Individual Authentication Medium, Method for Producing Same, and Authentication System Using Same

Abstract

In authentication of machineries and cards (artifact) that are used in social acts such as economic acts, an approach of artifact metrics corresponding to biometrics is effective. Therefore, the subject is to find out a material that satisfies requirements of artifact metrics and is, preferably, suppliable stably and also economically, to establish the production method thereof, and to apply these to an individual authentication system of artifact. Porous glass, which possesses a spinodal phase separation structure, is an individual authentication medium as artifact metrics. There is provided a production method thereof, and an individual authentication system utilizing the individual authentication medium.

Claims

1. An individual authentication medium, the medium being a fragment of a material possessing a spinodal phase separation structure and having at least one flat surface.

2. The individual authentication medium according to claim 1, the medium being formed of a phase-separated method porous glass prepared by using a phase separation structure of borosilicate glass.

3. The individual authentication medium according to claim 1, the medium having been subjected to at least one of processing for attaching a filler to pores of the individual authentication medium or processing of smoothing the surface of the individual authentication medium, and being formed of a phase-separated method porous glass having an arbitrary pore structure having an average pore diameter of 1 nm up to 100 m.

4. The individual authentication medium according to claim 2, the medium having been subjected to at least one of processing for attaching a filler to pores of the individual authentication medium or processing of smoothing the surface of the individual authentication medium, and being formed of a phase-separated method porous glass having an arbitrary pore structure having an average pore diameter of 1 nm up to 100 m.

5. A method for producing an individual authentication medium, comprising: a raw material-mixing step of mixing raw materials of glass; a fusion step of fusing the mixed materials to prepare a borosilicate glass host material; a molding step of molding the prepared borosilicate glass host material; a phase separation step of conducting phase separation by subjecting the molded borosilicate glass host material to a heat treatment; a chemical treatment step of subjecting the phase-separated borosilicate glass host material to a chemical treatment to prepare porous glass; and a step of conducting stabilization by resin sealing and the like

6. An individual authentication system including: an individual authentication medium that is a fragment of a material possessing a spinodal phase separation structure and has at least one flat surface; a data processor; and an observation device connected to the data processor, and: configured such that; the individual authentication medium is the individual authentication medium according to claim 1; the observation device acquires a surface image of the individual authentication medium and sends it to the data processor; and the data processor checks calculated characteristic point information against characteristic point information of the individual authentication medium, having been registered in advance for the database, and practices the individual authentication.

7. The individual authentication system according to claim 6, wherein the individual authentication medium is formed of a phase-separated method porous glass prepared using a phase separation structure of borosilicate glass.

8. The individual authentication system according to claim 6, wherein the individual authentication medium has been subjected to at least one of processing for attaching a filler to pores of the individual authentication medium or processing of smoothing the surface of the individual authentication medium, and is formed of a phase-separated method porous glass having an arbitrary pore structure having an average pore diameter of 1 nm up to 100 m.

9. The individual authentication system according to claim 7, wherein the individual authentication medium has been subjected to at least one of processing for attaching a filler to pores of the individual authentication medium or processing of smoothing the surface of the individual authentication medium, and is formed of a phase-separated method porous glass having an arbitrary pore structure having an average pore diameter of 1 nm up to 100 m.

Description

BRIEF DESCRIPTION OF DRAWINGS

[0083] FIG. 1 is a scanning electron microscope photograph of representative porous glass of an embodiment of the present invention, which is a photograph that shows three kinds of porous glasses having different pore diameters have similar structures;

[0084] FIG. 2 is a schematic diagram of an evaluation apparatus of an embodiment of the present invention;

[0085] FIG. 3 is a photograph of appearance of a porous glass medium of an embodiment of the present invention (8 mm*8 mm*1.1 mm);

[0086] FIG. 4 is a diagram showing a production process of a porous glass medium of an embodiment of the present invention;

[0087] FIG. 5 is a scanning electron microscope photograph of porous glass having pore diameter of 200 nm used for evaluating an embodiment of the present invention;

[0088] FIG. 6 is a graph of FAR and FRR of 1-MNN: in a case where measurement is performed at the shortest distance (n=1), of an embodiment of the present invention;

[0089] FIG. 7 is a graph of FAR and FRR of 10-MNN: in a case where 10 distances are obtained in ascending order and measurement is performed by an average thereof (n=10), of an embodiment of the present invention; and

[0090] FIG. 8 is a graph of FAR and FRR of 100-MNN: in a case where 100 distances are obtained in ascending order and measurement is performed by an average thereof (n=100), of an embodiment of the present invention.

