Show Summary Details
More options …

# Nanophotonics

Editor-in-Chief: Sorger, Volker

CiteScore 2017: 6.57

IMPACT FACTOR 2017: 6.014
5-year IMPACT FACTOR: 7.020

In co-publication with Science Wise Publishing

Open Access
Online
ISSN
2192-8614
See all formats and pricing
More options …

# Nanoscale hierarchical optical interactions for secure information

Naoya Tate
/ Makoto Naruse
• National Institute of Information and Communications Technology, 4-2-1 Nukui-kita, Koganei, Tokyo 184-8795, Japan
• Other articles by this author:
Published Online: 2016-12-28 | DOI: https://doi.org/10.1515/nanoph-2016-0134

## Abstract

There is increasing demand for novel physical security that can differentiate between real and false specific artifact that have been added to bank bills, certifications, and other vouchers. The most simple and effective method for improving the security level is to scale down the elemental structures so that they cannot be duplicated by attackers. While there is a paradox that the achieved fabrication resolution by a defender can also be realized by an attacker, further improvement in security is possible by the functional fusion of artifact metrics and nanophotonics. The fundamental advantages of this concept are the high-level clone resistance and individuality of nanoscale artifacts, which are based on the super-resolution fabrication and nanoscale hierarchical structure of optical near-field interactions, respectively. In this paper, the basis for the fabrication of nanoscale artifacts by utilizing random phenomena is described, and a quantitative evaluation of the security level is presented. An experimental demonstration using a nano-/macro-hierarchical hologram is presented to demonstrate the fundamental procedure for retrieving nanoscale features as hidden information. Finally, the concept and a simple demonstration of non-scanning probe microscopy are described as a practical application of the retrieval and authentication of nanoscale artifact metrics.

## 1 Introduction

Despite rapid advances in information security and the corresponding active computerization for various applications, there is still a high demand for the development of novel physical security for secure authentication. Optical information security is one of the most active research fields for both physical security and various forms of optical computing. Various algorithms and related technologies have recently been applied to information security for data encryption [1], [2], [3], [4], [5], [6], [7], [8], [9], [10], information hiding [11], [12], [13], [14], watermarking [15], [16], and authentication [17], [18], [19], [20], [21], [22], [23]. These proposals effectively utilize two-dimensional imaging capabilities and parallelism, which are known as the fundamentals of optical computing. Especially in the case of authentication, artifact metrics [24] utilize the physical features of individual artifacts as hidden information. Generally, several performance indicators on each artifact are utilized as artifact metrics, such as the clone resistance, individuality, measurement stability, and durability. Under such conditions, the defender, who wants to prevent counterfeiting, must fabricate fine-structured artifacts such that the attacker, who wants to copy the authentic artifact, will not be able to simply reproduce the artifacts from the obtained appearance.

In optical implementation of artifact metrics, such physical features of artifacts are simply represented by their structural patterns. In view of recent technological advances, the fabrication of micro-patterns is one of the most promising approaches to realizing fine-structured artifacts. At the same time, advances in conventional fabrication, which is typified by electron-beam lithography, have been technologically limited to a fabrication resolution of the submicron scale. The use of conventional fabrication has become popular. At present, it is difficult to ensure an appropriate security level for fine-structured artifacts fabricated by conventional techniques. Therefore, not only the technological development of fabrication techniques but also the contribution of novel concepts are greatly needed for further development of artifact metrics.

The functional fusion of nanophotonics with optical information security [25], [26] is being actively discussed. Nanophotonics is a novel optical technology that utilizes the local interactions between nanometric structures via optical near fields [27], [28], [29]. The optical near fields behave as a virtual cloud of photons that is constantly localized around nanometric structures illuminated by incident light. Because the virtual cloud of photons is localized in a region close to the electrons in the material, they can effectively interact with the structures in a unique manner [30]. Based on the fundamentals of nanophotonics, closed multiple nanometric structures reveal strong interactions via individual optical near fields. The results of such interactions are represented by the scattering of propagating light and can simply be detected as a macro-optical signal. For the implementation of nanoscale artifact metrics [31], the physical features of a nanoscale artifact should be obtained by a nanoscale retriever as a specific macro-optical signal via optical near-field interactions between the artifact and retriever. Because such signals necessarily reveal specific individuality not only about the artifact but also about the retriever from the position of the attacker, authentic signals are fundamentally difficult to reproduce. This theoretically and technologically assures the clone resistance and individuality of the artifact and retriever. Based on these factors, nanoscale artifact metrics require consideration and can be expected to lead to the realization of novel optical information security technologies that will be essential for a sophisticated information society in the future. In this paper, this research concept is defined as nanophotonics-based optical information security. The basics are briefly introduced, and the results of some experimental demonstrations are presented.

