JP6683033B2 - Biometric authentication device, biometric authentication method, and biometric authentication program - Google Patents

Description

  The present invention relates to a biometric authentication device, a biometric authentication method, and a biometric authentication program.

  In modern society, identity verification is required in various situations, and biometric authentication technology that uses part of the human body such as fingerprints, veins, and facial images has become widespread in recent years as one method of accurately verifying identity. ing.

  This type of biometrics technology can be used to access various services, such as determining whether a person can enter or leave a restricted area where only authorized people can enter, determining whether to log in to a personal computer, and confirming users in online transactions. It is already widely used depending on whether or not.

  Fingerprint authentication and the like have been widely known as biometric authentication technology, but they have a drawback in that there is a risk that information such as fingerprints appearing on the surface of the body is forged. On the other hand, a technique that uses subcutaneous information such as a vein pattern cannot be easily forged, and is therefore put into practical use as a highly secure authentication technique.

  For example, when the palm is irradiated with near-infrared rays and an image in the near-infrared wavelength region is captured, an image in which the subcutaneous veins of the palm appear black can be obtained. When the image pattern information of veins flowing like a mesh through the palm is extracted, the image pattern information is different for all people. It is said that the venous pattern does not depend on genetic information, and its vascular network is determined by various accidental factors during body formation during fetus development. Therefore, as described above, the vein pattern is different for all persons, and also different for twins. Furthermore, the vein patterns of the left and right hands of the same person are different. Therefore, by preliminarily recording the image pattern information in a database, an IC card, or the like, and comparing the image patterns in various situations where individual authentication is required, and calculating the pattern similarity, the identity verification is performed. It can be used for such purposes. The vein pattern can be acquired by analyzing an image obtained by imaging a living body in a wavelength band of light that passes through the skin, such as near infrared rays. In a living body, a portion through which a vein has a higher absorption rate of near infrared rays than a portion through which a vein does not pass. Therefore, the vein running pattern can be obtained by extracting the portion that appears black in the image captured with the near-infrared wavelength.

  Basically, vein patterns can be photographed on any part of the human body, but biometric authentication using palm vein patterns is widely used in terms of the amount of information related to usability and authentication accuracy. .

  In addition, as one of biometric authentication techniques using a vein pattern, there is known a technique that uses information on the lightness and darkness of a blood vessel in addition to a traveling pattern when comparing vein patterns (for example, see Patent Document 1). .

  Further, in the above-mentioned biometric authentication technology, for example, there is known a technology that provides more secure authentication accuracy by a method of combining authentication using a fingerprint with authentication using a vein pattern (for example, , Patent Document 2).

JP 2012-098974 A JP, 2004-054698, A

  In vein authentication using a vein travel pattern, it is determined whether the person is the other person or another person based on the degree of coincidence between the two-dimensional vein travel patterns. However, looking only at the two-dimensional traveling pattern, there are cases where there are blood vessels that follow similar traveling paths, although they are actually different blood vessels. Therefore, in vein authentication that uses a two-dimensional travel pattern, different travel patterns may actually be erroneously determined to match. In the authentication method that compares the travel patterns of blood vessels, such an error may cause deterioration in authentication accuracy.

  In one aspect, the present invention aims to improve authentication accuracy in biometric authentication using a vein pattern of a living body.

  In one aspect of the present invention, the biometric authentication device includes a blood vessel image extraction unit, a blood vessel pattern creation unit, a stability determination unit, and an authentication unit. The blood vessel image extraction unit acquires a biometric image including an image of a blood vessel traveling in the living body and extracts a two-dimensional blood vessel pattern image from the biometric image. The blood vessel pattern creation unit adds the blood vessel width or brightness to the image of the two-dimensional blood vessel pattern as a dimension in the depth direction based on the blood vessel width or brightness at each position in the running direction of the blood vessel in the two-dimensional blood vessel pattern. Create a three-dimensional blood vessel pattern. The stability determination unit, the cross-sectional shape of the luminance in the direction of the blood vessel width at each position in the running direction of the blood vessel in the image of the two-dimensional blood vessel pattern, the relationship between the extraction stability of the blood vessel and the cross-sectional shape of the luminance of the blood vessel. , The extraction stability at each position in the running direction of the blood vessel is determined. The blood vessel extraction stability represents the degree to which blood vessels are extracted from a biometric image. The authentication unit determines whether the authentication is possible based on the biometric data including the three-dimensional blood vessel pattern and the extraction stability.

  According to the above aspect, the authentication accuracy of biometric authentication using the vein pattern of the living body is improved.

It is a figure which shows the functional structure of the biometrics authentication apparatus which concerns on 1st Embodiment. It is a figure which shows the functional structure of a data preparation part. It is a figure which shows the functional structure of the authentication part in the biometrics authentication apparatus which concerns on 1st Embodiment. It is a figure which shows the structural example of a template database. It is a flow chart explaining the contents of the processing which a biometrics device performs. It is a flow chart explaining the contents of data creation processing. It is a flow chart explaining the contents of three-dimensional pattern creation processing. It is a flowchart explaining the content of extraction stability determination processing. It is a figure which shows the image of the blood vessel extracted from the biological image, and the example of a two-dimensional blood vessel pattern. It is the figure which expanded the area | region AR1 and area | region AR2 of FIG. It is a figure which shows the example of the three-dimensional-ized blood vessel pattern. It is a figure which shows a part of blood vessel in a biometric image. It is a figure which shows the three-dimensional blood-vessel pattern corresponding to the blood vessel of FIG. It is a figure explaining the example of the determination method of extraction stability. It is a flow chart explaining the contents of attestation processing. It is a flowchart (the 1) explaining the content of the collation process of input biometric data and a template. It is a flowchart (the 2) explaining the content of the collation process of input biometric data and a template. It is a flow chart (the 1) explaining the contents of rough alignment processing. It is a flow chart (the 2) explaining the contents of rough alignment processing. It is a flow chart (the 3) explaining the contents of rough alignment processing. It is a figure which shows another structural example of a template database. It is a figure which shows the functional structure of the authentication part in the biometrics authentication apparatus which concerns on 2nd Embodiment. It is a flow chart explaining a part of collation processing of input biometrics data and a template concerning a 2nd embodiment. It is a figure which shows the hardware constitutions of a computer.

[First Embodiment]
FIG. 1 is a diagram showing a functional configuration of the biometric authentication device according to the first embodiment.
As shown in FIG. 1, the biometric authentication device 1 of this embodiment includes a process selection unit 110, an image acquisition unit 120, a data creation unit 130, an authentication unit 140, and a template database 150.

  The processing selection unit 110 selects, based on an input signal from the input device 2 connected to the biometric authentication device 1, whether to perform template registration processing or authentication processing in the biometric authentication device 1 (determination). To). The input device 2 is, for example, a keyboard device, a touch panel device, or the like. The process of registering the template is, for example, a process of creating biometric data (template) for registration of a person (authentication target person) authenticated by the biometric authentication device 1 and registering it in the template database 150. In the authentication processing, biometric data of a person (authentication requesting person) who is going to be authenticated by the biometric authentication device 1 is created, and the authentication requesting person is identified based on the biometric data and the template registered in the template database 150. This is a process of determining whether or not to authenticate. The biometric authentication device 1 according to the present embodiment uses a travel pattern of blood vessels (veins) in the palm of a person as biometric data. In the following description, the biometric data used as a template is also referred to as registration data. Further, in the following description, the biometric data of the authentication applicant, which is collated with the registration data, is also referred to as input biometric data.

  The image acquisition unit 120 acquires the biometric image of the authentication applicant including the authentication target person from the biometric image capturing device 3 connected to the biometric authentication device 1. The biometric image capturing device 3 used in the present embodiment is a device that irradiates the palm with near-infrared rays and captures a biometric image including a subcutaneous vein of the palm. In the following description, the biological image capturing device 3 is also simply referred to as the image capturing device 3.