DESCRIPTION OF EMBODIMENTS

(Production Method)

[0091] Hereinafter, there will be explained in more detail a preparation method and characteristics of a material in individual authentication using this phase-separated method porous glass. Glass is an inorganic material having an amorphous network, and can be seen as an amorphous material that is antithetical to systematization or ordering, as a material representative of disorder, from the viewpoint of atom/molecule of an angstrom level. On the other hand, from the viewpoint of nano or larger levels, it can be understood as a uniform raw material.

[0092] Glasses that we usually see include soda-lime glass, lead glass, borosilicate glass, and quartz glass. Soda-lime glass is the most common glass formed mainly of three components of silicic acid, sodium oxide and calcium oxide. Lead glass is a glass having lead oxide as main component, and is known as a crystal glass. Borosilicate glass is a glass also referred to as a laboratory glass. Quartz glass is a glass formed from silicic acid alone.

[0093] Among these, borosilicate glass that has a low expansion coefficient, high heat shock resistance and very excellent chemical stability shows a phenomenon that it suddenly becomes chemically weak and tends to crack when used under a specific condition, which was formerly known as borate anomaly. On the basis of a fact that the cause of this was a phenomena referred to as phase separation of glass and expensive quartz glass could be prepared inexpensively if this phenomenon was used efficiently, synthesized quartz glass: Vycor glass was invented by Corning Incorporated, USA, in 1934. Then, as an intermediate, phase-separated method porous glass was invented.

[0094] As the phase-separated method porous glass, there are reported that of a high silicic acid type of SiO2: 96% or more in the final component, and that of a borosilicic acid type containing alumina or zirconia. These two kinds have different ranges of pore diameters to be prepared, that is, for high silicic acid type: 1 nm-300 nm, and for borosilicic acid type: 200 nm-50 m.

[0095] Referring to FIG. 4, a preparation process of high silicic acid type porous glass is shown below. [0096] In a raw material-mixing step 402, raw materials are mixed using silica sand and borax, boric acid and sodium carbonate, alumina etc. [0097] In a fusion step 404, raw materials mixed in the raw material-mixing step 402 are fused at around 1200 C.-1500 C. to prepare a borosilicate glass host material having SiO.sub.2, B.sub.2O.sub.3 and Na.sub.2O as main components. [0098] In a molding step 406, the borosilicate glass host material prepared in the fusion step 404 is molded at around 800 C.-1100 C. [0099] In a phase separation step 408, the borosilicate glass host material molded in the molding step 406 is subjected to a heat treatment, in which it is held at a temperature of a glass transition point or higher, to bring about phase separation. On this occasion, when the borosilicate glass host material is exposed to a temperature of a glass transition point or higher, respective component atoms fluctuate in the inside and begin to move. This phenomena closely resembles crystallization, but the glass keeps an amorphous state and is separated into two glass phases that take more stable states at the temperature, and, while fine atomic arrangement keeps a random amorphous phase, an ordered structure is formed as a somewhat macro cluster.

[0100] In the two phases, one is a silica phase formed of almost silicic acid, and the remnant is a sodium borate phase formed of boric acid, sodium oxide and silicic acid. On this occasion, depending on a composition and temperature of starting glass, there is a case where a spinodal structure, in which organization is intertwined such as sponge, is formed, and a case where a droplet structure, in which one phase is isolated such as a liquid drop, is formed. Porous glass is obtained from a spinodal phase-separated substance that has a continuous phase structure.

[0101] When continuously exposed to certain or higher temperature in a state separated into two phases, these two phases bring about rearrangement and an ordered structure grows. Meanwhile, when temperature exceeds a range in which the phase separation is brought about, the glass returns again to homogeneous composition. It is possible to consider that this phase separation phenomenon is a phenomenon in which component molecules composing glass repeat self-assembly and dissipative structure for stabilization, and that two phases that are most stable at the temperature self-organize toward a direction in which an interfacial area is minimized. This is pattern formation due to spinodal phase separation of a liquid, and is structure formation, which can be taken out by freezing the liquid due to self-organization, because a reaction of glass has an extremely long time axis.

[0102] A characteristic of structure formation due to spinodal phase separation is ordering in a macro view point, but it is believed that, unlike crystal growth, random formation in order formation is included, and that a structure formed of many branched structures are all different although in similar figures. Elaboration of a pore structure by control from an outside can control pore diameter, but cannot give control of a branched structure. Moreover, this structure can also give random structures of various modes by varying the pore diameter.