## 2 Nanoscale artifact metrics

Generally, a smaller scale of fabrication is preferred for higher levels of artifact metrics. Actually, there have been many recent proposals to apply advanced optical technologies and nanotechnologies to artifact metrics to accommodate social demands [32], [33], [34], [35], [36]. However, the general situation of artifact metrics is paradoxical in that novel technologies to realize state-of-the-art fine structures for authentic artifacts can also be utilized by attackers to clone similar artifacts. Here, cloning is defined as the illegal duplication of authentic artifacts for abuse. To overcome this paradox, controlled randomness has been exploited for novel fabrication. This approach realizes much smaller structures than the resolution limit of any conventional fabrication techniques, and at the same time, it is nearly impossible to clone a similar artifact, even by the defender who designed and fabricated the artifact. Here, a well-known phenomenon is utilized as a basic technique: the random collapse of resist patterns [37]. Resist collapse inevitably occurs during the rinse process of fabrication by electron-beam lithography; a schematic is shown in Figure 1B. Because this phenomenon depends on the pattern resolution, resist thickness, and duration of electron-beam exposure, appropriate electron-beam lithography conditions must be applied to avert resist collapse and produce a desired resist pattern. However, from the standpoint of nanoscale artifact metrics, resist collapse occasionally provides structures finer than the fabrication limitations of electron-beam lithography. Furthermore, the randomness of resist collapse is surely almost impossible to replicate by anyone else and any other fabrication technique.

Figure 1:

(A) SEM image of collapsed resist pillars. (B) Common procedure for electron-beam lithography. (C) Entire region of the fabricated silicon pattern based on the random collapse of resist pillars and (D) a magnified view of an area of the pattern. Source: Adapted with permission from Macmillan Publishers Ltd: Scientific Reports [31].

Figure 1A shows a cross-sectional view of randomly collapsed resist pillars. Each pillar has a cross-sectional area of 60 nm×60 nm and height of 200 nm. Before the collapse, they were aligned on a grid of 120 nm×120 nm2. Under such conditions, the random collapse of resist pillars occurred during the rinse process to remove unexposed resists. The sample was then subjected to several post-processing steps, and a nanostructured silicon pattern was finally obtained. Figure 1C and D shows scanning electron microscopy (SEM) images of the pattern. As shown in Figure 1D, the structural details of the method were as small as 9.23 nm, which is much finer than the general fabrication limitations of electron-beam lithography. Here, 2383 samples were fabricated to quantitatively evaluate their specifications for security applications.

In the evaluation, the false match rate (FMR) and false non-match rate (FNMR) were calculated using SEM images of the 2383 samples. The FMR is defined as the rate of false decisions including the acceptance of non-authorized artifacts, and the FNMR is the rate of false decisions including the rejection of identical artifacts. These factors quantitatively indicate the individuality and measurement stability of patterns, respectively. The similarity between two patterns is given by:

$R=∑i∑j[A^(i,j)−A^¯][B^(i,j)−B^¯]∑i∑j[A^(i,j)−A^¯]2[B^(i,j)−B^¯]2,$(1)

where $\overline{\stackrel{^}{A}}\text{\hspace{0.17em}}\text{and}\text{\hspace{0.17em}}\overline{\stackrel{^}{B}}$ indicate the averages of patterns A(i, j) and B(i, j), respectively. If R is greater than the given threshold, the two patterns are considered to be similar to each other, which is extremely likely to be a false decision. The 2383 samples were used to compare 2383×2382 pairs with each other when calculating the FMR for threshold values between 0 and 1. In a false decision, more than one case among the 2383×2382 comparisons would be similar. In this case, the error rate would be calculated as 1/(2383×2383)≈1.76×107. The leftmost curve in Figure 2 shows the resulting FMR. The FMR dropped below the error rate of 106 when the threshold was just above zero, which indicates that the occurrence of false decisions among all 2383×2382 comparisons was extremely low. Next, to calculate the FNMR, 100 images were obtained from 74 samples. If the similarity is less than the given threshold, the two compared images would be considered different from each other. In other words, in the case of a false decision, identical samples would be considered different. The rightmost curve in Figure 2 shows the resulting calculated FNMR. The FMR and FNMR curves are well separated from each other, which means that it is possible to obtain sufficiently low FMR and FNMR by choosing appropriate threshold values.

