JP5647297B2 - Authentication method and system using connecting line generated from optical information difference - Google Patents

Description

  The present invention relates to an authentication method and system using a connecting line generated from optical information difference, and extracts a feature unique to a target image from image data, and an authentication system and authentication information registration and inquiry method using the extracted feature About.

  Conventionally, various authentication technologies have been used to determine whether or not a legitimate subject is legitimate in every aspect of human activity, and more advanced authentication technologies have been demanded from the progress of computerization of the Internet and information. Yes. Classic authentication technology has been used to mark unique shapes such as seals and signatures, but due to counterfeiting, holograms, creases (watermarks), latent image patterns, pearl ink, micro characters In addition to anti-counterfeiting using technologies such as special light-emitting ink and intaglio printing, along with the progress of computerization, certain information such as passwords is currently kept secret by encrypting it. By checking, we try to prevent forgery. In addition, biometric information unique to an individual human body, such as a fingerprint, is registered in advance, and when authentication is required, various biometrics that perform authentication based on whether the biometric information matches the registered biometric information by reading the actual biometric information Authentication technology is also used (see, for example, Patent Document 1).

JP 2009-205393 A

  However, among the authentication techniques described above, a technique using a password has a problem that the amount of information is extremely small and theft and counterfeiting are easy because the password must be determined in advance and the user itself must be stored.

  In the technique using biological information, only a human can determine whether or not it is a legitimate object, and basically it cannot be used for other objects, animals and plants. In recent years, an act of forging the biometric information itself has also occurred, and there is a problem that it is not necessarily a completely secure authentication technique.

  Furthermore, many advanced authentication techniques often add authentication information to an object in advance physically or electronically, and processing to the object is necessary, which hinders the spread of the authentication technique. There is a problem.

  The present invention has been made in view of the above-described conventional problems, and does not require processing for an object to be authenticated, and uses a standard digital imaging device having a certain configuration for each individual feature of the object, and certain configuration conditions It is an object of the present invention to provide an authentication system and an authentication information registration / inquiry method for extracting unique information from image data photographed in the above and determining whether or not it is a legitimate object based on the extracted unique information.

  The first aspect calculates an optical information difference at a specified corresponding position of a plurality of digital data obtained by resolving an authentication target by scanning of the digital imaging means, and subjects corresponding to the pixel size of the digital imaging means A connection line generated by connecting the midpoint of the physical position of the subject corresponding to the center position of the pixel at the corresponding position calculated based on the physical size in FIG. As individual difference information uniquely specified, obtained by resolving digital data by scanning the authentication object a plurality of times with the digital imaging means under the configuration conditions of the storage means for storing the connecting line and the predetermined digital imaging means Compare multiple digital data, and calculate the movement and rotation conversion to match the position where each digital data maps the object to be authenticated. A conversion value calculating means for determining the corresponding position of each pixel array of a plurality of digital data using the calculated movement and rotation conversion, and connecting lines of individual difference information stored in the storage means at the corresponding position And tracking the position of the connection point on the read connection line so as to follow the broken line, and calculating the optical information difference of the pixel corresponding to the physical surface of the superimposed pixel including the position of the connection point. When the optical information difference calculated in the above is determined to be in a predetermined relationship with respect to the order of the connecting lines, it is characterized by comprising individual difference information extraction determining means for determining that the authentication is successful.

  According to a second aspect, in the first aspect, the configuration condition of the digital imaging unit includes the resolution of the digital imaging unit, and the resolution of the digital imaging unit is lower than the size of the particles forming the image to be authenticated. It is characterized by that.

  A third aspect is characterized in that, in the first aspect or the second aspect, the conversion value calculation means continuously scans the authentication target a plurality of times.

  According to a fourth aspect, in any one of the first to third aspects, a correction unit that performs correction to standardize a difference in image data due to a difference in configuration conditions of the digital imaging unit according to preset information. Is further provided.

  According to a fifth aspect, in any one of the first to fourth aspects, the storage unit further stores a connection comparison setting including a parameter for comparison of connection lines and a comparison method, and individual difference information extraction determination The means uses the comparison setting stored in the storage means to track the position of the connection point on the read connection line so as to follow the broken line, and the calculated optical information difference is calculated with respect to the order of the connection lines. When it has a predetermined relationship defined in the comparison setting, it is determined that the authentication is successful.

  A sixth aspect is any one of the first aspect to the fifth aspect, wherein the individual difference information extraction determination unit is configured such that the calculated optical information difference is in a predetermined descending order with respect to the order of the connecting lines. Is characterized in that authentication is determined to be successful.

  A seventh aspect is an authentication information registration method, wherein a plurality of digital data obtained by resolving digital data by scanning an authentication target a plurality of times with a digital imaging unit under a predetermined configuration condition of the digital imaging unit And a conversion value calculation step for calculating a movement and rotation conversion for matching the positions where the respective digital data map the authentication target, and each of the plurality of digital data using the calculated movement and rotation conversion. The step of calculating the optical information difference at the specified corresponding position by specifying the corresponding position of the pixel array, and calculating the physical size in the subject corresponding to the pixel size of the digital data from the configuration conditions of the digital imaging means Depending on the physical size, the midpoint of the physical position of the subject corresponding to the center position of the pixel at the corresponding position is linked. Calculating the points, connecting the linked stores in the descending order of the calculated light information difference, extracting the connecting lines as individual difference information, and registering the extracted connecting lines in the storage means; It is characterized by providing.

  According to an eighth aspect, a subject corresponding to the pixel size of the digital imaging means is calculated by calculating an optical information difference at a specified corresponding position of a plurality of digital data obtained by resolving an authentication target by scanning of the digital imaging means. A connection line generated by connecting the midpoint of the physical position of the subject corresponding to the center position of the pixel at the corresponding position calculated based on the physical size in FIG. In an authentication system including a storage unit that stores connection lines as uniquely-identified individual difference information, the digital imaging unit scans an authentication target a plurality of times and resolves it as digital data under a predetermined configuration of the digital imaging unit. The obtained digital data are compared, and the positions where each digital data is mapped to the authentication target are matched. A conversion value calculation step for calculating movement and rotation conversion for the image, and a corresponding position of each pixel array of the plurality of digital data is specified using the calculated movement and rotation conversion, and stored in the storage means at the corresponding position. The connected line of the individual difference information is read, the position of the connection point on the read connection line is traced so as to follow the broken line, and the pixel corresponding to the physical surface of the superimposed pixel including the position of the connection point An optical information difference calculated by calculating the optical information difference includes an individual difference information extraction determination step for determining that the authentication is successful when it is determined that the optical information difference is in a predetermined relationship with respect to the order of the connecting lines. It is characterized by.