  The data creation unit 130, based on the image of the blood vessel in the biometric image acquired from the imaging device 3, the three-dimensional blood vessel pattern represented by a plurality of segments (partial blood vessels) and the extraction stability of each segment in the blood vessel pattern. Create biometric data including and. Each of the plurality of segments in the three-dimensional blood vessel pattern represents two-dimensional information (two coordinate values) indicating the position on the flattened palm in the biometric image and the feature amount of the blood vessel corresponding to the segment. It has one-dimensional information (one coordinate value). The characteristic amount of the blood vessel is, for example, the blood vessel width (the thickness of the blood vessel), the brightness of the blood vessel in the biological image, or the like. The extraction stability is a value indicating the degree to which a segment is stably extracted when the data creation unit 103 extracts a blood vessel image from a biological image.

  When the process selecting unit 110 selects the process of registering the template, the data creating unit 130 registers the created biometric data as a template in the template database 150. On the other hand, when the authentication process is selected by the process selection unit 110, the data creation unit 130 transmits the created biometric data to the authentication unit 140 as input biometric data.

  The data extraction unit 130 includes a blood vessel image extraction unit 131, a blood vessel division unit 132, a blood vessel pattern creation unit 133, and a stability determination unit 134. The blood vessel image extraction unit 131 extracts a blood vessel image from the biometric image. The blood vessel dividing unit 132 divides the extracted blood vessel image into a plurality of segments. The blood vessel pattern creation unit 133 calculates, for each segment, two coordinate values representing a position in the biometric image (xy plane) and a coordinate value in the depth direction (z direction) based on a predetermined feature amount of the blood vessel. Then, a three-dimensional blood vessel pattern is created. The stability determination unit 134 determines the extraction stability of each segment in the blood vessel pattern.

  The authentication unit 140 performs the authentication process based on the input biometric data received from the data creation unit 130 and the template registered in the template database 150. The authentication unit 140 outputs the result of the authentication process to the display device 4. The display device 4 is a device that displays the result of the authentication process, that is, whether the authentication has succeeded or failed. If the authentication is successful, the authentication unit 140 causes the authentication-requiring device 5 to execute a predetermined operation or process. The authentication-requiring device 5 is a device that performs various operations or processes including operations or processes performed only when authentication is successful. The authentication-requiring device 5 is, for example, an automatic locking device that locks the door so that the door does not open when the door is closed, and unlocks only when the authentication is successful.

FIG. 2 is a diagram showing a functional configuration of the data creation unit.
As described above, the data creation unit 130 in the biometric authentication device 1 of the present embodiment includes the blood vessel image extraction unit 131, the blood vessel division unit 132, the blood vessel pattern creation unit 133, and the stability determination unit 134. The data creation unit 130 further includes a data output unit 135, as shown in FIG.

  The blood vessel image extraction unit 131 extracts a blood vessel image from the biometric image captured by the imaging device 3. When the blood vessel image extraction unit 131 receives the information on the selected process from the process selection unit 110, the blood vessel image extraction unit 131 acquires the biometric image captured by the imaging device 3 via the image acquisition unit 120. After that, the blood vessel image extraction unit 131 extracts a blood vessel image from the acquired biological image.

  The blood vessel dividing unit 132 divides the extracted blood vessel image into a plurality of segments (partial blood vessels), and calculates, for each segment, a two-dimensional coordinate value representing the position of the segment in the xy plane of the virtual three-dimensional space. The virtual three-dimensional space is a three-dimensional orthogonal coordinate system in which the x axis, the y axis, and the z axis are orthogonal to each other.

  The blood vessel pattern creation unit 133 calculates the coordinate value (depth value) in the z-axis direction of the virtual three-dimensional space for each segment based on the blood vessel width (blood vessel thickness) in the biometric image, and represents the position of the segment. A three-dimensional blood vessel pattern is created in association with two-dimensional coordinate values. The blood vessel pattern creation unit 133 includes a blood vessel width extraction unit 133a, a depth value calculation unit 133b, and a pattern three-dimensionalization unit 133c.

  The blood vessel width extraction unit 133a extracts the width (thickness) of the partial blood vessel corresponding to the segment in the biometric image as the line width information of the segment.

  The depth value calculation unit 133b calculates the depth value of the segment in the virtual three-dimensional space expressing the blood vessel pattern, based on the extracted line width information of the segment.

  The pattern three-dimensionalization unit 133c associates the two-dimensional coordinate values representing the position of the segment in the xy plane with the depth value calculated by the depth value calculation unit 133b, and converts the two-dimensional blood vessel pattern into the three-dimensional blood vessel pattern. Rebuild (convert) to.

  The stability determination unit 134 determines the extraction stability of each segment in the blood vessel pattern. The extraction stability is a value indicating the degree of stable extraction of the segment when the blood vessel image is extracted from the biological image as described above. The stability determination unit 134 includes a luminance cross section extraction unit 134a and a stability calculation unit 134b.

  The luminance cross section extracting unit 134a extracts a luminance cross section in the blood vessel width direction in the partial blood vessel in the biometric image corresponding to the segment. The luminance cross section is information representing a change in luminance in the blood vessel width direction.

  The stability calculation unit 134b calculates the extraction stability of the segment based on the change pattern of the brightness in the brightness cross section of the segment.

  The data output unit 135 outputs the biometric data including the three-dimensional blood vessel pattern and the extraction stability associated with each segment in the blood vessel pattern, based on the selection result of the processing selection unit 110. When the process selecting unit 110 selects the process of registering the template, the data output unit 135 performs a process of registering the created biometric data in the template database 150 as a process of outputting the created biometric data. On the other hand, when the process selecting unit 110 selects the process of registering the template, the data output unit 135 outputs the created biometric data to the authentication unit 140 as the input biometric data.

  FIG. 3 is a diagram showing a functional configuration of an authentication unit in the biometric authentication device according to the first embodiment.

  As shown in FIG. 3, the authentication unit 140 in the biometric authentication device 1 of the present embodiment includes a simple alignment unit 141, a matching reliability calculation unit 142, a matching unit 143, and a determination unit 144.

  The simple alignment unit 141 performs simple alignment (rough alignment) between the blood vessel pattern of the input biometric data acquired from the data creation unit 130 and the blood vessel pattern of the template registered in the template database 150. The simple alignment unit 141 roughly aligns the blood vessel patterns in the virtual three-dimensional space based on, for example, the segment having high extraction stability in the input biometric data and the segment having high extraction stability in the template. .

  The matching reliability calculation unit 142 calculates the matching reliability of the segment pair of the segment of the input biometric data and the segment of the template based on the positional relationship of the segments in each blood vessel pattern after the rough alignment. . The matching reliability calculation unit 142 calculates the matching reliability of the segment pair, with the segment of the input biometric data and the segment of the template having the distance in the virtual three-dimensional space equal to or less than the threshold as the segment pair. The matching reliability calculation unit 142 calculates, as the matching reliability of the segment pair, a product of the extraction stability of the segment of the input biometric data and the extraction stability of the segment of the template.

  The matching unit 143 calculates a matching score between the blood vessel pattern in the input biometric data and the blood vessel pattern in the template based on the matching reliability of each segment pair calculated by the matching reliability calculation unit 142.

  The determination unit 144 determines whether or not to authenticate the authentication applicant based on the matching score calculated by the matching unit 143, and outputs the determination result to the display device 4 or the authentication required device 5.

FIG. 4 is a diagram showing a configuration example of the template database.
As shown in FIG. 4, the template database 150 includes a registrant ID, a segment ID, coordinates of a start point, coordinates of an end point, and extraction stability.

  The registrant ID is information for identifying a person (authentication target person) authenticated by the biometric authentication device 1, and, for example, a positive integer starting from 1 is assigned in the order of registration.

  The segment ID is information for identifying a segment (partial blood vessel) in a blood vessel pattern for one registrant ID (one recognition target person), and is given, for example, a positive integer starting from 1.

  The coordinates of the starting point are coordinate values representing the position of the starting point (one end) of the segment in the virtual three-dimensional space. The first two coordinate values (XSn, YSn) of the coordinates (XSn, YSn, ZSn) of the start point of the segment whose segment ID is n are values representing the position of the segment in the biometric image. Further, the last coordinate value ZSn of the coordinates (XSn, YSn, ZSn) of the start point of the segment whose segment ID is n represents the depth value in the virtual three-dimensional space calculated based on the line width information of the segment. .