Figure 2:

Quantitative evaluation of the security performance of nanoscale artifact metrics according to the FMR, FNMR, and CMR. Source: Adapted with permission from Macmillan Publishers Ltd: Scientific Reports [31].

The clone match rate (CMR) was calculated for further evaluation. The CMR is defined as the rate of false matching by a cloned sample. To calculate the CMR, the attacker was assumed to be able to obtain an authentic artifact. This would allow fabrication of the artifact, and the average over every k×k pixel area of the cloned artifacts would essentially be equivalent to the authentic artifact. To quantify such a cloning process by the attacker, each of the 2383 images was transformed into a virtually cloned binary image. The pixel values within a k×k area were denoted as p(i, j), where 0≤p(i, j)≤255. If the averaged pixel value in this area is higher than the threshold T, the area is defined as having a higher average value. Thus, this pixel value of the cloned artifact was defined as p′(i, j)=130. If the averaged pixel value in the k×k area is less than or equal to the threshold T, the area is defined as having a lower average value. Thus, this pixel value of the cloned artifact was defined as p′(i, j)=80. The cloned artifact was compared with the authentic artifact using Eq. (1). If the similarity R exceeds the threshold value, the clone surely mimics the original. The CMR was calculated by performing the above evaluation for all 2383 samples. Here, we needed to consider the correspondence between the unit tile size of the virtually cloned artifact and the actual size. For example, if k=3, a 3×3 pixel area corresponds to a 10 nm×10 nm square area because one pixel occupies approximately a 3.3-nm2 area. This is generally regarded as state-of-the-art for current nanofabrication technologies. Nevertheless, the similarity for a unit tile size of 10 nm was less than 0.4, as indicated by the CMR curve in Figure 2. Considering that the CMR was lower than the similarity between patterns of identical artifacts, such a clone can be considered as a nonthreatening matter. Based on the results of the FMR, FNMR, and CMR analyses, it can be concluded that nanostructured random patterns realized by the random collapse of resist pillars provide a sufficiently high level of performance as nanoscale artifact metrics.

The large capacity of identities based on such nanoscale artifact metrics has been quantitatively evaluated with eigenanalysis [38]. Matsumoto et al. developed image retrieval methods of such nanoscale morphology for physical security that exploit the fine-scale precision ability of confocal laser microscopy [39].

## 3 Nano-/macro-hierarchical hologram

Overtness and covertness are fundamental concepts of optical information security. The former means showing something openly to be simply recognized by the human eye, and the latter needs specific equipment to be recognized. For instance, owing to the diversity of coding rules for optical information processing, confidential optical information can be hidden in any of the physical attributes of light, such as the phase, wavelength, spatial frequency, or polarization, so that simply one optical component is implemented for anti-counterfeiting [40], [41], [42]. Based on this concept, functional co-implementation of optical far fields and near fields seems to realize higher-level optical information security, while optical near-field interactions are fundamentally independent of the behavior of propagating light. Because the existence of nanoscale artifacts and the corresponding optical near fields does not affect the behavior of propagating light, lossless co-implementation of overtness due to optical near fields and covertness due to propagating light is available. This characteristic concept has been experimentally demonstrated in nano-/macro-hierarchical holograms [43], [44], where independent functions are associated with both optical near and far fields in a single device. Figure 3 shows the schematic concept of implementing a nano-/macro-hierarchical hologram.

Figure 3:

Schematic diagram for the concept of a nano-/macro-hierarchical hologram with a nanophotonic code embedded within the embossed structure of the hologram.

A sample device was developed based on the embossed hologram Virtuagram®, which was commercially developed by Dai Nippon Printing Co., Ltd., Japan. It is a high-definition computer-generated hologram composed of binary-level one-dimensional modulated gratings. Generally, the physical scale of the elemental periodic structure of the hologram is greater than 100 nm to induce a predefined diffraction of light. We additionally embedded slightly modified nanometric structures within the structure by electron-beam lithography and then sputtered a 50-nm-thick Au layer. As described above, in principle, a structural change occurring at the subwavelength scale does not affect the overt function, which is dominated by propagating light. Therefore, the visual aspects of the hologram were not affected by the additional induction of the covert function. Figure 4 shows the embedded nanometric structures. For comparison, they were embedded both outside and inside the periodic structures. The unit size of each structure ranged from 40 nm to 160 nm.