It is a system configuration figure of an authentication system in one embodiment of the present invention. It is a functional block diagram of one embodiment of the present invention. It is a flowchart which shows the authentication process of one Embodiment of this invention. It is a figure for demonstrating the principle of this invention. It is a figure for demonstrating the extraction process of the individual difference authentication shape of one Embodiment of this invention. It is a figure for demonstrating the extraction process of the individual difference authentication shape of one Embodiment of this invention. It is a figure for demonstrating the extraction process of the individual difference authentication shape of one Embodiment of this invention. It is a figure for demonstrating the authentication process performed using the individual difference authentication shape previously extracted of one Embodiment of this invention. It is a figure for demonstrating an example of the authentication process coincidence determination performed using the individual difference authentication shape previously extracted of one Embodiment of this invention. It is a figure for demonstrating an example of the correction process of the extracted target image of one Embodiment of this invention. It is a figure for demonstrating the extraction process of the specific application example of one Embodiment of this invention. It is a figure for demonstrating the coincidence determination process of the specific application example of one Embodiment of this invention. It is a figure for demonstrating another specific application example of one Embodiment of this invention. It is a figure for demonstrating another specific application example of one Embodiment of this invention. It is a figure for demonstrating another specific application example of one Embodiment of this invention. It is a figure for demonstrating another specific application example of one Embodiment of this invention. It is a figure which shows the example of the connection eyeglasses of one Embodiment of this invention. It is a figure which shows an example of the connection system of the starfish line drawing of one Embodiment of this invention. It is a figure which shows another example of the connection system of the starfish line drawing of one Embodiment of this invention. It is a flowchart which shows the process which registers the individual difference information of this embodiment. It is a flowchart which shows the authentication process of this embodiment. It is a figure for demonstrating the detail of calculation of the optical information difference of this embodiment. It is a figure for demonstrating an example of the authentication process coincidence determination performed using the individual difference authentication shape previously extracted of one Embodiment of this invention.

  Embodiments of an authentication system and authentication information registration and inquiry method according to the present invention will be described below with reference to the drawings.

<Individual difference authentication system principle>
The system according to the present embodiment does not require any processing on the object to be authenticated, and the individual features of the object are extracted from image data captured under a certain configuration condition using a standard digital imaging device having a certain configuration. The identity of the object is determined according to the principle relating to the unique information, the unique information extraction method for extracting the unique information from the read image, and the individual difference authentication method for authenticating the object as needed based on the extracted unique information. The object may include any material such as printed matter, various finished products, parts such as parts, and the human body as long as the surface can be scanned or imaged to read an image.

<Specific information based on individual differences>
Scanning a substance that actually exists with a digital imaging device, such as a scanner or a camera, is to digitize the surface state, but it is not possible to read the entire material configuration of interest. Scanning the surface of a material can be said to be a mapping of a phenomenon that occurs on the surface of the material momentarily. For example, an image sensor (CCD / CMOS) that arranges light receiving parts such as image elements that respond to light in the light wavelength (visible light region between 360 nm to 400 nm on the short wavelength side and 760 nm to 830 nm on the long wavelength side) in a plane. In image scanning (scanning), there is a difference in image processing and the spectral method and element configuration for the three primary colors, but the collected light is split into RGB and other components, and the charge reaction of each light receiving unit is digitized to obtain a numerical value. The phenomena are digitally reproduced by a color imaging process (numerical imaging process) arranged in sequence and output as array data. The mapping information of the light is a numerical value replaced with a scale of resolution corresponding to the XYZ arrangement. Further, in an image sensor, if the resolution is high, that is, each light receiving part constituting the sensor is smaller than the length of the wavelength of light, it cannot be observed as light. It becomes impossible to observe. In addition, since the natural substances are randomly arranged, it is impossible to reproduce the natural objects themselves with an image sensor in which the light receiving portions are arranged in an array.

  This is the resolution and reacts to the light received according to the size of the element that is the scanning unit of the image sensor, and this response is digitized and arranged, that is, the light received by each element arranged in a predetermined form on the image sensor. Image data is an array of the amount of electricity generated according to the above. Therefore, this digitization is also the result of digitizing the phenomenon in the target portion corresponding to one scanning unit and performing color compression processing. The image data in which this is arranged is resolved according to the size of the scanning unit. It can be said that it was a phenomenon. In other words, the scanned image data changes according to the configuration conditions such as the resolution. In this embodiment, the target specific information is extracted by focusing on this phenomenon. Here, the configuration conditions are the physical position of the image sensor, type, size, resolution, lens configuration, lens magnification, physical position of lighting equipment such as LEDs, brightness, frequency characteristics, color temperature, and the distance between subjects to be authenticated. The imaging device configuration and shooting conditions are configured by parameters such as a distance, a focus position, a focal length, a shutter speed, a frame rate, a relative position relationship of an optical system including an authentication target and a lens, and an imaging device.

  Since the object whose validity is actually judged is composed of natural substances, the minimum size is “particle” which is smaller than the nano / micro size. Photons are larger than that, and in visible light, the wavelength is in the visible light region between the shorter wavelength side of 360 nm to 400 nm and the longer wavelength side of 760 nm to 830 nm. Does not reflect light (visible light). Therefore, the phenomenon of the natural substance itself, which is much finer than the resolution, cannot be deciphered even by the digitization by the sensor, but the reading result, that is, the value output by the sensor depends on the phenomenon within the size. That is, the scanning position is shifted even a little (micro order), that is, the corresponding position in the data array of the image data obtained when the same object is read a plurality of times and the relative position of the read element are slightly If shifted, the numerical value of the optical information read by the element of the corresponding image sensor is affected, so that the same scanning information cannot be obtained even if the same object is scanned under the same conditions. This always occurs unless there is an image sensor having a resolution smaller than the light wavelength size, and this difference in the reading result is referred to as a resolution difference in the present application. FIG. 4 illustrates such a resolution difference using a specific CCD sensor. Here, the example shown in FIG. 4 is merely an example to which the present principle is applied, and the present invention is not limited by any technique specifically shown in this example.