  The coordinates of the end point are coordinate values indicating the position of the end point (other end) of the segment in the virtual three-dimensional space.

  As described above, the extraction stability is a value indicating the degree of stable extraction of a segment when extracting a blood vessel image from a living body image.

  When the biometric authentication device 1 according to the present embodiment starts the operation, it performs the process shown in FIG. FIG. 5 is a flowchart illustrating the content of processing performed by the biometric authentication device.

  The biometric authentication device 1 first waits for the input of information for selecting the processing content from the input device 2 (step S1). When the user (authentication target person and authentication applicant) performs an operation of selecting the processing content, the input device 2 transmits information corresponding to the operation to the biometric authentication device 1. In the present embodiment, the processes selectable by the user are two types of processes, that is, a template registration process and an authentication process.

  When the information for selecting the processing content is input from the input device 2, the biometric authentication device 1 determines, based on the input information, whether to perform the process of registering the template or the authentication process (step). S2). The process selection unit 110 performs the process of step S2. The process selection unit 110 outputs to the data creation unit 130 which of the template registration process and the authentication process is to be performed.

  Next, the biometric authentication device 1 performs a data creation process (step S3) in the data creation unit 130. In step S3, the data creation unit 130 first causes the image acquisition unit 120 to acquire a biometric image of the user. The image acquisition unit 120 causes, for example, the imaging device 3 to capture a biometric image of the user, and acquires the data of the biometric image output by the imaging device 3. The image acquisition unit 120 transmits the acquired biometric image data (hereinafter, simply referred to as “biometric image”) to the data creation unit 130. When the data creation unit 130 acquires the biometric image of the user via the image acquisition unit 120, the data creation unit 130 creates biometric data including a three-dimensional blood vessel pattern from the biometric image and the extraction stability of each segment in the blood vessel pattern. To do.

  When the data creation process (step S3) is completed, the data creation unit 130 next determines whether the process determined in step S2 is a template registration process or an authentication process (step S4). The determination in step S4 is performed by the data output unit 135 of the data creation unit 130.

  When the determined process is a process of registering a template (step S4; registration), the data output unit 135 registers the created biometric data in the template database 150 as a template (registered data) (step S5). On the other hand, when the determined process is the authentication process (step S4; authentication), the data output unit 135 transmits the generated biometric data to the authentication unit 140 as the input biometric data. In this case (step S4; authentication), the biometric authentication device 1 then performs authentication processing (step S6) in the authentication unit 140.

  In step S6, the authentication unit 140 determines whether to authenticate the user based on the input biometric data received from the data creation unit 130 and the template registered in the template database 150, and the determination result (authentication Process to output the result).

  After finishing the process of step S5 or S6, the biometric authentication device 1 determines whether to continue the process (step S7). When continuing the processing (step S7; Yes), the biometric authentication device 1 returns to the state of waiting for input of information for selecting the processing content from the input device 2 (step S1). On the other hand, when the process is ended for some reason (step S7; No), the biometric authentication device 1 ends the process. Further, even when a command to end the process is input from the input device 2 or the like to the biometric authentication device 1 while waiting for the input of information for selecting the processing content from the input device 2 (step S1), The authentication device 1 ends the process shown in FIG.

  The data creation process (step S3) in the flowchart of FIG. 5 is performed by the data creation unit 130 as described above. The data creation unit 130 performs the processing illustrated in FIGS. 6 to 8 as the data creation processing.

  FIG. 6 is a flowchart illustrating the contents of the data creation process. FIG. 7 is a flowchart illustrating the content of the three-dimensional pattern creation process. FIG. 8 is a flowchart illustrating the contents of the extraction stability determination process.

  When the data creation process (step S3) is started, the data creation unit 130 first acquires a biometric image of the user as shown in FIG. 6 (step S301). The process of step S301 is performed by the blood vessel image extraction unit 131 of the data creation unit 130 and the image acquisition unit 120. When the blood vessel image extraction unit 131 receives the information about the process performed from the process selection unit 110, the blood vessel image extraction unit 131 causes the image acquisition unit 120 to acquire the biometric image of the user. The image acquisition unit 120 causes the imaging device 3 to capture a biometric image of the user and acquires the biometric image output by the imaging device 3. The image acquisition unit 120 also passes the acquired biometric image to the blood vessel image extraction unit 131.

  Next, the data creation unit 130 causes the blood vessel image extraction unit 131 to extract a blood vessel image from the acquired biometric image (step S302). The blood vessel image extraction unit 131 extracts a blood vessel image from the biometric image according to any known extraction method, for example.

  Next, the data creation unit 130 divides the extracted blood vessel image into a plurality of segments (step S303). The blood vessel dividing unit 132 performs step S303. The blood vessel dividing unit 132 first divides the blood vessel at a point where the blood vessel is branched or a point where the blood vessel is crossed, for example, according to any known division method. Then, the blood vessel dividing unit 132, for example, for each divided blood vessel, a section in which the blood vessel width (the thickness of the blood vessel) is substantially constant, and the blood vessel width is substantially monotonic from one end to the other end in the running direction of the blood vessel. It is further divided such that the decreasing section becomes one segment.

  Next, the data creation unit 130 performs a three-dimensional pattern creation process (step S304) and an extraction stability determination process (step S305).

  The three-dimensional pattern creation process of step S304 is performed by the blood vessel pattern creation unit 133. The blood vessel pattern creation unit 133 creates a three-dimensional blood vessel pattern from the image of the two-dimensional blood vessel extracted from the biometric image. The blood vessel pattern creation unit 133 calculates, for each segment, a depth value in a virtual three-dimensional space expressing the blood vessel pattern based on the width of the blood vessel corresponding to the segment, and the calculated depth value and the segment in the biological image. The two coordinate values representing the position are associated with each other.

  The stability determination unit 134 performs the extraction stability determination processing in step S305. The stability determination unit 134 calculates, for each segment, a change in brightness in the width direction (luminance cross section) in the blood vessel corresponding to the segment, and calculates (determines) extraction stability of the segment based on the brightness cross section. When calculating the extraction stability of the segment based on the luminance cross section, the stability determination unit 134 may calculate the extraction stability by using a predetermined mathematical expression, or a pattern of the luminance cross section prepared in advance and the extraction stability. The extraction stability may be calculated (determined) based on a table representing the relationship with

  In FIG. 6, the processes of steps S304 and S305 are performed in parallel, but the process is not limited to this, and the process of step S305 may be performed after the process of step S304. Further, in the processing of steps S304 and S305, the processing of step S304 may be performed after the processing of step S305.

  When the processes of steps S304 and S305 are finished, the data creation unit 130 then holds the segments in the three-dimensional blood vessel pattern and the extraction stability in association with each other (step S306), and ends the data generation process (step S306). return).

  The three-dimensional pattern creation process of step S304 in the data creation process (steps S301 to S306) of FIG. 6 is performed by the blood vessel pattern creation unit 133 as described above. The blood vessel pattern creation unit 133 performs the processing shown in FIG. 7 as the three-dimensional pattern creation processing.

  When starting the process of step S304, the blood vessel pattern creation unit 133 first selects one segment (step S304a). In step S304a, the blood vessel pattern creation unit 133 selects a segment based on a predetermined selection rule. The selection rule is, for example, that a segment with the smallest segment ID is selected from the segments not selected in step S304a.

  Next, the blood vessel pattern creation unit 133 extracts the line width information of the selected segment (step S304b). The process of step S304b is performed by the blood vessel width extraction unit 133a. The blood vessel width extraction unit 133a extracts the width (thickness) of the blood vessel in the section corresponding to the selected segment from the image of the blood vessel in the biological image.

  Next, the blood vessel pattern creation unit 133 calculates the depth value of the segment based on the extracted line width information (step S304c). The depth value calculation unit 133b performs the process of step S304c. The depth value calculation unit 133b represents, for example, the correspondence relationship between the width of the blood vessel and the coordinate value in the z-axis direction in the virtual three-dimensional space, which is prepared in advance based on the statistical data of the blood vessel width in the palm of a person. The depth value of the segment is calculated based on the table.