[0103] In the random structure due to phase separation, randomness appears without exception on the basis of a distance unit shown by pore diameter, and, as a consequence of generation of subsequent randomness on the basis of subsequent distance unit, finally a high degree of randomness is achieved.

[0104] A spinodal structure of porous glass gives a complex structure of two-dimension from the viewpoint of surface and of three-dimension as an overall structure. A phase separation structure is determined by factors of temperature and time period, and properties of porous glass to be prepared is different when temperature is different even under conditions that give the same structure, but, because they have similar figures, dissimilarity thereof is not distinct apparently, which is also a characteristic.

[0105] In a chemical treatment step 410 in FIG. 4, the borosilicate glass host material phase-separated in the phase separation step 408 is subjected to a chemical treatment using an acid solution. Usually, several-normal sulfuric acid or nitric acid is used, which is held at 90 C. or higher to elute a sodium borate phase. After the end of the treatment, and water-wash and drying, an A type porous glass of SiO2: about 96% shown in FIG. 5 is obtained.

[0106] A pore structure of the A type porous glass does not reflect a phase separation structure, and actually is a structure in which silica gel originating in a sodium borate phase accumulates in a skeleton structure formed of silica glass. Furthermore, in the chemical treatment step 410 in FIG. 4, a B type porous glass reflecting the phase separation structure is obtained by removing the silica gel in some way, such as removal thereof with an alkali aqueous solution.

[0107] Furthermore, as to characteristics of this porous glass, a skeleton supporting the structure is glass, which is firm and solid, and the glass is unlikely to break as compared with ordinary glass because strain has been eliminated in phase separation, to give a firm material mechanically and chemically. Furthermore, after a chemical treatment has been performed once to result in porous glass, phase separation does not continue to give a material that is stable thermally too.

[0108] In the above-described production method, main raw materials are silicic acid (silica sand), boric acid, sodium carbonate, alumina, zirconia, etc., which are the same as those of ordinary borosilicate glass and therefore the material costs are not expensive.

[0109] Moreover, the production process includes mixing of raw materials, fusion, molding, phase separation, and chemical treatment, and equipment necessary for the production is a mixer, fusion crucible, electric furnace (air atmosphere), chemical treatment reactor (level of flask), and particularly expensive equipment is unnecessary, but price of porous glass at this time is high. However, this is mainly due to process costs, and is mainly due to labor cost in small-lot production. Furthermore, processing of porous glass is not different from ordinary glass processing, and facilities and procedures having been established already can be utilized. Porous glass may easily be supplied in a mass scale, if needed. Phase-separated method porous glass to be prepared in this way is finished as a medium, for example, of 1 mm square and 50 m in thickness, is stabilized by resin sealing or the like, and is fixed firmly to a target.

[0110] (Specific Authentication Method)

[0111] An authentication system that realizes artifact metrics is referred to as an artifact metric system (see, for example, Non-Patent Literature No. 2). Here, an artifact metric system is configured basically of following two phases. [0112] Registration phase: to register a target artifact, data representing characteristics thereof (randomness) (hereinafter, referred to as registration data) are acquired with a sensor and are recorded in a database. [0113] Verification phase: to perform authentication of a target artifact, data representing characteristics thereof (hereinafter, referred to as verification data) are acquired with a sensor. Then, using the verification data, and registration data recorded in the database, verification is performed and, finally, a verification result (acceptable or rejectable) is output.

[0114] A purpose of the artifact metric system is to determine whether or not an artifact presented in the registration phase and an artifact presented in the verification phase are the same substance. When they are determined to be the same, acceptable is output, and when they are determined to be not the same, rejectable is output. As indices for objectively measuring basic accuracy of an artifact metric system, following FAR (or FMR: False Match Rate) and FRR (or FNMR: False Non-Match Rate) are known. [0115] FAR (False Acceptance Rate): probability that different artifacts are erroneously determined to be acceptable. [0116] FRR (False Rejection Rate): probability that the same one artifact is erroneously determined to be rejectable.

[0117] Above-described two kinds of error probability (the error probability that false is determined as true, and the error probability that true is determined as false) are not limited to artifact metric systems. These can be considered generally in all authentication systems (such as a biometric system), determination systems and testing systems (such as a hypothesis testing), and are indices for measuring basic accuracy of these systems.