  Referring to FIG. 4, for example, in a CCD sensor that is generally used in a scanner that reads an authentication target, as shown in FIG. Visible light having an incoming wavelength (visible light region between 360 nm to 400 nm on the short wavelength side and 760 nm to 830 nm on the long wavelength side) is converted into charges and accumulated. The accumulated charge is digitized as the intensity of light, and according to the digitized numerical values such as RGB, the object to be imaged is incident light limited by the pitch of the image sensor and the aperture, and one color per element. And are expressed as color arrangement data. The color arrangement data obtained in this way is expressed as numerical data representing the coordinates representing the position of the image sensor and the respective components of the color system such as RGB. For example, assuming that (X coordinate, Y coordinate: R, G, B), as shown in the figure, (1,1: 112, 28, 93), (1,2: 17, 30,104), (1,3: 39, 36, 107)..., But these values are different because the scanning position changes in micro order each time the data is read by the scanner.

  This resolution difference also occurs in the printed matter printed in the nano / micro size. In a reproduction product (mass production product) such as a printed matter, an image obtained by scanning is influenced by a nano / micro level printing error and a difference in paper surface that cannot be actually observed from an actual phenomenon. In other words, the resolution difference reflects the individual characteristics of the nano-sized particles that are made unintentionally or that cannot be mass-produced and manufactured by the intention. In the present application, the difference due to the difference in resolution at the time of object reading caused by the characteristics of the individual is referred to as “individual difference”.

<Extraction of information specific to the target>
In the present invention, this individual difference is acquired from a target such as, but not limited to, a printed matter, and information unique to one target is extracted from the acquired individual difference and used for authentication processing. One method of extracting information specific to such an object will be described. As described above, the image data output for each scan has unstable reproducibility of resolution, and individual differences are reflected here. However, since the value itself obtained as a result of simple resolution is usually not constant, it cannot be used as information that uniquely represents the object as it is. Therefore, information that can obtain a certain relationship from individual differences is extracted. That is, the obtained read data is arranged in the order of the change in comparison with each corresponding position of each read data, the relative positional relationship is replaced with three-dimensional coordinates and sequentially connected, and the shape of a fixed broken line, That is, a “line” is generated in which “unstable reproduction element points” that are elements that hinder the same resolution due to individual differences are arranged in the order of unstable reproduction intensity. As described above, this “line” can always be generated if the object is read as an image regardless of the resolution of the sensor of the scanner. The order of generating this “line” includes information specific to the object. A certain universality based on the uniqueness of appears.

  Such a line can be obtained by scanning the same object at least twice and specifying the rank of the difference because the generation rank causes the unstable reproduction. The basic principle of the present invention is that even if the same object is scanned, the same result cannot be obtained due to physical (light wavelength / physical properties) conditions. Even if possible, for the problem that identification is difficult in mass-produced products such as printed matter, it is not a comparison of individual differences itself, but a ranking relationship of instability of resolution reproducibility for each read sequence This can be realized by a more specific method that enables individual identification by observing. This is derived from the physical presence of the object and the fact that there is also a physical image sensor alignment. In other words, since the arrangement of the image sensors is physically constant, the possibility that the internal configuration of the image sensor will collapse is lower than at least the possibility of occurrence of the deviation of the scanning position even if the deviation of the scanning position for each scanning occurs. Therefore, it can be considered that the arrangement is maintained in the configuration state. Accordingly, the positional relationship is maintained in the “line” obtained from the image data (array value) of the scanning area. That is, the value of each corresponding element of the image sensor changes with each scanning, but the order of the difference is maintained. Also, for example, in the case of different printed materials, even if this “line” shape is scanned in order, the probability of matching is low, and the probability of being described later depends on the number of “unstable reproducible element points”. It is possible. This “line” is also the order of instability reproducibility because it reproduces instability resolution, that is, the instability of values obtained in each corresponding element of image data. Here, since it is sufficient that this “line” can express the degree of reproducibility of instability, the same object can be scanned continuously at least twice and determined by the rank of the difference. That is, for example, the image data obtained by scanning twice in succession is slightly different depending on the (fine) feature of the object smaller than the resolution level of the scan, and the difference is processed by the concept of rank, for example. By doing so, it is possible to grasp the fine features of the object. Here, the two scans are usually performed continuously, that is, within a certain period of time. However, if the time interval is too long, the phenomenon occurring in the object changes greatly, and the above-described principle does not work. Because there is a possibility. For example, when the power of the imaging device is turned off after the first shooting, the imaging conditions change greatly due to irreversibility such as charge amount in the image sensor and sensitivity to the light quantity, and it is difficult to adjust. There is a possibility. Therefore, it is preferable that a plurality of shootings be performed within a certain close time, or the shooting is continued in a state where substantially the same shooting conditions are maintained. In this way, according to the principle of the present invention, a phenomenon that does not require any new equipment and cannot be detected by a single resolution process is estimated from a plurality of image data. The phenomenon can be acquired.

  Applying the principle as described above, one of the processes for extracting information unique to the object will be described below. Referring to FIG. 5, as described with reference to FIG. 4, the object 502 is scanned continuously a plurality of times under the same configuration conditions by the camera 501 equipped with a CCD or CMOS image sensor, and is resolved as digital data. . FIG. 20 is a flowchart showing processing for registering individual difference information according to the present embodiment.