Figure 4:

SEM images of embedded nanometric structures inside the periodic structures. Source: Adapted with permission from The Optical Society: Optics Express [44].

As a practical demonstration, the retrieved results for the spatial distribution of optical near fields were defined as nanophotonic codes. Because optical near fields behave as localized energy fields, their existence cannot be recognized by conventional optical methods, which only detect propagating light. Therefore, as a basic mechanism of scanning near-field microscopy (SNOM), the induced optical near fields must be scattered by interactions between the tip of the closed scanning probe. This is schematically indicated in Figure 3. During retrieval, the SNOM was operated in illumination-collection mode with a SiO2-fiber probe having a tip with a radius of curvature of 5 nm. The tip was sharpened by an original etching process using an HF solution. The probe was connected to a tuning fork to sense the shear force between the probe and sample device. The vertical position of the probe was finely regulated by referring to the sensed shear force, which was fed back to a piezoelectric actuator of the probe stage. The observation distance between the probe and sample device was set to be less than 50 nm. The light source for inducing optical near fields was a laser diode with an operating wavelength of 785 nm, and scattered light was detected by using a photomultiplier tube. To verify the polarization dependency of the retrieval, a Glan-Thompson polarizer with an extinction ratio of 10−6 was set to irradiate only linearly polarized light, and the polarization was rotated by a half-wave plate.

Figure 5A and B shows the retrieved results for the nanophotonic codes that were outside and inside of the periodic structures, respectively. Linearly polarized light that was rotated from 0° to 180° at 20° intervals was used for the retrieval. Although Figure 5A shows that small and noisy intensity distributions were obtained, Figure 5B shows clear polarization dependence. For example, from the area of the nanophotonic code located in the center, a high-contrast signal intensity distribution was obtained with polarization around 80°.

Figure 5:

Observed SNOM images as nanophotonic codes located (A) outside and (B) inside the periodic structures. Source: Adapted with permission from The Optical Society: Optics Express [44].

To quantitatively evaluate the characteristics of the retrieved nanophotonic codes, an original parameter was defined. The recognizability Rexp, which quantitatively indicates the visibility of each code in SNOM images is given by

$Rexp=∑x|I(x)−〈I(x)〉env|,$(2)

where x and I(x) represent the horizontal position and horizontal intensity profile, respectively, along the dashed line in Figure 5B, which crosses the area of the nanophotonic code. While ${〈I\text{(}x\text{)}〉}_{\text{env}}$ represents the average intensity of the environmental signal distribution, as shown in Figure 5B, the difference between I(x) and ${〈I\text{(}x\text{)}〉}_{\text{env}}$ indicates the visibility of the nanophotonic code. Figure 6 shows the calculated Rexp as a function of the input light polarization.

Figure 6:

Experimentally obtained recognizability Rexp corresponding to the embedded nanoscale structure (red circles) and isolated nanoscale structure (blue circles). Source: Adapted with permission from The Optical Society: Optics Express [44].

The nanophotonic code embedded in the periodic structures exhibited much greater polarization dependency, as indicated in Figure 5B; the maximum Rexp was obtained at 80° input polarization. In contrast, only slight polarization dependency was observed with the isolated nanophotonic code, as indicated in Figure 5A. These characteristics of retrieving the nanophotonic code in an environmental periodic structure can be explained by the basic mechanism of inducing optical near fields. In other words, when the polarization of the excitation is orthogonal to the periodic structure, the induced energy fields are localized and enhanced at the edges of structures and are obtained as enhanced signal distribution. However, when the polarization is parallel to the periodic structures, the induced energy fields subsequently dissipate along the edge of structures and are difficult to obtain. The results revealed the polarization dependency of the retrieval, as shown in Figure 6.

From the viewpoint of practical application of these concepts as nanoscale artifact metrics, the nanophotonic code is defined not only by the embedded nanoscale structure and corresponding optical near fields but also by specifications of the retrieval, such as the irradiated light and tip of the scanning probe. This means that it is fundamentally difficult for the attacker to associate and duplicate the authentic artifact and retriever from a nanophotonic code and vice versa. Moreover, several novel features of nanophotonics, such as energy transfer [25] and hierarchy [45], may be exploited to achieve further functional improvements of the nanophotonic codes.