[0118] Moreover, as an index for measuring enhanced security for an artifact metric system, the following index is known (see, for example, Non-Patent Literature No. 3). [0119] CAR (Clone Acceptance Rate): probability of erroneous acceptance of clone, which is a false artifact

[0120] Comprehensive analysis of CAR for producibility of all clones is generally difficult, but it would be desirable to analyze CAR as widely as possible.

[0121] An embodiment of an individual authentication system 200 is shown in FIG. 2. The individual authentication system 200 includes an individual authentication medium 202, a data processor 206 and an observation device 204 connected to the data processor. In authentication, in advance, characteristic information on an image in the individual authentication medium 202 is registered for a database 214. In individual authentication, the observation device 204 acquires a surface image of the individual authentication medium 202 in 220. The observation device 204 sends the acquired surface image to the connected data processor 206. The data processor 206 can be equipped with a characteristic point calculator 210 that calculates characteristic points from the surface image received from the observation device 204. In addition, it can be equipped with a characteristic point checker 212 that checks calculated characteristic points against characteristic point information of the individual authentication medium 202 having been registered in advance for the database and practices authentication.

[0122] Here, the observation device 204 can be any device such as an optical camera, scanning microscopes such as a scanning electron microscope, laser microscope or AFM, or the like. The data processor 206 can be any equipment such as a server and personal computer (PC). Moreover, the characteristic point calculator 210 can include any calculation technique.

[0123] However, to perform this work operation for all authentication is accompanied with measurement, and therefore it is considered, except for performing absolute authentication, usually to construct a system in which digital information to be acquired from an image is registered for a device and the information therein is to be checked.

[0124] In other words, in this authentication system, a huge number of random information is included in the phase-separated method porous glass being artifact metrics, and therefore it is possible to consider this as a table of random numbers given from nature. Then, ordinary authentication can be practiced with extremely high stability by giving randomly a part of porous glass information possessed by the device side in accordance with request from a checker.

[0125] Meanwhile, in absolute authentication, the use of access to artifact metrics itself is preferred. That is, this authentication system is configured of an individual authentication system and devices that authenticate individuality by acquiring an image on the surface of the individual authentication medium with an optical camera, a scanning microscope such as a scanning electron microscope, laser microscope or AFM, or the like and using characteristic point information of the image. It is possible to use an individual authentication system and devices matched with a material as a consequence of selection in accordance with a security level requested by the various observation systems as described above.

Example 1

[0126] According to phase-separated method porous glass of a borosilicic acid type described in Patent Literature Nos. 7, 8, six plate-like substances having pore diameter of 200 nm are prepared. Then, on the basis of images obtained by observing each with a scanning electron microscope (SEM) at magnifications of 5,000 and 10,000, a simple experiment for examining randomness of distribution of pores in porous glass was performed.

[0127] For determining match/unmatch of distribution pattern of pores relative to the image of porous glass, a technique on the basis of pattern matching using an amount of characteristics was utilized. Details thereof will be explained in the following paragraph or later.

(Preparation of Porous Glass and Method for Acquiring Image)

[0128] As a sample, typical borosilicic acid type porous glass (average pore diameter; 200 nm) was prepared, and the different eight sites were photographed with a SEM (magnification: 10,000) to give eight original images I, II, . . . , VIII (hereinafter, these are referred to as parent images). In the same way, the same eight sites were photographed with a SEM at a changed magnification (magnification: 5,000) to give additional eight original images i, ii, . . . , viii (hereinafter, these are referred to as child images). Here, there is assumed a case where accuracies when an image is acquired are different between a registration phase and verification phase in an artifact metric system. To examine whether or not verification functions effectively even under a bad circumstance such as different accuracies of sensors for acquiring an image, images of varied magnifications were set as experiment objects. For example, it was considered that the parent image was registration data, and the child image was verification data.

(Verification Method in Artifact Metric System)

[0129] As a verification method for the parent image and the child image, a technique on the basis of pattern matching utilizing a Sift characteristic amount was utilized. Here, Sift (Scale-Invariant Feature Transform) (see, for example, Non-Patent Literature No. 6) is a method for characteristic point detection and description thereof presented by D. Lowe in 1999. Sift is characterized in that it never changes regardless of rotation or scale change of an image and is resistant to change in illumination. Sift is a characteristic amount detecting technique that is installed in OpenCV, which is an image processing library of open source, and is widely used in image discrimination etc. Furthermore, in verification of the parent image and the child image, an approach with a local characteristic amount is also utilized. Here, the local characteristic amount is an amount obtained by detecting a point having highly contrasting density of image and expressing differentiation relative to the circumference thereof with a vector (see, for example, Non-Patent Literature No. 11).