  Since different imaging devices may be used during authentication, position correction, resolution, sensor light-receiving color temperature correction, white balance, etc. used for calculating individual difference information in advance using standard teaching data Correction, ISO sensitivity correction, lens distortion correction, chromatic aberration correction, condensing correction, illumination light amount correction, spectral correction, conversion according to the type of imaging device, etc. In addition, a filter configured to apply the correction value is applied to the resolved image data. As a result, each pixel value of the image data is converted, and the image data can be read under a certain condition regardless of the configuration of the imaging device, but the correction value data is stored separately in the storage device without being limited thereto. Or other image data can be generated by methods known in the art. For example, even when using an imaging device with different resolutions at the time of registration and at the time of authentication, the resolution is recorded in advance, and the resolution of the image data with a low resolution is increased (the numerical value itself remains the same as the resolution Can be processed by converting to the same resolution.

  The actual object images 512 and 522 are the same for the first time and the second time, but the image data 511 resolved for the first time and the image data 521 resolved for the second time are compared. It is understood that the resolved array data is different due to a difference in the scanning position. Specifically, it is an array of image data 511 at the time of the first reading (1,1: 112,28,93) (2,1: 17,30,104) (3,1: 39,36,107) (4, 1: 221, 38,108) ..., the second sequence of image data 521 is (1,1: 110,45,99) (2,1: 28,24,167) (3,1: 35,41,100 ) (4,1: 209, 43,111) ... a little different.

  Next, a resolution difference is extracted from a plurality of image data obtained by reading a plurality of times as described above. Specifically, as shown in FIG. 6, the difference image 603 is extracted from a plurality of image data 601 and 602 in this case, and the resolution difference that occurs when the image is resolved a plurality of times is revealed as an optical information difference. Capture individual differences. When actually calculating the resolution difference, it is meaningless to compare different points. Therefore, the conversion value of movement and rotation is calculated to superimpose the positions where each image data maps the same part of the subject. . Using each conversion value calculated for the image data, the pixel array of each image data is overlapped, and the light information of the pixel at the overlapping position is compared to calculate the light information difference. Any method known in this technical field can be used to calculate the conversion value for superimposing the positions.

  For example, such a conversion value can be calculated by extracting some characteristic points from each image data and performing an affine transformation so that the feature points coincide with each other. I can't. In this embodiment, in order to compare a plurality of image data, the points corresponding to each other are specified, how to move the image data in parallel or the like in order to overlap those points, Find a conversion value that indicates whether it is necessary to apply rotation, actually change the image data, do not superimpose it, identify the corresponding point using the conversion value, and find the pixel at the corresponding position. The difference between the optical information is taken because this may cause subtle deviations when the image data is actually converted and superimposed. Therefore, if such a shift can be avoided, the above processing can be performed by taking a difference after actually converting the image data.

  FIG. 22 is a diagram for explaining a detailed example of the calculation of the optical information difference according to the present embodiment. As described above, in the present embodiment, the image data obtained by scanning the object a plurality of times is converted to superimpose the position where the object is mapped, and the light information of each pixel corresponding to each other is compared. Although the optical information difference is obtained, specifically, it can be executed by a process as shown in FIG. First, conversion values are obtained so that the positions where the subject is mapped are superimposed on the image data A2201 and the image data B2202. Next, as a result of converting each of the image data A and B, the light information of the corresponding pixels (in this example, the RBG component) is compared, and the difference is set as the light information difference of the corresponding physical position. More specifically, virtual image data P2203 is formed as shown in FIG. 22 (therefore, image data P has a smaller size than image data A and B), and an image is obtained from each point of image data P2203. The corresponding points of data A and B are obtained. As shown in enlarged views A2205 and B2206 of the image data, the image data A and B do not completely coincide with each other at the pill cell level even if they are overlapped. Overlapping with multiple pixels.

  Therefore, in this embodiment, for example, in order to obtain the optical information difference, the optical information difference is obtained by taking the difference between the optical information value of one pixel and the optical information values of a plurality of pixels overlapping with the pixel. Although it calculates, it is not restricted to this, It can calculate by any method known in this technical field. In this case, a connection point to be described later can be a midpoint 2211 of the center point 2209 of one pixel and the center 2210 of the center point of a plurality of corresponding pixels. As a result, the obtained optical information difference is color-represented as an image. However, in reality, such an intermediate process does not exist, and a human balloon graphic is directly obtained from the optical information difference data. The sum of these component values is the optical information difference for the image data A and B.

  In this way, the resolution difference can be expressed as a phase difference at the overlapping position of the corresponding images, for example, as a sum of absolute values of differences for each RGB component at each overlapping position as shown in FIG. . Here, in order to obtain the difference between the two image data 601 and 602, it is necessary to specify the overlapping position of the pixels of each image data as described above. The overlapping positions are specified, that is, the corresponding positions are specified so that the light information of the same part of the subject can be compared. As a result of such superposition, each pixel value corresponding to the physically corresponding position of the A and B image data 601 and 602 is compared for each RGB component, and the color representation of the difference as an optical information difference is an optical information difference. C. That is, array A (1,1: 112,28,93) (2,1: 17,30,104) (3,1: 39,36,107) (4,1: 221,38,108) ... of image data 601 and B (1,1: 110,45,99) (2,1: 28,24,167) (3,1: 35,41,100) (4,1: 209, 43,111) ... is converted into the actual distance, A '(0.009,5.391,0.0: 112, 28, 93) (0.027,5.391,0.0: 17, 30,104) (0.045,5.391,0.0: 39, 36,107) (0.063,5.391,0.0: 221, 38,108) ...・ (Coordinate unit is μm) and B '(0.011,5.392,0.0: 110, 45, 99) (0.031,5.392,0.0: 28, 24,167) (0.048,5.392,0.0: 35, 41,100) (0.066,5.392 , 0.0: 209, 43, 111) ... A '(+) B' = CB '(0.011, 5.392, 0.0: 110, 45, 99) (0.031, 5.392, 0.0: 28, 24, 167) (0.048, 5.392,0.0: 35, 41, 100) (0.066, 5.392, 0.0: 209, 43, 111).