## 4 Non-scanning nanoretrieval

As described in the previous section, SNOM has been widely utilized as the most typical detector of optical near fields. However, for practical application as a retriever in nanoscale optical information security, the large system scale of SNOM is a critical disadvantage. In other words, the observation of a high-resolution two-dimensional image of spatially distributed optical near fields with a long processing time is not necessarily required for authentication because only the individual physical features of each artifact are required. In this section, the concept of non-scanning nanoretrieval [46] based on SNOM is described, and the results of some experimental demonstrations are presented.

During the retrieval process with SNOM, the probe is kept at a constant distance from the target by employing an autonomous control mechanism via a feedback system that utilizes the sensed shear force and optical response. Owing to this mechanism, there are inevitable fluctuations in the probe position. Based on the concept of the inherent hierarchy of optical near-field interactions [47], interactions at various spatial scales are necessarily induced during these fluctuations. In other words, a fluctuating probe induces various scales of interactions with the target. As a result, SNOM sequentially observes compressed information [48], [49], [50] on the spatial patterns of the target by fluctuations at a single retrieving point without any scanning process. Figure 7 schematically summarizes the basics of non-scanning nanoretrieval.

Figure 7:

Basic procedure of non-scanning nanoretrieval: (A) subsequent sensing of the optical response and shear force, (B) producing two-dimensional plots of the obtained signals, (C) image coding to produce filled-in binary patterns of the plotted results, and (D) extraction of corner coordinates as a physical feature of each artifact. Source: Adapted with permission from Springer-Verlag: Appl. Phys. A [46].

First, constantly fluctuating signals corresponding to the optical response and shear force due to the optical near-field interactions between the probe and artifact are sensed. Because the two fluctuating signals in Figure 7A are obtained with the same time sequence, they can be combined as two-dimensional results, as shown in Figure 7B. Here, signals due to shear force are plotted along the horizontal axis, and signals due to optical responses are plotted along the vertical axis. Owing to the signal feedback for keeping a constant distance between the two and the corresponding hysteresis, the results reveal an orbit-like modulation during the processing time. To simply evaluate the result, binary image coding is performed by filling in the result, as shown in Figure 7C, and the filled-in binary image is subjected to feature extraction. Here, the coordinates of the corners in the image are defined and utilized as physical features of the artifact. While various feature extraction methods have been actively studied [51], [52], [53] for realizing high-speed image classification and searching, the features from the accelerated segment test (FAST) [54] method was applied, which is known to be one of the most effective methods. Finally, the coordinates of the corners found in each image are defined as individual features of the retrieved signals and corresponding artifact.

For experimental verification of non-scanning nanoretrieval, the results of two retrievals with different artifacts were quantitatively compared. Non-scanning NOM with a SiO2-fiber probe and Al nanorods were used as a retriever and artifact, respectively. The radius of curvature of the tip of the probe was less than 20 nm, and the tip was covered with a 50-nm-thick Au layer formed by sputtering, as shown in Figure 8A. The nanorods were grown on a Si substrate by the glancing angle deposition (GLAD) method [55]. This method can realize various sizes, shapes, and constituents of nanorods by controlling multiple independent magnetron sources and a three-dimensional rotation stage, which perform the deposition of materials by positioning a sample substrate over each magnetron at the appropriate distance and angle. The distance between the probe and nanorods was set to be less than 50 nm. Artifacts 1 and 2, which consisted of different sizes of nanorods with diameters of 30 and 70 nm, respectively, are shown in Figure 8B. All nanorods were grown in the vertical direction to the surface of the substrate. A laser with a wavelength of 532 nm was used to induce optical near-field interactions and the corresponding optical response and shear force. Signals due to the optical response were detected using a photo-multiplier R-3896 manufactured by Hamamatsu Photonics, Japan, and the shear force was retrieved using electrical signals synchronous with the optical response.

Figure 8:

SEM images of the (A) tip of the fiber probe and (B) top views of nanorods in artifacts 1 and 2. Source: Adapted with permission from Springer-Verlag: Appl. Phys. A [46].

Figure 9A and B shows the two-dimensional plots of the retrieved signals due to the shear force and the optical response, respectively, with artifacts 1 and 2. The processing time for each retrieval was 20 ms. A clear difference can be observed between the two results.

