the decoded “OK” is clear and recognizable directly for naked eyes, which demonstrates the proposed method successfully. It should be pointed out that the reconstructed VKs have some distortions, especially at the edge area, due to the inhomogeneous thermal contraction in the postprocessing of the HOE-VKs. This problem affects the decoded quality. But we can optimize the fabrication procedure to improve the distortion issue.
Fig. 7 (a) and (b) The extracted results from the fabricated HOE-VKs, (c) Revealed image of “OK” by overlaying the two VKs.
3.2 Robust analysis
As an effective image hiding method with high security, the robustness is the other important aspect to be considered in real applications. For most traditional hiding systems, the security and robustness restrict each other, which is to say that high security usually needs strict realization of poor robustness, and vice versa. However, in our proposed method, such contradiction can be balanced by the combination of the classical VC and the holographic fabrication. On the one hand, the VC algorithm ensures the VCIH system a very high security level. As mentioned in the previous sections, the shared-keys are completely unrelated with each other, thus no information about the secret image leaks from each individual shared-key. On the other hand, the fabrication of the VKs is essentially based on the optical holography. As a powerful optical processing technique, the optical holography has strong adaptation to the illumination, which means that the VCIH system theoretically has good robustness on multiform noise attack. In fact, the security comes from the intrinsic property of the VC algorithm, while the robustness is balanced by the holographic optics. Therefore, the proposed system has high security and good robustness at the same time.
To analyze the robust performance of VCIH, we add noise to the fabricated VKs and then test the decrypted results in optical experiments. Although the noise has multiple types, such as white noise, salt-pepper noise and random noise etc., we mainly chose the random noise to analyze, because it is typical and common in the practical applications. We smear some dust as random noise to the shared VKs and perform the decryption again in the same optical extraction path of Fig. 6. The VKs with dust are shown in Figs. 8(a) and 8(b). From them, we can see that the dust-noise covers all areas of the VKs and the dust pollution is severe enough to be observed by eyes directly.
Figures 9(a) and 9(b) are the extracted VKs with dust-noise pollution, and Fig. 9(c) is the revealed image of “OK” by superimposing them together. Compared with the noiseless status, the reconstructed VKs mingle with some speckles and the background contrast is influenced as well. The line edges of the VKs turn to be blurred to a certain degree, especially for the VK2 of Fig. 9(b), that is mainly because the dusts on VK2 have bigger particle size. Although the noise is severe enough, the decoded message “OK” still can be recognized by the eyes. These results demonstrate that the proposed VCIH method is very robust to the random noise pollution, which is quite suitable for practical applications.
Fig. 9 (a) and (b) are the reconstructed VKs with dust-noise pollution, (c) is the revealed image of “OK”.
Secondly, we also tested the proposed VCIH with occlusion attack. As is mentioned before, the generated VKs are printed on the transparent glass substrate (see Fig. 5(a)) by aid of the chrome plated printing technique. The phase distributions of the VKs are basically constant, which means that their diffracted patterns diffuse in a small angle, if we use the so-printed VKs as the object in the interference exposure. Therefore, the fabricated HOE-VKs are sensitive to the block attack in fact. We block 40% areas of the HOE-VKs, as shown in Figs. 10(a) and 10(b). The corresponding rebuilt VKs are then given in Figs. 10(c) and 10(d), respectively. As can be seen, some information is lost in the recorded images, so the decoded message “OK” shown in Fig. 10(e) is also incomplete.
Fig. 10 Occlusion attack analysis, (a) and (b) are the HOE-VKs with 40% zone occlusion, (c) and (d) are the reconstructed two VKs, (e) is the decoded image of “OK”.
This foible of VCIH originally comes from the pure-amplitude property of the printed VKs. However, if we add random phase part (for example, the random phase distributed in [0, 2π] to express the diffuse reflection effect) to the VKs in the fabrication process, this problem can be solved to improve the resistance on occlusion attack. Actually, in our previous work, we have applied the random phase to the fabrication of the DOE-VKs and the results showed that the decryption is robust to the occlusion attack [38,39]. Since the main purpose of this paper is to demonstrate the validity of the proposed VCIH, this extended research will be considered in our future study steps.
3.3 Discussions
The proposed VCIH method employs the advantages of the classical VC and the holographic fabrication technique. In this way, the visible shared-keys are hidden as invisible ones, which inherits the invisible VK concept described in our prior work. Therefore, the corresponding encryption system obtains a higher security degree.