  Next, the blood vessel pattern creation unit 133 adds the depth value (dimension in the depth direction) to the two-dimensional coordinates indicating the position of the segment in the biometric image, and three-dimensionalizes the position information of the segment (step S304d). Step S304d is performed by the pattern three-dimensionalization unit 133c.

  Next, the blood vessel pattern creation unit 133 determines whether or not all the segments have been three-dimensionalized (step S304e). When there is a segment that has not been three-dimensionalized (step S304e; No), the blood vessel pattern creation unit 133 repeats the processing from step S304a. Then, when the processing for three-dimensionalization is performed on all the segments (step S304e; Yes), the blood vessel pattern creation unit 133 transmits the three-dimensional information of the segments to the data output unit 135, and the three-dimensional pattern creation is performed. The process ends (return).

  On the other hand, the extraction stability processing of step S305 in the data creation processing (steps S301 to S306) of FIG. 5 is performed by the stability determination unit 134 as described above. The stability determination unit 134 performs the process shown in FIG. 8 as the extraction stability determination process.

  When the stability determination unit 134 starts the process of step S305, it first selects one segment (step S305a). In step S305a, the stability determination unit 134 selects a segment based on a predetermined selection rule. The selection rule is, for example, that a segment with the smallest segment ID is selected from the segments not selected in step S305a.

  Next, the stability determination unit 134 extracts the luminance cross section of the selected segment (step S305b). The processing in step S305b is performed by the luminance cross section extraction unit 134a. The luminance cross-section extraction unit 134a extracts the distribution of luminance values in the width direction in the partial blood vessel corresponding to the selected segment in the blood vessel image in the biological image.

  Next, the stability determination unit 134 calculates the extraction stability of the segment based on the extracted luminance cross section (step S305c). The stability calculation unit 134b performs the process of step S305c. The stability calculation unit 134b, for example, based on a table, which is prepared in advance and shows the correspondence relationship between the cross-sectional luminance distribution pattern of the blood vessel in the palm of a human and the extraction stability of the blood vessel, the extraction stability of the segment. Calculate and determine the degree.

  Next, the stability determination unit 134 determines whether or not the process of determining the extraction stability has been performed for all the segments (step S305d). When there is a segment for which the extraction stability has not been determined (step S305d; No), the stability determination unit 134 repeats the processing from step S305a. Then, when the process of determining the extraction stability is performed for all the segments (step S305d; Yes), the stability determination unit 134 transmits the extraction stability of each segment to the data output unit 135 and performs extraction. The stability determination process ends (return).

  FIG. 9 is a diagram showing an example of a blood vessel image and a two-dimensional blood vessel pattern extracted from a biometric image. FIG. 10 is an enlarged view of the areas AR1 and AR2 in FIG. FIG. 11 is a diagram showing an example of a three-dimensionalized blood vessel pattern. FIG. 12 is a diagram showing a part of a blood vessel in a biometric image. FIG. 13 is a diagram showing a three-dimensional blood vessel pattern corresponding to the blood vessels in FIG.

  As described above, the biometric authentication device 1 according to the present embodiment acquires a biometric image showing a blood vessel in the palm of a human and performs template registration processing or authentication processing. At this time, the biometric authentication device 1 first extracts an image of the blood vessel 602 from the biometric image 6 showing the palm 601 of the user, as shown in (a) of FIG. 9 and (a) of FIG. 10. In the image of the blood vessel 602 extracted from the biometric image 6, the width (thickness) of each portion of the blood vessel substantially matches the width of the blood vessel running in the palm 601 (in the living body) of the user.

  However, in the conventional biometric authentication device, when creating a blood vessel pattern based on the extracted image of the blood vessel 602, the blood vessel in the image of the blood vessel is divided into a plurality of partial blood vessels (segments) as described above, A two-dimensional coordinate value representing the position of is calculated. When the information held by the biometric data is only the two-dimensional coordinate value indicating the position of the segment, the biometric data is, for example, as shown in (b) of FIG. 9 and (b) of FIG. All blood vessels 603 (segments) shown in the image 6 are represented by the same line width. Therefore, in the conventional biometric authentication device, there is a possibility that an error may occur in the association of the segments due to the fact that the positions where the positional relationship (the blood vessel running pattern) of the plurality of segments are similar are close to each other.

  For example, in the image of the blood vessel extracted from the biometric image 6 as in the area AR1 shown in FIG. 10A, the traveling pattern of the partial blood vessel 602a in the area B11 and the traveling pattern of the partial blood vessel 602b in the area B12 are shown. , The shape is similar, but the width (thickness) of the blood vessel is different. Therefore, in the blood vessel image extracted from the biometric image, the difference between the blood vessel patterns of the two regions B11 and B12 is clear. However, in a blood vessel pattern having a constant line width like the area AR2 shown in FIG. 10B, there is no difference between the running pattern of the partial blood vessel 603a in the area B21 and the running pattern of the partial blood vessel 603b in the area B22. Be clear. Therefore, in the authentication process, a partial blood vessel that is actually associated with the region B21 may be mistakenly associated with the region B22, resulting in an error in the authentication result.

  On the other hand, in the biometric authentication device 1 of the present embodiment, the depth value of the segment is calculated based on the width of the blood vessel in the image of the blood vessel, and as shown in FIG. 11, the blood vessel pattern (traveling pattern of the blood vessel 604) is cubic. Make it original. In FIG. 11, the xy plane of the virtual three-dimensional space 7 is a plane that represents the position of the segment in the biometric image 6. Further, in FIG. 11, the value in the z-axis direction represents the depth value according to the width (thickness) of the partial blood vessel corresponding to the segment.

  Many of the blood vessels 602 in the biometric image 6 change in width (thickness) in the middle as shown in FIG. 12, for example. When the blood vessel 602 whose width changes in this way is divided into segments, the blood vessel 602 includes, for example, a first segment 602c whose width is approximately R1 and a second segment 602d whose width changes from approximately R1 to approximately R2. , And a third segment 602e having a width of approximately R2.

  When the two-dimensional coordinates of one end of the first segment 602c in the biometric image 6 of FIG. 12 are Q1 and the two-dimensional coordinates of the other end are Q2, the first segment 602c is shown in FIG. 13 in the virtual three-dimensional space. It is represented by the first segment 604c shown. That is, it is represented by a straight line connecting the start point Q1 and the end point Q2 on the xy plane, and is represented by a straight line of z = Z1 corresponding to the blood vessel width R1 on the zx plane.

  Similarly, when the two-dimensional coordinate of one end of the third segment 602e in the biometric image 6 of FIG. 12 is Q3 and the two-dimensional coordinate of the other end is Q4, the third segment 602e is in the virtual three-dimensional space, It is represented by the third segment 604e shown in FIG. That is, it is represented by a straight line connecting the start point Q3 and the end point Q4 on the xy plane, and is represented by a straight line of z = Z2 corresponding to the blood vessel width R2 on the zx plane.

  Further, the second segment 602d in the biometric image 6 of FIG. 12 is represented by the second segment 604d shown in FIG. 13 in the virtual three-dimensional space. That is, it is represented by a straight line connecting the start point Q2 and the end point Q3 on the xy plane, and is represented by a straight line changing from z = Z1 to z = Z2 from the start point Q2 to the end point Q3 on the zx plane.

  As described above, by reconstructing the two-dimensional blood vessel pattern shown in FIG. 9B or the like into a three-dimensional blood vessel pattern based on the line width information (blood vessel width) of the segment, a plurality of continuous segments can be obtained. It is possible to reduce the locations where the positional relationship (the running pattern of blood vessels) is similar. Therefore, in the three-dimensionalized blood vessel pattern, for example, the difference between the segment 603a of the region B21 and the segment 603b of the region B22 in FIG. 10B becomes clear. Therefore, in the authentication process, it is possible to prevent the partial blood vessel that is actually associated with the partial blood vessel 603a of the region B21 from being mistakenly associated with the segment 603b of the region B22, and reduce the error in the authentication result. .