Figure 9:

Two-dimensional plots of retrieved signals by using non-scanning nanoretrieval with (A) artifact 1 and (B) artifact 2. Source: Adapted with permission from Springer-Verlag: Appl. Phys. A [46].

To quantitatively evaluate the difference, both results were coded to produce filled-in binary patterns. Then, their corners were extracted as their individual features. Similarity values were calculated by comparing the coordinates of the extracted corners of both patterns. In general, a similarity value is defined as the Hamming distance between two sets of data. In the case of the proposed method, the coordinates of the corners were coded to a binary dataset, and the two sets of data were compared with the other. In the results, the similarity value between the two was calculated to be 0.10 when normalized to a maximum value of 1.00 in the case of self-similarity. The difference between the two values of 0.10 and 1.00 is sufficiently large for actual authentication between artifacts 1 and 2. Figure 10 shows the results of comparing other patterns that were obtained at other retrieval points in artifact 1.

Figure 10:

Similarity values between the retrieved result at the reference point and other points in artifact 1. Source: Adapted with permission from Springer-Verlag: Appl. Phys. A [46].

In the case of the same artifact, the similarity values were more than 0.40. This result indicates that valid authentication can be implemented by setting an appropriate threshold for the similarity value, such as 0.40.

Here, filled-in binary patterns were produced to extract the physical features of each artifact. However, this idea is just one possibility that was used to demonstrate the basics of the proposed method. The corners in each pattern, which were extracted by the FAST method, are another possible definition. Further studies will need to be performed to optimize each procedure in non-scanning nanoretrieval.

## 5 Conclusion

Nanophotonics-based optical information security is the functional fusion of nanoscale artifact metrics and nanophotonics. Nanoscale artifacts with a super-resolution structure based on a random procedure, namely, the random collapse of resist patterns, were found to provide high-level security, as quantified by the calculated results of FMR, FNMR, and CMR. Such structures behave well as sources of optical near fields, and the distribution of optical near fields can be defined as a nanophotonic code that corresponds to both specifications of the artifact and retriever. This was demonstrated experimentally through the use of a nano-/macro-hierarchical hologram. Non-scanning nanoretrieval is proposed for a compact authentication system based on the concept of nanophotonics-based optical information security.

The most important matter of nanophotonics-based optical information security is that its security conditions, especially clone resistance and individuality, depend on the randomness of the procedure and fundamental physics of nanophotonics. This is fundamentally different from conventional information security in that their security levels are basically constrained by the specifications of fabrication technologies. The proposed approach can be expected to break down the cat-and-mouse game between defenders and attackers in the field of physical security. For the practical application and wide popularization of this concept, both the artifact and retriever need to be implemented with a simple procedure and compact setup, respectively. From this point of view, the realized nanoscale random patterns presented in Section 2 and the addition of nanoscale structures to a hologram as described in Section 3 can both be processed by a conventional lithography procedure. No additional steps are required. While the demonstrations presented in Sections 2 and 3 utilized SEM and SNOM, respectively, non-scanning nanoretrieval with a much shorter processing time than existing SNOM is proposed and was successfully demonstrated, as presented in Section 4. Based on the described concept and demonstrations of nanophotonics-based optical information security, further theoretical and experimental activities are expected to be developed for practical use in daily life applications of very high-level artifact metrics in the near future.

## Acknowledgments

The authors thank T. Matsumoto and N. Yoshida from Yokohama National University, Japan, and Y. Ohyagi, M. Hoga, and S. Nishio from Dai Nippon Printing, Co. Ltd., Japan, for their valuable collaborations. This work was supported in part by a Grant-in-Aid in Scientific Research and the Core-to-Core Program, A. Advanced Research Networks from the Japan Society for the Promotion of Science.

## References

• [1]

Javidi B, ed. Optical and digital techniques for information security. Springer Verlag, 2005. Google Scholar

• [2]

Réfrégier P, Javidi B. Optical image encryption based on input plane and Fourier plane random encoding. Opt Lett 1995;20:767–9.

• [3]

Wang RK, Watson LA, Chatwin C. Random phase encoding for optical security. Opt Eng 1996;35:2464–9.

• [4]

Mogensen PC, Glückstad J. Phase-only optical encryption. Opt Lett 2000;25:566–8.