  The resolution difference obtained as described above has unique attributes for the target because the characteristics unique to the target are reflected, and the results obtained by further reading and extracting the same target multiple times There is a certain correlation in. In order to make this relationship clearer, that is, to enable authentication of the object, unstable reproducible element points are connected from the obtained resolution difference. In other words, if the unstable reproduction element points are arranged from the resolution difference to the unstable reproduction intensity, that is, the middle point of each overlapping position in the above example in descending order of the optical information difference value, this order is the resolution of the object to be imaged. Reproducibility for processing, that is, a certain correlation with respect to the same object. The numerical value of the unstable reproduction intensity, that is, the optical information difference itself of each image data is affected by the error at the time of resolution more than the individual difference, and thus is not suitable as information unique to the object. Therefore, the numerical value of the unstable reproduction intensity itself is ignored, and this point is connected in order in the order in which only the position information is retained and rearranged. The position information corresponds to the case where the ratio of the size of the subject and the image data changes due to the difference in the imaging device configuration at the time of registration and at the time of authentication. Using the optical system configuration information such as size, lens configuration, and distance between subjects, calculate the physical size of the subject corresponding to the pixel size of the image data, and find the midpoint of the physical mapping position of the center point of the overlapping pixels. To use.

  Specifically, as shown in FIG. 7, the physical position of the subject that is mapped to the pixel center point from the overlapping pixel position from the converted values of the respective image data that are superimposed to obtain the optical information difference in FIG. 6. It is converted into (three-dimensional position), and the midpoint is adopted as position information of each connecting point constituting a connecting line having a polygonal line shape representing individual differences. The light information differences 701D (25) (80) (16) (20)... Are rearranged in the order of the light information difference values, and the corresponding pixels of the two image data used for the calculation of the difference values from the head are calculated. Physical position A '(3.353, 0.072, 0.0) (1.771, 0.960, 0.0) (5.295, 5.245, 0.0) (4.276, 0.928, 0.0) ..., B' (3.357, 0.072) , 0.0) (1.773, 0.963, 0.0) (5.292, 5.247, 0.0) (4.275, 0.927, 0.0) ... L (3.355, 0.072, 0.0) (1.772, 0.9615, 0.0) (5.2935, 5.246, 0.0) (4.2755, 0.9275, 0.0) ... As shown in an image 702, a connection line representing an individual difference obtained from the image data of the scanning region is generated. Here, such a connection line or an image showing the connection line is referred to as a human balloon line drawing. As will be described later, the obtained human balloon line drawing has a reproducible order that cannot be obtained by a single resolving process, and represents a characteristic unique to the object.

  Through the above processing, it is possible to obtain the connection data of unstable reproduction element points, which is individual difference information extracted from individual differences of data obtained by scanning the target. In this embodiment, this individual difference information is The individual difference information and the extraction method described above are only one method of the present invention, and are reflected in the resolution difference of data obtained as a result of reading the target as an image. Any information indicating information unique to a target to be detected can be used as an individual difference information in the present invention by being extracted by any method known in the art. In the above description, the values of the RGB components are used. However, as can be understood from the principle of the present invention, the light in each scanning unit obtained by optically scanning the surface state of the target with a constant resolution. Any data can be used as long as it is an array of data of the same intensity.

<Target authentication process>
Hereinafter, referring to FIGS. 8, 10 and 23, the object is authenticated from the individual difference information obtained above, that is, whether the object such as a printed matter is the same as the object from which the individual difference information has already been extracted. One method for determination using difference information will be described. First, as a premise, with reference to FIG. 8, when the reproducibility of the ranking relationship of unstable reproduction element points, that is, the position information of the connection points arranged in order of the intensity of the optical information difference is for the same object, Explain that it has a certain similarity. FIG. 21 is a flowchart showing the authentication processing of this embodiment.

  As shown in FIG. 8, the object 502 is obtained by extracting individual difference information as represented by a connecting line shown in an image 801 from image data read in advance by a scanner or the like a plurality of times and storing it in some storage means such as a database. Has been. Here, the target 502 is further scanned with the camera 501 a plurality of times, and as shown in an image 802, optical information at a position corresponding to the pixel where the connected positions of the human balloon lines are overlapped according to the human balloon lines drawn in the image 801 extracted in advance. The difference is obtained, and specifically, the connection positions are compared according to the connection order. As a result of this comparison, since the individual difference is reflected in the resolution difference, this human balloon line drawing is not completely the same as shown in FIG. 8, but for example, the point on the human balloon line drawing shown in the image 801 and the image 802 With this point, the order of the unstable reproducible element points has a certain reproducibility, and a human balloon line drawing 803 can be obtained. For this reason, between the human line drawings obtained for the same object, there is reproducibility in the connection order of the corresponding connection positions, and therefore the identity can be determined based on the degree of the order matching.

  Here, in this embodiment, the matching of the objects is determined by comparing the positions of connection points arranged in descending order of the intensity of the optical information difference. Various pattern comparison methods, that is, variations of the connection comparison method, such as connecting and comparing in ascending order of stable reproduction element strength, are conceivable. In the case of using the present invention, it is possible to limit the authentication target by using the parameters and comparison method optimal for the imaging target. Call. The connected eyeglasses use various parameters as described below, and the combination method of the human line drawing is determined by this combination. An example of connected eyeglasses is shown in FIG. FIG. 18 and FIG. 19 show the connection method of the human line drawing line drawing. FIG. 18 shows an index formula, and FIG. 19 shows a sequence system. Parameters used in connected eyeglasses: ((rank of unstable reproduction element strength (forward / reverse)) (lowest connected distance) (threshold of intensity) (change intensity) (number of connections) (scanning region) + α).

In addition, an example of the format of the human balloon line drawing corresponding to this is shown below.
"Hitofudaki line drawing" index formula:
((Start point 3D coordinates XYZ) (start point 3D coordinate vector) (distance)) + number x ((3D vector) (distance))
“Hitokigaki Line Drawing” Sequence Formula: Number x 3D Coordinate XYZ
In other words, in addition to the descending order detailed below, ascending order, skip the unstable reproducible element points less than the minimum connection distance, do not connect the unstable reproducible element points below the intensity threshold, skip the points whose intensity change is below a certain level Various comparison methods such as connecting to the maximum number of connections, connecting only within the range of the definition area, and the like are possible. In addition, it is possible to select the most suitable linked eyeglasses depending on the material on which the object appears and the expression method of the object, that is, what printing method using what ink, and the actual authentication process. in, it is necessary to define whether to use the advance what connection colored glasses. For this reason, in addition to the individual difference information, a database or the like for storing the individual difference information described above stores, as ancillary information, what kind of connected glasses is used or is associated with a certain method. Thus, the authentication process can be performed using the individual difference information under the optimum conditions. In addition to the connected eyeglasses, the accompanying information can also include other information effective for determining the identity of the object using the individual difference information.