  As described above, in the biometric authentication device 1 according to the present embodiment, when the blood vessel pattern is created by extracting the image of the blood vessel 602 from the biometric image 6, the two-dimensional information indicating the position in the biometric image 6 is separated. A three-dimensional blood vessel pattern is created by adding the blood vessel feature amount. Therefore, in the biometrics authentication system 1 according to the present embodiment, it is possible to clarify the difference between similar parts only in the positional relationship in the biometrics image of a plurality of segments, and reduce the error in the authentication result. It will be possible. Therefore, according to the present embodiment, it is possible to improve the authentication accuracy of biometric authentication.

  Further, in the biometric authentication device 1 according to the present embodiment, it is possible to further improve the authentication accuracy by associating the extraction stability with each segment in the blood vessel pattern.

FIG. 14 is a diagram illustrating an example of a method of determining the extraction stability.
FIG. 14A shows four examples of the relationship between how the blood vessel appears in the biometric image 6 and the luminance cross section in the width direction of the blood vessel.

  The image (pattern 1) of the blood vessel 602f at the left end in (a) of FIG. 14 clearly shows the contour which is the end in the width direction (x direction). The brightness value in the width direction of the blood vessel 602f changes rapidly at the end of the blood vessel 602f, and the brightness of the middle portion becomes substantially constant.

  In the image (pattern 2) of the blood vessel 602g that is the second from the left, the contour portion that is the end in the width direction (x direction) is slightly blurred. The brightness value of the blood vessel 602g in the width direction changes more gently at the end portion of the blood vessel 602g than the blood vessel 602f, and the brightness of the middle portion is substantially constant.

  On the other hand, the image of the third blood vessel 602h from the left (Pattern 3) is not entirely clear, not only in the contour portion which is the end in the width direction (x direction). The brightness value in the width direction of the blood vessel 602h changes more gently at the end portion of the blood vessel 602h than the blood vessel 602g, and the section in which the brightness value is the minimum is short.

  The image of the blood vessel 602i at the right end (pattern 4) is clearer than the image of the blood vessel 602h (pattern 3), but is less clear than the images of the blood vessels 602f and 602g. Further, in the image of the blood vessel 602i, a streak-like bright portion extending in the traveling direction of the blood vessel can be seen in the middle portion in the width direction. Such a brightness value in the width direction of the blood vessel 602i changes more gently at the end of the blood vessel 602i than the blood vessel 602g, and the section in which the brightness value is the minimum is short.

  As in the four blood vessels 602f to 602i shown in FIG. 14A, even if the blood vessel width and the minimum luminance value in the cross section in the width direction are the same, the luminance change in the width direction of the extracted blood vessel. The (luminance cross section) differs depending on the blood vessel (segment).

  Further, the pattern of the luminance cross section of the blood vessel has a correlation with, for example, the degree of stable extraction of the blood vessel (extraction stability) when imaged under a plurality of imaging conditions. For example, like the blood vessel 602f of the pattern 1 in FIG. 14A, a blood vessel that changes rapidly at the end of the blood vessel 602f and the brightness of the middle portion is substantially constant is stable even when the imaging condition changes. To be extracted. Similarly, the blood vessel 602g of pattern 2 is also stably extracted even if the imaging conditions change. On the other hand, a blood vessel having an unclear edge in the width direction such as the blood vessel 602h of the pattern 3 or the blood vessel 602i of the pattern 4 has many imaging conditions in which the blood vessel 602f is not extracted as compared with the blood vessel having the sharp edge. . Therefore, compared to the case where the blood vessel pattern is only three-dimensionalized using the width (thickness) of the blood vessel, the method using the luminance cross section of the blood vessel captures the features of the blood vessel in the biological image more clearly and uses it. It is possible to clarify the difference in blood vessel characteristics between persons.

  Therefore, in the present embodiment, as described above, biometric data in which the extraction stability based on the luminance cross section of the segment is associated with each segment (partial blood vessel) in the three-dimensional blood vessel pattern is created, and the template and the input biometric data are created. Used as. As a result, in the authentication processing according to the present embodiment, the difference in blood vessel characteristics between users is clarified, and the error in the authentication result can be further reduced.

  When determining the extraction stability, for example, after extracting the blood vessel image from the biological image, the extraction stability S (p) of the blood vessel at an arbitrary point p on the extracted blood vessel is calculated and determined. To do. Here, the point p is the center point of the blood vessel on the blood vessel path.

  The extraction stability S (p) is based on the amount of change in luminance (that is, the luminance cross section) in a cross-sectional direction (that is, the width direction of the blood vessel) that crosses the blood vessel path vertically at an arbitrary point p on the blood vessel path. To calculate. To obtain the stability from the luminance cross section, a function F such as the following formula (1) is defined and calculated from the distance from the center of the blood vessel and the amount of change in luminance.

S (p) = F (i- _w (p), ..., _iw (p)) (1)

In equation (1), i _u (p) is a luminance value at a distance u in the cross-sectional direction across the blood vessel at an arbitrary point p on the blood vessel path. Further, w in the equation (1) is a distance (that is, a half value of the width of the blood vessel) from the center of the blood vessel to the position where the change in brightness is 0 in the cross-sectional direction crossing the blood vessel.

  The function F is analytically defined by a method such as Point Spread Function (PSF) based on the diffusion characteristics of the skin. Further, the function F is, for example, based on experimental data obtained by photographing blood vessel images of many subjects under various conditions, and obtains a correlation between the probability that a segment is extracted and the feature of the luminance cross section of the segment, and obtains this data. A method of slicing and holding it may be used.

  Further, as an experimental method, for example, a table defining the shape of the luminance cross section of the blood vessel image, which is known to be stably acquired, is prepared in advance, and the stability value is calculated based on the table. You may. For example, as shown in FIG. 14B, a table 160 in which the pattern of the luminance cross section and the extraction stability are associated with each other is created and stored in a storage unit (not shown in FIG. 1) in the biometric authentication device 1. It may be stored and the extraction stability of each segment may be calculated based on the table 160. In this case, for example, how close the luminance cross section of the blood vessel extracted from the biometric image is to the pattern of the typical luminance cross section of the blood vessel that is stably extracted is compared, and the stability is based on the similarity of the luminance cross section. The value of may be calculated.

  When creating the table 160 as shown in (b) of FIG. 14, the extraction stability is set to 1 when stable extraction is always possible. In addition, the value of the extraction stability when stable extraction is not possible is set in the range of 0 <S <1 based on the relationship between the imaging conditions that can be extracted and the imaging conditions that cannot be extracted. For example, the extraction stability of a segment that is stably extracted under 7 to 9 imaging conditions out of 10 imaging conditions is a value close to 1 out of 0 <S <1. On the contrary, the extraction stability of the segment that is stably extracted only in 4 to 5 of the 10 imaging conditions is set to a value close to 0 in 0 <S <1.

  In addition, due to the influence of external light at the time of imaging and exposure, the biological image 6 appears to be bright as a whole, and the image of the blood vessel is thinly extracted in the entire biological image 6, or conversely, The blood vessel image may be thickly extracted in the entire biometric image 6. In such a case, the brightness value of each pixel in the biometric image 6 is corrected based on the brightness of the entire biometric image 6, and the extraction stability is calculated (determined) based on the shape of the corrected brightness cross section. Good.

  The blood vessel width and brightness itself obtained from the blood vessel image are relatively unstable values, but the shape of the brightness cross section, which is the local brightness distribution around the blood vessel, can be acquired relatively stably. . For example, the blood vessel width and the brightness reflected in the biometric image 6 change depending on various conditions such as the brightness of the illumination and the distance between the sensor and the living body part, but the amount of change in the brightness in the direction perpendicular to the blood vessel is different in the brightness itself. But the same changes. Therefore, for example, by normalizing so that either the minimum value or the maximum value of the luminance change amount of the luminance cross section becomes equal, and utilizing the characteristic shape of the luminance cross section for the comparison of blood vessels, the depth of the blood vessel can be determined by the luminance alone. As compared with the case of estimating the height, it is possible to more stably contribute to the identification of blood vessels.

  Next, the authentication process (step S6) in the flowchart of FIG. 5 will be described. The authentication processing is performed by the authentication unit 140 as described above. The authentication unit 140 performs the processing shown in FIGS. 15, 16A and 16B, and FIGS. 17A to 17C as the authentication processing.