The VKs are fabricated as HOEs by the exposure of photopolymer on glass substrate. Typically to say, the recorded HOEs have a certain thickness (possibly from tens micrometers to sub-millimeters, depended on the specific type of the recording material). Therefore, the fabricated VK basically belongs to a kind of volume holographic grating, according to the volume optics concept. Only the illuminated light matches the specific incident angle, polarization and wavelength, the hidden VKs can be extracted from the HOE plates and then stacked to reveal the encoded image. This property makes the shared VKs look transparent with invisible effect. Besides, the volume grating nature also brings additional benefits to the encryption system: the multiplexing technique can be used to expand the system bandwidth [45], which means we can hide multiple VKs in only one HOE plate. Figures 11(a), (b) and 11(c) depicts the multiplexing ways of incident angles, polarization and wavelength, respectively. In Fig. 11(a), the light beams kr1, kr2 and kr3 with incident angles of β1, β2 and β3 intervene with the object beam ko of VK1, VK2 and VK3, respectively. In this way, we hide three VKs in one HOE, so the bandwidth is expanded by three times. Theoretically, the multiplexed angles are related to the thickness of the exposure material, so the number of the angles can be increased more with the thickness raise. In Figs. 11(b) and 11(c), the multiplexing factor is switched to the polarization and wavelength, respectively. The expanded bandwidth improves the delivery efficiency of the shared-VKs.
Fig. 11 Bandwidth expansion with multiplexing technique of (a) incident angles, (b) polarization, and (c) wavelength.
In step 2 of the decryption process (see Fig. 2), the VKs is embellished to the QR -code-like appearance. Actually, we adopt this operation to decrease the attention of the potential attackers and assist alignment in the overlaying procedure. In the optical experiments, although the VCIH has strong robustness on the random noise pollution, it is sensitive to the occlusion attack. The reason for this problem is mainly attributed to the pure-amplitude distribution of the printed VKs. If we apply the random phase part to enhance the diffusion, the robustness can be improved to resist the occlusion situations. Besides, the electrical intensity modulator (such as SLM) can be further employed in the optical fabrication process. In this way, the printed VKs are replaced by electrical images, so the fabrication efficiency is improved.
4. Conclusion
We have proposed and developed a low-cost VCIH system based on the classical VC and the volume holographic optics. The encryption process of the VCIH contains three steps. Firstly, the normal VC algorithm is used to convert the secret image “OK” to the encrypted VKs. Then, the VKs are further disguised with QR-code-like appearance to lower the vulnerability of the VC and benefits to the VKs alignment for secret message extraction. The generated VKs are finally hidden as HOEs fabricated by the holographic exposure of photopolymer. In this way, the visible VKs are transformed into invisible shared ones, which avoids the vulnerable weakness of the conventional VC. We have also built the optical extraction facility to test the decryption quality. The secret message “OK” can be revealed by simply overlaying the reconstructed VKs directly in the optical path, without additional computation. The HOE-VKs improve the image hiding system with high security and good robustness to random noise attack. Besides, the system bandwidth can be expanded to improve the delivery efficiency, thanks to the volume grating nature of the fabricated VKs. The proposed method may provide a promising potential for the practical image hiding systems.
This preliminary study encourages us to continue the exploration for advancing the encryption process and optimizing the optical fabrication quality. The multiplexing ability of angle, polarization and wavelength will be investigated to improve the system capacity in our next research schedule. We will also focus on the robustness issue and employ the random phase element to enhance the resistance on occlusion attack.
Funding
National Natural Science Foundation of China (61575197); Fusion Foundation of Research and Education of CAS; Youth Innovation Promotion Association CAS (2017489).
Acknowledgments
The authors thank the reviewers for giving the pertinent comments, valuable questions and constructive suggestions on this work.