• [5]

Unnikrishnan G, Joseph J, Singh K. Optical encryption by double-random phase encoding in the fractional Fourier domain. Opt Lett 2000;25:887–9.

• [6]

Hennelly K, Sheridan JT. Fractional Fourier transform based image encryption: phase retrieval algorithm. Opt Commun 2003;226:61–80.

• [7]

Matoba O, Javidi B. Encrypted optical memory system using three-dimensional keys in the Fresnel domain. Opt Lett 1999;24:762–4.

• [8]

Situ G., Zhang JA. Lensless optical security system based on computer-generated phase only masks. Opt Commun 2004;232:115–22.

• [9]

Situ G, Zhang J. Double random-phase encoding in the Fresnel domain. Opt Lett 2004;29:1584–6.

• [10]

Chen L, Zhao D. Optical image encryption with Hartley transforms. Opt Lett 2006;31, 3438–40.

• [11]

Rosen J, Javidi B. Hidden images in halftone pictures. Appl Opt 2001;40:3346–53.

• [12]

Kim JJ, Kim JH, Kim ES. Optodigital implementation of multiple information hiding and extraction system. Opt Eng 2004;43:113–25.

• [13]

Hayasaki Y, Matsuba Y, Nagaoka A, Yamamoto H, Nishida N. Hiding an image with a light-scattering medium and use of a contrast-discrimination method for readout. Appl Opt 2004;43:1552–58.

• [14]

Situ G, Zhang J. Image hiding with computer-generated phase codes for optical authentication. Opt Commun 2005;245:55–65.

• [15]

Takai N, Mifune Y. Digital watermarking by a holographic technique. Appl Opt 2002;43:3078–84. Google Scholar

• [16]

Kishk S, Javidi B. 3D object watermarking by a 3D hidden object. Opt Express 2003;11:874–88.

• [17]

Li Y, Kreske K, Rosen J. Security and encryption optical systems based on a correlator with significant output images. Appl Opt 2000;39:5295–301.

• [18]

Abookasis D, Arazi O, Rosen J, Javidi B. Security optical systems based on a joint transform correlator with significant output images. Opt Eng 2001;40:1584–9.

• [19]

Yamazaki M, Ohtsubo J. Optimization of encrypted holograms in optical security systems. Opt Eng 2001;40:132–7.

• [20]

Abookasis D, Batikoff A, Famini H, Rosen J. Performance comparison of iterative algorithms for generating digital correlation holograms used in optical security systems. Appl Opt 2006;45:4617–24.

• [21]

Chang HT, Lu WC, Kuo CJ. Multiple-phase retrieval for optical security systems by use of random-phase encoding. Appl Opt 2002;41:4815–34.

• [22]

Situ G, Zhang J. A cascaded iterative Fourier transform algorithm for optical security applications. Optik 2003;114:473–7.

• [23]

Meng XF, Cai LZ, Yang XL, Shen XX, Dong GY. Information security system by iterative multiple-phase retrieval and pixel random permutation. Appl Opt 2006;45:3289–97.

• [24]

Matsumoto H, Matsumoto T. Clone match rate evaluation for an artifact-metric system. IPSJ J 2003;44:1991–2001. Google Scholar

• [25]

Naruse M, Tate N, Aono M, Ohtsu M. Information physics fundamentals of nanophotonics. Rep Prog Phys 2013;76:056401.

• [26]

Naruse M, Tate N, Ohtsu M. Optical security based on near-field processes at the nanoscale. J Opt 2012;14:094002.

• [27]

Kobayashi K, Sangu S, Ito H, Ohtsu M. Near-field optical potential for a neutral atom. Phys Rev A 2001;63:013806.

• [28]

Ohtsu M, Kobayashi K. Optical near fields. Berlin: Springer, 2004. Google Scholar

• [29]

Tanaka Y, Kobayashi K. Spatial localization of an optical near field in one-dimensional nanomaterial system. Phys E 2007;40:297–300.

• [30]

Ohtsu M. Dressed photons—concepts of light-matter fusion technology. Berlin: Springer, 2013. Google Scholar

• [31]

Matsumoto T, Hoga M, Ohyagi Y, et al. Nano-artifact metrics based on random collapse of resist. Sci Rep 2014;4:1–5. Google Scholar

• [32]

Matoba O, Nomura T, Pérez-Cabré E, Millan MS, Javidi B. Optical techniques for information security. Proc IEEE J 2009;97:1128–1148.