  FIG. 15 is a flowchart illustrating the contents of the authentication process. FIG. 16A is a flowchart (part 1) explaining the content of the matching process of the input biometric data and the template. FIG. 16B is a flowchart (part 2) explaining the content of the matching process between the input biometric data and the template. FIG. 17A is a flowchart (part 1) explaining the content of the rough alignment processing. FIG. 17B is a flowchart (part 2) explaining the content of the rough alignment processing. FIG. 17C is a flowchart (part 3) explaining the content of the rough alignment processing.

  When the authentication process (step S6) is started, the authentication unit 140 first performs a matching process between the input biometric data and the template as shown in FIG. 15 (step S61). The process of step S61 is performed by the simple alignment unit 141 of the authentication unit 140, the matching reliability calculation unit 142, and the matching unit 143. In step S61, the simple alignment unit 141 roughly aligns the input biometric data and the template on the basis of the set (segment pair) of the segments having high extraction stability in the input biometric data and the template. In step S61, the matching reliability calculation unit 142 calculates the matching reliability based on the segment extraction stability in the input biometric data and the segment extraction stability in the template after performing rough alignment. . Further, in step S61, the matching unit 143 calculates a matching score representing the degree of matching between the input biometric data and the template for each segment based on the calculated matching reliability, and adds the final matching score to the determination unit 144. Pass to.

  Next, the authentication unit 140 determines whether or not the collation is successful in step S61 (step S62). The determination unit 144 performs the determination in step S62. The determination unit 144 determines whether or not the final matching score received from the matching unit 143 is equal to or higher than the threshold value, and when the final matching score is equal to or higher than the threshold value, determines that the matching is successful (that is, the authentication is successful). To do.

  When the collation is successful (step S62; Yes), the determination unit 144 notifies that the authentication is successful and causes the authentication-requiring device 5 to execute a predetermined process (step S63). In step S63, for example, the determination unit 144 causes the display device 4 to display that the authentication has succeeded, and causes the authentication-requiring device 5 to execute the process when the authentication succeeds. When the authentication-requiring device 5 is the automatic locking device, the determination unit 144 transmits an unlock command to the authentication-requiring device 5 in step S63.

  On the other hand, when the collation fails (step S62; No), the determination unit 144 notifies that the authentication has failed (step S64). In step S63, the determination unit 144 causes the display device 4 to display that authentication has failed, for example.

  When the determination unit 144 finishes the process of step S63 or S64, the authentication unit 140 finishes the authentication process (return).

  The authentication unit 140 performs the processing shown in FIGS. 16A and 16B as the matching processing (step S61) in the flowchart of FIG.

  In the matching process of step S61, the authentication unit 140 first selects and reads one of the templates registered in the template database 150, as shown in FIG. 16A (step S6101). The process of step S6101 is performed by the simple alignment unit 141. The simple alignment unit 141 selects one template from the template database 150 according to a predetermined selection rule. The selection rule is, for example, that a template with a minimum registrant ID is selected from templates that have not been subjected to collation processing.

  Next, the authentication unit 140 causes the simple alignment unit 141 to roughly align the input biometric data with the template based on the extraction stability (step S6102). In step S6102, the simple alignment unit 141 extracts the same number of segments with high extraction stability from the input biometric data and the template, and roughly aligns the segments extracted (simple alignment).

  Next, the authentication unit 140 initializes a final matching score indicating a matching result (matching degree) between the input biometric data and the selected template (step S6103).

  Next, as illustrated in FIG. 16B, the authentication unit 140 selects one of the segment pairs based on the rough alignment result (step S6104). The process of step S6104 is performed by the matching reliability calculation unit 142. The matching reliability calculation unit 142 selects one segment pair out of the segment pairs set based on the distance between the segment in the input biometric data and the segment in the template after rough alignment. .

  The matching reliability calculation unit 142 selects one segment pair according to a predetermined selection rule. It is assumed that the selection rule is, for example, to select the segment pair having the minimum segment ID of the segment on the input biometric data side among the segment pairs not selected in step S6104.

  Next, the authentication unit 140 causes the matching reliability calculation unit 142 to calculate the matching reliability of the selected segment pair (step S6105). In step S6105, the matching reliability calculation unit 142 calculates, as the matching reliability, the product of the extraction stability of the segment on the input biometric data side of the segment pair and the extraction stability of the segment on the template side.

  Next, the authentication unit 140 collates the segment pairs in the three-dimensional space and calculates a collation score that is corrected based on the collation reliability (step S6106). The collation unit 143 performs the process of step S6106. The closer the three-dimensional distance of the segment pair is, the more likely the segment pair is the same segment (partial blood vessel). For this reason, the matching unit 143 sets the additional point of the matching score for the segment pair to be larger as the segment pair is closer. Further, the matching unit 143 corrects the matching score by multiplying the matching score of the segment pair by the matching reliability calculated from the extraction stability of the segment pair.

  Next, the authentication unit 140 causes the matching unit 143 to add the matching score to the final matching score (step S6107).

  Next, the matching unit 143 determines whether or not the processing for calculating the matching score has been performed for all the segment pairs (step S6108). When there is a segment pair for which the process for calculating the matching score is not performed (step S6108; No), the matching unit 143 cooperates with the matching reliability calculation unit 142 and repeats the processes from step S6104. Then, when the process of calculating the matching score for all the segment pairs is performed (step S6108; Yes), the matching unit 143 next determines whether or not the final matching score is equal to or more than a threshold value (step S6109). .

  When the final matching score is equal to or higher than the threshold value (step S6109; Yes), the matching unit 143 notifies the determining unit 144 that the matching is successful (step S6110), and ends the matching process (return).

  On the other hand, when the final matching score is smaller than the threshold value (step S6109; No), the matching unit 143, for example, causes the simple alignment unit 141 to determine whether or not there is an unmatched template (step S6111). If there is a template that has not been matched (step S6111; Yes), the matching unit 143 cooperates with the simple alignment unit 141 and the matching reliability calculation unit 142 to repeat the processing from step S6101.

  On the other hand, when the collation is completed with all the templates (step S6111; No), the collating unit 143 notifies the determining unit 144 that the collation has failed (step S6112), and the collating process is ended (return). . That is, when the input biometric data is matched with all the templates, but there is no template whose final matching score is equal to or higher than the threshold value, the matching unit 143 determines that the matching has failed.

  As described above, in the authentication processing according to the present embodiment, when the matching data and the template are matched, the matching score is corrected by the matching reliability calculated based on the extraction stability of the segment pair. When the extraction stability of two segments as a segment pair is high, the result of collation using the segment pair by the two segments becomes reliable. Therefore, the weight that the matching score of the segment pair with high extraction stability contributes to the final matching score should be large. On the contrary, when the extraction stability of two segments as a segment pair is low, it can be said that the result of collation using the segment pair of the two segments has low reliability. Therefore, the weight that the matching score of the segment pair having low extraction stability contributes to the final matching score should be small.

  In the matching process (step S61) according to the present embodiment, the product of the extraction stability of segment pairs is used as the matching reliability, as described above. When the extraction stability S is 0 <S ≦ 1, the higher the extraction stability of the segment pair, the greater the matching reliability is 1 or a value close to 1. Therefore, in the matching process according to the present embodiment, the weight that the matching score of each segment pair contributes to the final matching score can be adjusted based on the degree of extraction stability of the segment pair.

  Further, in the matching process according to the present embodiment, in step S6102, rough alignment of the blood vessel pattern (input biometric data) newly created for matching with the blood vessel pattern (template) registered in the template database 150 is performed. I do.

  In the conventional blood vessel pattern matching process in the biometric authentication device, rough matching is performed on all the segments in the blood vessel pattern, and the alignment parameter is determined so that the degree of matching between the two blood vessel patterns is the highest. To be done. However, in the registration in the virtual three-dimensional space, it is necessary to consider the three-dimensional positional shift, rotation, enlargement / reduction, etc. of the position of the segment, and the processing amount is larger than that in the two-dimensional registration.