[0130] Characteristic amounts obtained by above-described method for any combination of the parent image and the child image are compared in a round robin manner, and a pair of characteristic points with the smallest difference d is found out, and the d is set as a distance. After that, the distances d are compared for all target pairs, n distances d were selected in ascending order, and an average thereof was defined as distance between both images. On the basis of this, pattern matching between the parent image and the child image was performed, in which, as values of the n, following three kinds were set to targets of examination. [0131] 1-MNN: a case where measurement is performed by the smallest distance alone (n=1) [0132] 10-MNN: a case where ten distances are obtained in ascending order and measurement is performed by an average thereof (n=10) [0133] 100-MNN: a case where hundred distances are obtained in ascending order and measurement is performed by an average thereof (n=100)

[0134] Utilizing distances of examination targets, a case where a distance between both images (parent image and child image) being targets is less than a certain value (hereinafter, referred to as a threshold) is determined to be acceptable, and a case where a distance is equal to or more than the threshold is determined to be rejectable.

(Results of Experiments)

[0135] With respect to parent images I, II, . . . , VIII and child images i, ii, . . . , viii, values of distances (round down to the nearest decimal) on the basis of 1-MNN are shown in Table 1.

TABLE-US-00001 TABLE 1 I II III IV V VI VII VIII i 11 109 159 184 206 228 231 222 ii 110 16 153 175 195 185 241 221 iii 167 171 17 230 240 225 245 232 iv 143 162 221 15 233 218 224 225 v 120 134 234 179 17 213 234 248 vi 181 202 207 214 201 19 241 221 vii 247 190 236 232 246 236 15 236 viii 239 221 246 231 230 226 240 19

[0136] In Table 1, values lying at diagonal elements designate the same photographing site in the porous glass, and therefore are distance values of cases to be determined as acceptable, and the others are distance values of cases to be determined as rejectable. Here, all values lying at diagonal elements are less than 20, while the other values are equal to or more than 100. As a consequence, a highly unique property of the image can be confirmed, and this shows that randomness of distribution of pores in the porous glass is high.

[0137] FIG. 6 is a graph of FAR and FRR in the case of 1-MNN. Here, the abscissa axis in FIG. 6 represents a threshold of distance, and the ordinate axis represents an error probability. In FIG. 6, total 64 distances, that is, 8 parent images by 8 child images, are calculated to draw the graph. This result is believed to show good accuracy with respect to 64 experimental data. If the difference between magnifications of the parent image and the child image is almost the same, furthermore high accuracy would be represented, as compared with this experimental result. In this description, as the first experiment of a case where a porous glass is applied to artifact metrics, the experiment was performed with a small number of experimental data, in the presence of difference in accuracy of sensors acquiring registration data and verification data, and according to a scenario assuming a simple verification method. Nevertheless, it is believed that the experiment result shows good accuracy. Therefore, it is believed that porous glass can be expected as an application material to artifact metrics.

[0138] FIG. 7 is a graph of FAR and FRR in the case of 10-MNN. Here, the abscissa axis in FIG. 7 represents a threshold of distance, and the ordinate axis represents an error probability. In FIG. 7, total 64 distances, that is, 8 parent images by 8 child images, are calculated to draw the graph.

[0139] FIG. 8 is a graph of FAR and FRR in the case of 100-MNN. Here, the abscissa axis in FIG. 8 represents a threshold of distance, and the ordinate axis represents an error probability. In FIG. 8, total 64 distances, that is, 8 parent images by 8 child images, are calculated to draw the graph.

[0140] The embodiment described in the present description and drawings is an example, and should not be construed as limitation or restriction for the attached Claims.

INDUSTRIAL APPLICABILITY

[0141] The present invention relates to individual authentication that is needed in practice of network environment and credit acts such as various commercial transactions and contracts in explosive spread of mobile devices or in an IoT community. In particular, for various computers, mobile phones, machineries such as automobiles and cards (artifact) that are used in social acts such as economic acts, which do not have sure individual identifiability at this time, artifact metrics having spinodal phase separation structure prepared as self-organization can be utilized. Consequently, it becomes possible to give an ultimate individual authentication system. Moreover, by cooperating with biometrics identifying an individual person, forgery and falsification are made difficult. Furthermore, network systems for performing individual identification, including dishonest act become easily discriminated, authentication can be performed with improved accuracy, and a safe network environment at low cost.