References
1. M. Khan and T. Shah, “A literature review on image encryption techniques,” 3D Res. 5(4), 29–1722 (2014). [CrossRef]
2. W. Chen, B. Javidi, and X. Chen, “Advances in optical security systems,” Adv. Opt. Photonics 6(2), 120–155 (2014). [CrossRef]
3. S. Liu, C. Guo, and J. T. Sheridan, “A review of optical image encryption techniques,” Opt. Laser Technol. 57, 327–342 (2014). [CrossRef]
4. Y. Shi, G. Situ, and J. Zhang, “Multiple-image hiding in the Fresnel domain,” Opt. Lett. 32(13), 1914–1916 (2007). [CrossRef] [PubMed]
5. Y. Shi, G. Situ, and J. Zhang, “Multiple-image hiding by information prechoosing,” Opt. Lett. 33(6), 542–544 (2008). [CrossRef] [PubMed]
6. Y. Shi, T. Li, Y. Wang, Q. Gao, S. Zhang, and H. Li, “Optical image encryption via ptychography,” Opt. Lett. 38(9), 1425–1427 (2013). [CrossRef] [PubMed]
7. Q. Gao, Y. Wang, T. Li, and Y. Shi, “Optical encryption of unlimited-size images based on ptychographic scanning digital holography,” Appl. Opt. 53(21), 4700–4707 (2014). [CrossRef] [PubMed]
8. A. Chanana, A. Paulsen, S. Guruswamy, and A. Nahata, “Hiding multi-level multi-color images in terahertz metasurfaces,” Optica 3(12), 1466–1470 (2016). [CrossRef]
9. W. Xu, H. Xu, Y. Luo, T. Li, and Y. Shi, “Optical watermarking based on single-shot-ptychography encoding,” Opt. Express 24(24), 27922–27936 (2016). [CrossRef] [PubMed]
10. T. Shimobaba, Y. Endo, R. Hirayama, D. Hiyama, Y. Nagahama, S. Hasegawa, M. Sano, T. Takahashi, T. Kakue, M. Oikawa, and T. Ito, “Holographic microinformation hiding,” Appl. Opt. 56(4), 833–837 (2017). [CrossRef] [PubMed]
11. D. Kong, X. Shen, L. Cao, and G. Jin, “Phase retrieval for attacking fractional Fourier transform encryption,” Appl. Opt. 56(12), 3449–3456 (2017). [CrossRef] [PubMed]
12. F. Xu, G. Shulkind, C. Thrampoulidis, J. H. Shapiro, A. Torralba, F. N. C. Wong, and G. W. Wornell, “Revealing hidden scenes by photon-efficient occlusion-based opportunistic active imaging,” Opt. Express 26(8), 9945–9962 (2018). [CrossRef] [PubMed]
13. L. Zhang, X. Yuan, D. Zhang, and J. Chen, “Research on multiple-image encryption scheme based on Fourier transform and ghost imaging algorithm,” Curr. Opt. Photon. 2(4), 315–323 (2018).
14. O. Kafri and E. Keren, “Encryption of pictures and shapes by random grids,” Opt. Lett. 12(6), 377–379 (1987). [CrossRef] [PubMed]
15. M. Naor and A. Shamir, “Visual cryptography,” Lect. Notes Comput. Sci. 950, 1–12 (1995). [CrossRef]
16. S. P. Kannojia and J. Kumar, “Role of visual cryptography schemes in information security: A review,” Int. J. Appl. Eng. Res. 13(13), 11340–11343 (2018).
17. S. Jiao, C. Zhou, Y. Shi, W. Zou, and X. Li, “Review on optical image hiding and watermarking techniques,” Opt. Laser Technol. 109, 370–380 (2019). [CrossRef]
18. C. Blundo, A. D. Bonis, and A. D. Santis, “Improved schemes for visual cryptography,” Des. Codes Cryptogr. 24(3), 255–278 (2001). [CrossRef]
19. S. Cimato, A. De Santis, A. L. Ferrara, and B. Masucci, “Ideal contrast visual cryptography schemes with reversing,” Inf. Process. Lett. 93(4), 199–206 (2005). [CrossRef]
20. C. Blundo, A. De Santis, and M. Naor, “Visual cryptography for grey level images,” Inf. Process. Lett. 75(6), 255–259 (2000). [CrossRef]
21. C. Lin and W. Tsai, “Visual cryptography for gray-level images by dithering techniques,” Pattern Recognit. Lett. 24(1–3), 349–358 (2003). [CrossRef]
22. R. Lukac and K. Plataniotis, “Bit-level based secret sharing for image encryption,” Pattern Recognit. 38(5), 767–772 (2005). [CrossRef]
23. Y. Hou, “Visual cryptography for color images,” Pattern Recognit. 36(7), 1619–1629 (2003). [CrossRef]
24. H. Yamamoto, Y. Hayasaki, and N. Nishida, “Secure information display with limited viewing zone by use of multi-color visual cryptography,” Opt. Express 12(7), 1258–1270 (2004). [CrossRef] [PubMed]