  Next, a method for specifically determining the identity, that is, determining the reproducibility of the order of unstable reproducible element points will be described. A specific processing example will be described with reference to FIG. 23. First, the authentication target match determination is performed by first tracing the human balloon line drawing 2301 extracted and registered to the connection order so that each connection position is read and overlapped at the time of authentication. This is done by checking whether the value of the light information difference of the pixel corresponding to the physical surface of the pixel of the combined image data 2302 has a predetermined relationship. Here, the coincidence determination is performed according to a comparison method defined by the connected glasses. For example, if the matching polygon shape is a three-point shape, it is determined whether or not a ranking relationship corresponding to each of the three points from the beginning of the connection of the line-up line drawing is established. Count polygons. Next, the correspondence is determined in order, and the search is performed up to the end point of the human line drawing. As a result, when the coincidence shape is stored in the coincidence polygon array and the number of coincidence shapes is equal to or greater than the coincidence determination threshold, it is determined that the authentication is successful. Of course, it is also possible to search for matching polygons up to a predetermined number of counts and terminate the processing when authentication is successful when the number of counts is exceeded.

  In this embodiment, without generating a connection line at the time of authentication, the image data scanned at the time of authentication along the registered connection line is traced and the number of matching ranks is counted. The connecting lines can be similarly compared to determine identity. Specifically, referring to FIG. 9, polygon matching will be described. That is, as one method for comparing the human balloon line drawings, the correspondence is determined in order from the start point of the human balloon line drawings, and the identity is determined by calculating a coefficient that a matching polygon exists when the three points match. The matching by polygon is also performed by comparing the human balloon line drawing shown in the image 901 of individual difference information extracted and registered first with the human balloon line drawing shown in the image 902 of individual difference information extracted again at the time of authentication. The two human balloon lines are compared according to any one of the connected iro-glasses, such as the above-mentioned connection order, and are determined to match, and a triangle is formed from three consecutive points (a matching polygon shape exists). As shown in the figure, it is determined whether or not a ranking relationship corresponding to each of the three points from the top of the line drawing is established, and if a corresponding relationship of three points is established, the matching polygons are counted. In this way, the correspondence is determined in order, and the search is performed up to the end point of the line drawing. As a result, when the matching shape is stored in the matching polygon array 904 and the number of matching shapes is equal to or greater than the matching determination threshold, it is determined that the authentication is successful.

<About the authentication probability>
Corresponding to the physical size of the authentication target resolved at the resolution of the digital photographic device and the size of the image data collected by the focus (pixel size), it is simple without duplicating unstable (resolution) reproduction intensity points. The following describes how the probability of matching when all pixels are connected in order of intensity changes when the score is reduced. Here, the number of image data to be read is two and the resolution is the same.

Based on the authentication method described above, the human balloon line drawing finally becomes the data of the relative relationship only in the coordinate position, and the probability that the human balloon line drawing coincides by chance is expressed by the following formula. If the instability of unstable reproducible element points is uniform, assuming that the size in the x, y, z direction of the cropped image is X _t , Y _t , Z _t , the degree of freedom of each element point is , X _t × Y _t × Z _t , where N is the total number of unstable reproduction element points, the probability that any one point on the image data is an unstable reproduction element point is

It becomes. The probability of matching this one point when matching a human line drawing is the size of the image at the time of registration is X _r , Y _r , Z _r, and the physical position corresponding to the pixel is different from that of the image data. If you think that it falls within the corresponding pixel area,

If the matching polygon shape is a three-point shape, the probability that the first three points all match is

It becomes. When this is applied to the second and subsequent three points, the authentication matching probability R when the number of matching shapes is P is

  In addition, when the number N of unstable reproducible element points to be connected as a human line drawing is all pixels of the image data,

It can be seen that the probability of coincidence increases as the number of connections is reduced. On the other hand, in order to increase the accuracy of authentication determination, it is necessary to use a condition that the number of connections is equal to or greater than a certain level. Also, if the number of matching shapes that match continuously increases from 3 points, the number of multiplications in {} shown in [Equation 5] increases, so the probability of coincidence decreases, and this also improves the authentication accuracy. It is possible to increase. Furthermore, it is self-evident that in [Equation 3], the probability that only the first and third points match is

It turns out that the probability of coincidence increases by chance.

  In the example described above with reference to FIG. 9, it is determined that the authentication has failed if there is no more match, with a match polygon three-point shape, with the number of match polygons being five as a threshold. Note that an optimum value can be determined as the threshold value depending on the characteristics of the authentication target, that is, what kind of material is printed by what method.

<Example>
FIG. 1 is a system configuration diagram of an authentication system according to an embodiment of the present invention. The authentication system 100 according to the present embodiment can connect a scanner 110 for reading an object and a display 120 for displaying a result to an apparatus having an interface with the CPU 101, ROM 102, RAM 103 and various external devices. Of course, in the present invention, since some kind of scanner is necessary to read the object, it can be built in the apparatus, or can be provided with a printer that outputs the result as another function, or can be connected to a network and used as a database. You can also exchange necessary data. Specifically, it can be a dedicated device for authentication, and various system configurations known in this technical field such as connecting a scanner to a camera-equipped mobile phone, a notebook computer, or a personal computer are possible.