  Therefore, in the matching process according to the present embodiment, the extraction stability associated with each segment in the blood vessel pattern is used, and the processes shown in FIGS. 17A to 17C are performed to make the rough alignment process efficient. . The rough alignment processing of FIGS. 17A to 17C is performed by the simple alignment unit 141.

  When the rough alignment process (step S6102) is started, the simple alignment unit 141 first initializes the extraction stability reference value α as shown in FIG. 17A (step S6102a). In step S6102a, simple alignment section 141 sets reference value α to a value that is 0 <α <1 and is close to 1.

  Next, the simple alignment unit 141 extracts a segment whose extraction stability is higher than the reference value α from the input biometric data and the template (step S6102b).

  Next, the simple alignment unit 141 initializes the best alignment score BAS and the best alignment parameter BAP (step S6102c). In step S6102c, the simple alignment unit 141 sets the best alignment score BAP to 0, for example. Further, in step S6102c, simple alignment section 141 sets the values of various parameters including position deviation, rotation, and deformation in best alignment parameter BAP to 0, respectively.

  Next, the simple alignment unit 141 sets the first alignment parameter AP (step S6102d). The alignment parameter is a parameter that specifies a point to be matched when aligning the blood vessel pattern in the input biometric data with the blood vessel pattern in the template, a direction (rotation angle) of the blood vessel pattern, and the like. In step S6102d, the simple alignment unit 141 sets the initial alignment parameter AP according to a preset alignment parameter setting rule.

  Next, as shown in FIG. 17B, the simple alignment unit 141 performs coordinate conversion on the segment groups extracted from the input biometric data and the template using the alignment parameter AP, and then collates them to align the degree of coincidence. The score AS is calculated (step S6102e).

  Next, the simple alignment unit 141 determines whether the calculated alignment score AP is larger than the best alignment score BAS (step S6102f). When AS> BAS (step S6102f; Yes), the simple alignment unit 141 updates the best alignment score BAS to the calculated alignment score AS (step S6102g). Also, in step S6102g, simple alignment section 141 updates the best alignment parameter BAP to the currently set alignment parameter AP.

  After step S6102g, the simple alignment unit 141 determines whether all alignment parameters have been processed (step S6102h). If AS ≦ BAS (step S6102f; No), the simple alignment unit 141 skips the process of step S6102g and makes the determination of step S6102h.

  If there is an unprocessed alignment parameter (step S6102h; No), the simple alignment unit 141 changes the alignment parameter AP (step S6102i), and repeats the processing from step S6102e. Then, when processing is performed with all the alignment parameters AP (step S6102h; Yes), the simple alignment unit 141 then changes the reference value α as shown in FIG. 17C (step S6102j). In step S6102j, the simple alignment unit 141 changes the reference value α to a value smaller than the current value according to a predetermined change rule.

  Next, the simple alignment unit 141 determines whether the changed reference value α is larger than the alignment reference cutoff reference value (step S6102k). The cutoff reference value may be the minimum value of the extraction stability or a value larger than the minimum value of the extraction stability.

  When the reference value α is larger than the cutoff reference value (step S6102k; Yes), the simple alignment unit 141 repeats the processing from step S6102b onward in FIG. 17A. On the other hand, when the changed reference value α becomes equal to or less than the cutoff reference value (step S6102k; No), the simple alignment unit 141 then sets the best alignment parameter BAP as the alignment parameter between the input biometric data and the template. (Step S6102m). When the alignment parameter is held in step S6102m, the simple alignment unit 141 ends the rough alignment process (return).

  As described above, in the rough alignment processing according to the present embodiment, the same number of blood vessel segments with high extraction stability are read out from each of the two pieces of blood vessel patterns of the input biometric data and the selected template, and matching is performed in order. That is, in the matching process (step S61) according to the present embodiment, the simple alignment unit 141 first determines the rough position between the blood vessel pattern of the input biometric data and the blood vessel pattern of the template based on the segment group with high extraction stability. Make a match.

  After that, the simple alignment unit 141 reads out a segment group whose extraction stability is greater than the abortion reference value for alignment little by little from each blood vessel pattern, and performs more detailed alignment using the segment group to which they are added.

  Note that the cutoff reference value may be the minimum value of the extraction stability as described above, or may be a value larger than the minimum value of the extraction stability. Therefore, in the rough alignment processing, alignment may be finally performed using all the segment groups. Also, in the rough alignment process, the segment group with low extraction stability is not used for alignment.For example, the process ends when alignment is performed using segment groups with moderate extraction stability. May be. As described above, in the matching processing according to the present embodiment, it is possible to speed up the positioning processing by performing the positioning stepwise using the information of the extraction stability or by canceling the positioning processing in the middle. it can.

  In the rough alignment process, the alignment method in the three-dimensional space, that is, the setting of the alignment parameter AP is also optional. Therefore, in the rough alignment processing according to the present embodiment, for example, the alignment parameter AP (various parameters for positional deviation / rotation / deformation) is set so that more segments match or are very close to each other. May be searched.

  As described above, according to this embodiment, when the image of the blood vessel extracted from the biometric image is divided into a plurality of segments, the segment (partial blood vessel) is added to the information indicating the position of the segment in the biometric image (in the plane). ), The depth value according to the width is added to make the blood vessel pattern three-dimensional. Therefore, it becomes possible to give a difference according to the width of the blood vessel to a plurality of locations where the positional relationship of the plurality of segments is very similar in the plane. Therefore, according to the present embodiment, it is possible to reduce the error in the authentication result and improve the authentication accuracy as compared with the case where the verification (authentication) is performed based on the two-dimensional blood vessel pattern.

  Further, when the blood vessel pattern is three-dimensionalized as in the present embodiment, by associating the extraction stability with each segment in the blood vessel pattern, the matching result of the segment pair with low extraction stability at the time of matching is the final matching result. Can be reduced. Further, by utilizing the extraction stability, it is possible to increase the influence of the matching result of the segment pair having high extraction stability on the final matching result at the time of matching. Therefore, according to the present embodiment, it is possible to reduce the error in the collation result caused by collating the collation data (blood vessel pattern) in which the segment having low extraction stability is not extracted with the template. Therefore, according to the present embodiment, it is possible to further reduce the error in the authentication result and further improve the authentication accuracy.

  Moreover, according to the present embodiment, rough matching can be efficiently performed based on the segment pair having high extraction stability when matching the matching data and the template. Therefore, the processing load of the biometric authentication device 1 can be reduced.

  As described above, the flowcharts of FIGS. 5 to 8, FIG. 15, FIG. 16A and FIG. 16B, and FIGS. 17A to 17C are merely examples of the processing performed by the biometric authentication device 1 according to the present embodiment. Absent. That is, in the processing performed by the biometric authentication device 1 according to the present embodiment, as described above, a part of the processing illustrated in FIGS. 5 to 8, FIG. 15, FIG. 16A and FIG. 16B, and FIG. 17A to FIG. 17C is performed. The processing may be changed or omitted. For example, in this embodiment, as shown in FIGS. 17A to 17C, the alignment process is performed by gradually increasing the number of segment groups based on the extraction stability associated with each segment of the blood vessel pattern. Has been made more efficient. However, the position adjustment is not limited to this, and the position adjustment may be performed using all the segments from the beginning.

  Further, the template database 150 of FIG. 4 is merely an example, and the configuration of the template database 150 can be appropriately changed. For example, the template database 150 may have the configuration shown in FIG.

FIG. 18 is a diagram showing another configuration example of the template database.
When the image of the blood vessel extracted from the biometric image is divided into a plurality of segments, for example, the blood vessel is divided at the points where the blood vessels branch or intersect. In this case, for example, one segment (partial blood vessel) may be running while curving or bending. Therefore, the template database 150 registers the coordinates of the start point and the coordinates of the end point, as well as the coordinates of the bending point existing between the start point and the end point, as the three-dimensional information of each segment, as shown in FIG. 18, for example. It may be configured.

  Further, in the present embodiment, the blood vessel width is mentioned as the characteristic amount of the blood vessel used for three-dimensionalization of the blood vessel pattern. However, the feature amount of the blood vessel is not limited to this, and may be the brightness of the blood vessel.