• [33]

Chen W, Javidi B, Chen X. Advances in optical security systems. Adv Opt Photonics 2014;6:120–155.

• [34]

Carnicer A, Hassanfiroozi A, Latorre-Carmona P, Huang Y.-P, Javidi B. Security authentication using phase-encoded nanoparticle structures and polarized light. Opt Lett 2015;40:135–138.

• [35]

Carnicer A, Arteaga O, Pascual E, et al. Optical security verification by synthesizing thin films with unique polarimetric signatures. Opt Lett 2015;40:5399–402.

• [36]

Javidi B, Carnicer A, Yamaguchi M, et al. Roadmap on optical security. J Opt 2016;18:083001.

• [37]

Namatsu H, Kurihara K, Nagase M, Iwadate K, Murase K. Dimensional limitations of silicon nanolines resulting from pattern distortion due to surface tension of rinse water. Appl Phys Lett 1995;66:2655–7.

• [38]

Naruse M, Hoga M, Ohyagi Y, et al. Eigenanalysis of morphological diversity in silicon random nanostructures formed via resist collapse. Physica A 2016;462:883–8.

• [39]

Matsumoto T, Yoshida N, Nishio S, et al. Optical nano artifact metrics using silicon random nanostructures. Sci Rep 2016;6:32438.

• [40]

Javidi B, Horner JL. Optical pattern recognition for validation and security verification. Opt Eng 1994;33:1752–6.

• [41]

Refregier P, Javidi B. Optical image encryption based on input plane and Fourier plane random encoding. Opt Lett 1995;20:767–9.

• [42]

Rakuljic GA, Leyva V, Yariv A. Optical data storage by using orthogonal wavelength-multiplexed volume holograms. Opt Lett 1992;17:1471–3.

• [43]

Tate N, Nomura W, Yatsui T, Naruse M, Ohtsu M. Hierarchical hologram based on optical near- and far-field responses. Opt Express 2008;16:607–12.

• [44]

Tate N, Naruse M, Yatsui T, et al. Nanophotonic code embedded in embossed hologram for hierarchical information retrieval. Opt Express 2010;18:7497–505.

• [45]

Naruse M, Yatsui T, Nomura W, Hirose N, Ohtsu M. Hierarchy in optical near-fields and its application to memory retrieval. Opt Express 2005;13:9265–71.

• [46]

Tate N, Naruse M, Nomura W, et al. Non-scanning optical near-field microscopy for nanophotonic security. App Phys A Mater 2015;121:1383–7.

• [47]

Naruse M, Hori H, Kobayashi K, et al. Information theoretical analysis of hierarchical nano-optical systems in the subwavelength regime. J Opt Soc Am B 2009;26:1772–9.

• [48]

Candès EJ, Romberg J, Tao T. Robust uncertainty principles: exact signal reconstruction from highly incomplete frequency information. IEEE Trans Inform Theory 2006;52:489–509.

• [49]

Donoho D. Compressed sensing. IEEE Trans Inf Theory 2006;52:1289–306.

• [50]

Haupt J, Nowak R. Signal reconstruction from noisy random projections. IEEE Trans Inf Theory 2006;52:4036–48.

• [51]

Harris C, Stephens MA. Combined corner and edge detector. Proc of the Fourth Alvey Vision Conf 1998;147–51. Google Scholar

• [52]

Smith SM, Brady JM. SUSAN—A new approach to low level image processing. Int J Comput Vision 1997;23:45–78.

• [53]

Lowe DG. Object recognition from local scale-invariant features. In Int Conf on Computer Vision 1999;1150–1157. Google Scholar

• [54]

Rosten E, Drummond T. Machine learning for high-speed corner detection. Volume 3951 of the series Lecture Notes in Computer Science 2006;430–43. Google Scholar

• [55]

Pihosh Y, Turkevych I, Ye J, et al. Physical and photocatalytic properties of TiO2 nanostructures fabricated by means of glancing angle deposition. ECS Trans 2009;16:49–58. Google Scholar

Revised: 2016-09-25

Accepted: 2016-10-06

Published Online: 2016-12-28

Citation Information: Nanophotonics, Volume 6, Issue 3, Pages 613–622, ISSN (Online) 2192-8614, ISSN (Print) 2192-8606,

Export Citation

## Citing Articles

[1]
V. Sasikala and K. Chitra
Journal of Optics, 2018