  FIG. 2 is a functional block diagram of an embodiment of the present invention. In the present embodiment, a software program is executed on the CPU 101, and various functions necessary for achieving the present invention are processed. For example, each function block as shown in FIG. 2 can be processed. That is, an image reading unit 201 that scans a target and extracts an image, an individual difference information calculation unit 202 that calculates individual difference information from the obtained image, an individual difference information comparison unit 203 that compares the calculated individual difference information, and Processing is executed by the determination unit 204 that determines whether or not to authenticate finally. In the present embodiment, the processing is executed with the functional blocks as shown in FIG. 2, but the present invention is not limited to this, and the functional blocks are further divided, or a plurality of functional blocks are integrated and the processing is executed with different functional blocks. You can also.

  The processing of this embodiment will be described with reference to FIGS. Here, in the present embodiment, a certain target is compared with a target that is registered or assumed in advance, and authentication or determination is made as to whether or not they match. It is assumed that the difference information is calculated in advance and stored in any one of them. For example, the individual difference information calculated in advance may be stored in a memory inside the device such as the ROM 102 or the RAM 103, or may be acquired via an external storage device or a network.

  FIG. 3 is a flowchart showing authentication processing according to an embodiment of the present invention. The image reading unit 201 reads the target using the scanner 110 and outputs image data (S301). Since the scanner 110 used for reading is usually different from the reading device used for calculating individual difference information in advance, the image data obtained in this way is subjected to a certain correction and read difference. (S302). In this embodiment, position correction of an imaging device or the like used for calculating individual difference information using standard teaching data in advance, sensor light reception color temperature correction, white balance correction, ISO sensitivity correction, lens distortion correction, Obtain correction values for standardizing resolution such as chromatic aberration correction, condensing correction, illumination light quantity correction, spectral correction, and conversion according to the type of imaging device, and use these correction values to solve the filter. By applying to imaged image data, each pixel value of the image data is converted, and image data can be read under certain conditions regardless of the configuration of the imaging device, but the storage device is not limited to this. The image data can be generated by exchanging correction value data separately, such as by storing the image data, or by other methods known in the art.

  Next, it is possible to authenticate using a part of the target image. FIG. 11 is a diagram for explaining extraction processing of a specific application example of an embodiment of the present invention. As shown in FIG. 11, a target is a display label attached to some product, which is a kind of printed matter. An image is read twice for a predetermined area of the label as a target, and a human line drawing is generated from the two obtained images (S303). Next, the pre-calculated human balloon line drawing is read from the memory or the like (S304), and compared with the human balloon line drawing generated in S303 as shown in FIG. 12 (S305). FIG. 12 is a diagram for explaining the coincidence determination process of a specific application example of the embodiment of the present invention. The comparison is performed using, for example, the above-described connection comparison method using polygons. If a match is recognized with a certain probability (S306), a successful authentication is notified (S307), and if not, an authentication failure is notified (S308). ). In this embodiment, authentication is determined using the printed portion of the display label. However, the present invention is not limited to this, and any method known in this technical field can be used. For example, it is possible to press a seal stamp on a label in advance as a mark and use it as a target, and such data can be used if color information can be read from any surface of the container.

  In this example, individual difference extraction and individual difference authentication processing was performed using a pharmaceutical label with the following devices and settings. The label to be certified is a film base label with a vertical and horizontal length of 44.0 [mm] x 74.0 [mm], and no special ink is used or special processing is performed during printing. The format of the line drawing is sequence, the imaging equipment is a CCD digital camera with 10.75 million pixels, the focus is manual focus, the sensitivity is ISO 100, the aperture is F = 8, the shutter speed is 1/20 [s], and the captured image size (W × H) was set to 3872 × 2592 [pixel], and the number of shootings was two.

Specifically, individual difference features were extracted by the following procedure.
(1) As shown in Fig. 11, set the target label so that the shooting range is 15.5 [mm] x 23.2 [mm] in the vertical and horizontal directions, and shoot twice under the above conditions. Get.
(2) The two image data obtained are trimmed to a size of 8.7 [mm] × 8.7 [mm], respectively, and the optical information difference D (x, y) defined by the following equation is obtained from the two image data. . Let image data 1 be g1 (x, y) and image data 2 be g2 (x, y). X and Y are a set of x and y coordinates of the image data.

However, g1 _R (x, y) represents the R component of the RGB value of g1 (x, y), and so on.
(3) Generate a starfish line drawing under the following conditions. The connection order of unstable reproduction elements is descending order connection, the minimum connection distance is 62 μm, the threshold value of unstable reproduction intensity is 10.0, and the number of connections is 250. From the optical information difference obtained above, the human line drawing shown in FIG. 1201 is generated.
(4) The obtained human balloon line drawing 1201 is stored in the database together with the accompanying information (setting of the used device, human balloon line drawing generation condition, etc.) as individual difference features to be authenticated.
(5) The object is imaged twice again under the same conditions as in (1) above, and a human balloon line drawing 1302 as shown in FIG. 12 is generated in the same manner.
(6) Based on the incidental information, the human balloon line drawing is referred to the database, and the human balloon line drawing 1201 called from the database and the human balloon line drawing 1302 obtained here are matched from the head of the connection. Matching conditions were set as follows. The matching shape is a three-point shape. Matching minimum shape number of teeth, 4 polygons.
(7) The matching result is an authentication result 1301 as shown in FIG. As shown in FIG. 12, since the number of matched polygons 1303 is 26 and exceeds the condition of 4 polygons or more, it is determined that the same target is the same as the registered human balloon line drawing and it is confirmed that the authentication is successful. it can.

  13 to 16 are diagrams for explaining another specific application example of the embodiment of the present invention. The example shown in FIG. 13 is for tablets. In general, a tablet is printed on the surface with a mark or name indicating what the tablet is. Using the present invention for this printed part, each tablet can be identified individually. For example, if individual difference information is created at the time of manufacture, it is possible to uniquely identify which tablet in which lot of which factory and when it is manufactured, and various uses can be expected. Further, if color information can be read from the tablet surface itself, individual identification can be performed using the present invention even for a tablet on which no mark or the like is printed.

  A specific authentication process will be described in the same manner as the above-described pharmaceutical label. As shown in FIG. 13, the tablet used was a round tablet with a gray stamp on a white background with a length and width of 7.5 [mm] × 7.5 [mm]. This is a commercially available drug that can generally be purchased. The imaging equipment was the same as that used in the above-mentioned example of the pharmaceutical label, the shooting conditions were 1/50 [s] only for the shutter speed, and the other values were set the same. Also, the number of times of shooting was the same two times.