25. J. Machizaud and T. Fournel, “Two-out-of-two color matching based visual cryptography schemes,” Opt. Express 20(20), 22847–22859 (2012). [CrossRef] [PubMed]
26. H. Wu and C. Chang, “Sharing visual multi-secrets using circle shares,” Comp. Stand. Inter. 28(1), 123–135 (2005). [CrossRef]
27. J. Feng, H. Wu, C. Tsai, Y. Chang, and Y. Chu, “Visual secret sharing for multiple secrets,” Pattern Recognit. 41(12), 3572–3581 (2008). [CrossRef]
28. A. Mishra and A. Gupta, “Multi secret sharing scheme using iterative method,” J. Inf. and Optim. Sci. 39(3), 631–641 (2018). [CrossRef]
29. C. Blundo, S. Cimato, and A. De Santis, “Visual cryptography schemes with optimal pixel expansion,” Theor. Comput. Sci. 369(1–3), 169–182 (2006). [CrossRef]
30. Y. Chen, Y. Chan, C. Huang, M. Tsai, and Y. Chu, “A multiple-level visual secret-sharing scheme without image size expansion,” Inf. Sci. 177(21), 4696–4710 (2007). [CrossRef]
31. W. Fang, “Offline QR code authorization based on visual cryptography,” in Proceedings of 2011 Seventh International Conference on Intelligent Information Hiding and Multimedia Signal Processing, (IEEE, 2011), pp. 89–92. [CrossRef]
32. S. Thamer and B. Ameen, “A new method for ciphering a message using QR code,” Comput. Sci. Eng. 6(2), 19–24 (2016).
33. X. Cao, L. Feng, P. Cao, and J. Hu, “Secure QR code scheme based on visual cryptography,” in Proceedings of 2nd International Conference on Artificial Intelligence and Industrial Engineering, (Atlantis, 2016), pp. 433–436.
34. J. Lu, Z. Yang, L. Li, W. Yuan, L. Li, and C. Chang, “Multiple schemes for mobile payment authentication using QR code and visual cryptography,” Mob. Inf. Syst. 2017, 4356038 (2017). [CrossRef]
35. M. Nakajima and Y. Yamaguchi, “Extended visual cryptography for natural images,” J. WSCG 10(2), 303–310 (2002).
36. Z. Zhou, G. R. Arce, and G. Di Crescenzo, “Halftone visual cryptography,” IEEE Trans. Image Process. 15(8), 2441–2453 (2006). [CrossRef] [PubMed]
37. Z. Wang, G. R. Arce, and G. Di Crescenzo, “Halftone visual cryptography via error diffusion,” IEEE Trans. Inf. Forensics Security 4(3), 383–396 (2009). [CrossRef]
38. Y. Shi and X. Yang, “Optical hiding with visual cryptography,” J. Opt. 19(11), 115703 (2017). [CrossRef]
39. Y. Shi and X. Yang, “Invisible visual cryptography,” Chin. Phys. Lett. 34(11), 114204 (2017). [CrossRef]
40. H. Yamamoto, Y. Hayasaki, and N. Nishida, “Securing information display by use of visual cryptography,” Opt. Lett. 28(17), 1564–1566 (2003). [CrossRef] [PubMed]
41. J. Machizaud, P. Chavel, and T. Fournel, “Fourier-based automatic alignment for improved Visual Cryptography schemes,” Opt. Express 19(23), 22709–22722 (2011). [CrossRef] [PubMed]
42. S. Jaya, Malik, A. Aggarwal, and A. Sardana, “Novel authentication system using visual cryptography,” in Proceedings of 2011 World Congress on Information and Communication Technologies, (IEEE, 2011), pp. 1181–1186. [CrossRef]
43. G. Wang, W. Yan, and M. Kankanhalli, “Content based authentication of visual cryptography,” Multimedia Tools Appl. 76(7), 9427–9441 (2017). [CrossRef]
44. V. Daisy, C. Joe, and S. Sugi, “An image based authentication technique using visual cryptography scheme,” in Proceedings of 2017 International Conference on Inventive Systems and Control (ICISC), (IEEE, 2017), pp. 1–6. [CrossRef]
45. S. Vyas, P. H. Wang, and Y. Luo, “Spatial mode multiplexing using volume holographic gratings,” Opt. Express 25(20), 23726–23737 (2017). [CrossRef] [PubMed]
46. L. Puyo, J. P. Huignard, and M. Atlan, “Off-axis digital holography with multiplexed volume Bragg gratings,” Appl. Opt. 57(12), 3281–3287 (2018). [CrossRef] [PubMed]
47. J. W. Goodman, Introduction to Fourier optics (Robert & Company Publishers, 2005).