Under the above conditions, individual difference features were extracted as follows.
(1) The target tablet is set so that the shooting range is 15.5 [mm] × 23.2 [mm] in the vertical and horizontal directions, and is shot twice under the above shooting conditions to obtain image data for two sheets.
(2) The two image data are trimmed in a size of 4.2 [mm] × 3.5 [mm] in length and width, respectively, and the digital reproduction difference is obtained from the two image data in the same manner as the above-described pharmaceutical label.
(3) The connected cardboard glasses have a human balloon line drawing format: sequence formula, connection rank: descending order, minimum connection distance: 63 μm, threshold of unstable reproduction intensity: 20.0, number of connections: 250, thereby obtaining a human balloon line drawing. FIG. 13 shows the obtained human balloon line drawing.
(4) The obtained human balloon graphic and incidental information are stored in the database, and the authentication registration is completed.
(5) The subject is imaged twice again under the same conditions as in (1) above, and a human balloon line drawing is extracted.
(6) Similarly, matching is performed from the head of the connection between the human graffiti line drawing called from the database and what has been obtained. Matching conditions were set to a matching shape: a three-point shape and a minimum number of matching determination shapes: 5 polygons.
(7) The result of matching is shown in FIG. The number of matching shapes is 13, which indicates that the authentication is successful.

  The example shown in FIG. 14 is an example in which the present invention is applied to a label such as sake or wine. However, in order to define a specific part of the label, for example, a part of a seal or a seal can be used, and a reading place is previously stored. It is possible to easily specify the reading area without specifying the details. Alternatively, for use in authentication according to the present invention, an authentication process can be performed by reading an image displayed by pressing a seal stamp.

  In the example shown in FIG. 15, the printed portion of the retort food is used. In this way, even if the object is printed on any material by any method, it can be read by a scanner or the like to generate an image. If possible, it can be used in the present invention.

  In the example shown in FIG. 16, the present invention is used to identify a picture. However, since the whole is a printed matter, the present invention can be applied to any portion. Further, even in such an object that is severely deteriorated over time, the connection points that coincide with each other at a certain rate are retained without being lost, so that the authentication accuracy can be maintained.

  In the above-described embodiments, it is necessary to specify which area is used for authentication except for tablets, but it is known in the technical field, for example, it is determined in advance as the right end part, It can be specified by separately transmitting information on the designated area.

Claims (8)

The optical information difference at the corresponding position of the plurality of first digital data obtained by resolving the authentication target by scanning with the first digital imaging means is calculated, and corresponds to the pixel size of the plurality of first digital data. A connection line generated by connecting points, which are calculated based on the physical size of the subject, corresponding positions of the plurality of first digital data as connection points in a predetermined order of the calculated optical information difference, Means for storing as individual difference information for uniquely identifying a target;
The corresponding position of each pixel array of the plurality of second digital data obtained by resolving the authentication target by scanning of the second digital imaging means is specified, and the connecting line stored in the storing means is determined. Reading, tracking the position of the connection point on the read connection line so as to follow a polygonal line, and the optical information of the pixel at the corresponding position of the plurality of second digital data corresponding to the position of the connection point Means for calculating a difference, and determining that the optical information difference of the plurality of second digital data is in a predetermined relationship with respect to the order of the connecting lines, and determining that the authentication is successful. An authentication system characterized by that.   The authentication system according to claim 1, wherein the resolutions of the first digital imaging unit and the second digital imaging unit are different.   The authentication system according to claim 1, wherein the first digital imaging unit and the second digital imaging unit scan an authentication target a plurality of times continuously.   According to another aspect of the present invention, there is further provided means for performing correction for standardizing a difference in image data due to a difference in configuration conditions between the first digital imaging unit and the second digital imaging unit based on preset information. The authentication system according to any one of claims 1 to 3.   The means for storing further stores a connection comparison setting including a parameter for comparison of the connection lines and a comparison method, and the means for determining uses the connection comparison setting on the read connection line. The position of the connection point is traced so as to follow a polygonal line, and the optical information difference of the second digital data has a predetermined relationship defined in the connection comparison setting with respect to the order of the connection lines. 5. The authentication system according to claim 1, wherein the authentication system determines that the authentication is successful.   The connection line is generated by connecting points formed by corresponding positions of the plurality of first digital data in descending order of the optical information difference of the plurality of first digital data. The optical information difference of the second digital data is determined to be successful if it is in a predetermined descending order with respect to the order of the connecting lines. The described authentication system. Scanning the authentication object a plurality of times by the digital imaging means and resolving the plurality of first digital data;
Calculating an optical information difference at a corresponding position of the plurality of first digital data;
A physical size of a subject corresponding to a pixel size of the plurality of first digital data is calculated, and a point formed by a corresponding position of the plurality of first digital data is calculated as a connection point based on the calculated physical size. Extracting a connection line generated by connecting the connection points in a predetermined order of the calculated optical information difference as individual difference information;
An authentication information registration method comprising: registering the extracted connection line in a storage means. The optical information difference at the corresponding position of the plurality of first digital data obtained by resolving the authentication target by scanning with the first digital imaging means is calculated, and corresponds to the pixel size of the plurality of first digital data. A connection line generated by connecting points, which are calculated based on the physical size of the subject, corresponding positions of the plurality of first digital data as connection points in a predetermined order of the calculated optical information difference, As an individual difference information that uniquely identifies a target, in an authentication system having means for storing,
Scanning the authentication object a plurality of times by the second digital imaging means and resolving the plurality of second digital data;
The corresponding position of each pixel arrangement of the plurality of second digital data is specified, the connection line stored in the storage means is read out, and the position of the connection point on the read connection line is a broken line The pixel light information difference at the corresponding position of the plurality of second digital data corresponding to the position of the connection point is calculated, and the light information difference of the plurality of second digital data is calculated as follows: An authentication method comprising: a step of determining that the authentication is successful when it is determined that the connection line order is in a predetermined relationship.