  The biometric authentication device 1 illustrated in FIG. 1 is merely an example of the functional configuration of the biometric authentication device according to this embodiment. In the biometric authentication device according to the present embodiment, for example, the template database 150 may be stored not in the biometric authentication device 1 itself but in another server or the like that is communicably connected to the biometric authentication device 1.

[Second Embodiment]
In this embodiment, an example in which a blood vessel position variation rate different from the matching reliability described in the first embodiment is calculated based on the segment extraction stability, and the authentication processing is performed in consideration of the blood vessel position variation rate. Will be explained.

The overall functional configuration of the biometric authentication device 1 according to the present embodiment is as shown in FIG.
That is, the biometric authentication device 1 according to the present embodiment includes the process selection unit 110, the image acquisition unit 120, the data creation unit 130, the authentication unit 140, and the template database 150. Further, the data creation unit 130, as shown in FIGS. 1 and 2, includes a blood vessel image extraction unit 131, a blood vessel division unit 132, a blood vessel pattern creation unit 133, a stability determination unit 134, and a data output unit 135. ,including. Similar to the first embodiment, the data creation unit 130 in the biometric authentication device 1 of the present embodiment creates a three-dimensional blood vessel pattern created from a biometric image and the extraction stability associated with each segment in the blood vessel pattern. Create biometric data including. A three-dimensional blood vessel pattern is created by adding, for example, a depth value according to the line width information (thickness of blood vessel) of the segment to the two-dimensional coordinate value that represents the position of the segment in the biological image (in the plane). To do. In addition, the extraction stability is calculated (determined) based on the luminance value variation pattern in the luminance cross section by extracting the luminance cross section in the blood vessel width direction of the segment.

  Next, the configuration of the authentication unit 140 in the biometric authentication device 1 of this embodiment will be described with reference to FIG.

  FIG. 19 is a diagram showing the functional configuration of the authentication unit in the biometric authentication device according to the second embodiment.

  As shown in FIG. 19, the authentication unit 140 in the biometric authentication device 1 according to this embodiment includes a simple alignment unit 141, a blood vessel position variation rate calculation unit 145, a collation unit 143, and a determination unit 144.

  As described in the first embodiment, the simple alignment unit 141 extracts a segment having high extraction stability from the blood vessel pattern of the input biometric data and the blood vessel pattern of the template, and roughly aligns the blood vessel pattern.

  The blood vessel position variation rate calculation unit 145 calculates the blood vessel position variation rate of the segment in the three-dimensional blood vessel pattern based on the extraction stability of each segment in the segment pair of the input biometric data segment and the template segment. In other words, the blood vessel position variation rate of the segment is the rate of variation of the depth value of the segment due to the variation of the brightness of the blood vessel that occurs when the biological image is captured. The blood vessel position variation rate calculation unit 145 calculates the blood vessel position variation rate such that the segment pair having a higher segment extraction stability has a smaller blood vessel position variation rate.

  The matching unit 143 performs detailed matching between the blood vessel pattern of the input biometric data that has been roughly aligned with the blood vessel pattern of the template. The matching unit 143 in the biometric authentication device 1 according to the present embodiment calculates the matching score of the segment pair in consideration of the blood vessel position variation rate instead of the matching reliability in the first embodiment.

  A segment pair having a high extraction stability has a high matching reliability and a low blood vessel position variation rate. On the contrary, a segment pair of segments with low extraction stability has low matching reliability and a high blood vessel position variation rate. Therefore, for example, when calculating the matching score of the segment pair having a high blood vessel position variation rate, the matching unit 143 increases the tolerance of the positional deviation with respect to the depth value of the segment. This makes it possible to prevent the matching score of the segment pair of the correct combination from becoming small due to the fluctuation of the blood vessel position in the z-axis direction in the virtual three-dimensional space.

  Further, for example, when calculating the matching score of the segment pair having a low blood vessel position variation rate, the matching unit 143 reduces the tolerance of the positional deviation regarding the depth value of the segment. As a result, it is possible to reduce the influence of the matching score of a segment pair that is not a correct combination, although it has high extraction stability, on the matching result.

  The determination unit 144 determines whether or not to authenticate the user (authentication-requesting person) based on the verification result of the verification unit 143, and outputs the authentication result to the display device 4, the authentication-required device 5, or the like.

  The biometric authentication device 1 according to the present embodiment performs the processing of steps S1 to S7 described in the first embodiment (see FIG. 5). Further, even when a command to end the process is input from the input device 2 or the like to the biometric authentication device 1 while waiting for the input of information for selecting the processing content from the input device 2 (step S1), The authentication device 1 ends the process shown in FIG.

  Furthermore, as the data creation processing (step S3) in the flowchart shown in FIG. 5, the biometric authentication device 1 (data creation unit 130) of the present embodiment performs the processing shown in FIGS. 6 to 8, for example.

  In addition, the authentication unit 140 in the biometric authentication device 1 according to the present embodiment is based on the input biometric data and the template in the procedure shown in FIGS. 15, 16A and 16B, and 17A to 17C, for example. Authentication processing (step S6) is performed. However, in the authentication process according to the present embodiment, the authentication unit 140 calculates the blood vessel position variation rate instead of steps S6105 and S6106 shown in FIG. 16B, and calculates the matching score based on the blood vessel position variation rate. I do.

  FIG. 20 is a flowchart illustrating a part of the matching process of the input biometric data and the template according to the second embodiment. Note that in the flowchart of FIG. 20, steps that are the same processing as the processing described in the first embodiment are given the same step numbers (for example, S6104, etc.) as those of the flowchart of FIG. 16B.

  When starting the authentication process (step S6), the authentication unit 140 in the biometric authentication device 1 of the present embodiment performs the processes of steps S61 to S64 shown in FIG. Of these, in the matching process between the input biometric data and the template in step S61, the authentication unit 140 first performs the processes in steps S6101 to S6103 illustrated in FIG. 16A. After that, the authentication unit 140 selects one of the segment pairs based on the rough alignment result, as shown in FIG. 20 (step S6104).

  When the segment pair is selected, the authentication unit 140 then calculates the blood vessel position variation rate of the selected segment pair (step S6115). The process of step S6115 is performed by the blood vessel position variation rate calculation unit 145. The blood vessel position variation rate calculation unit 145 calculates, for example, the reciprocal of the product of the extraction stability of each segment in the selected segment pair as the blood vessel position variation rate. As shown in the first embodiment, when the extraction stability S is 0 <S ≦ 1 and the reciprocal of the product of the extraction stability is the blood vessel position variation rate, the blood vessel position of the segment pair with high extraction stability is obtained. The fluctuation rate is 1 or a value close to 1. On the other hand, the blood vessel position variation rate of the segment pair having low extraction stability becomes larger than 1, and the absolute value becomes larger as the extraction stability becomes lower.

  Next, the authentication unit 140 collates the segment pairs in consideration of the variation of the segment pairs in the depth direction in the three-dimensional space based on the blood vessel position variation rate, and calculates a collation score (step S6116). The processing of step S6116 is performed by the matching unit 143. For example, for a segment pair having a high blood vessel position variation rate, the matching unit 143 increases the tolerance for positional deviation between the segments in the segment pair in the depth direction and calculates a matching score. The matching unit 143 calculates the matching score based on the distance between the segments in the pair of segments in the xy plane, the distance in the z direction (difference in depth value), and the like. At this time, for example, when the distance between the segments in the pair of segments in the z direction is equal to or less than the threshold, the matching unit 143 adds a value corresponding to the distance to the matching score. However, when the threshold value for the distance between the segments in the z direction is a constant value, the distance between the segments in the z direction exceeds the threshold value for the segment pair having a high blood vessel position variation rate. For this reason, even if the segment pair is a set of correct segments, there is a possibility that it will not be reflected in the matching score because the positional deviation in the z direction is large. Therefore, when calculating the matching score for the segment pair having a high blood vessel position variation rate in step S6116, the matching unit 143 sets the threshold value of the distance in the z direction between the segments to be added to the matching score to the blood vessel position fluctuation, for example. Lengthen according to the rate value. As a result, in a segment pair with a high rate of blood vessel position variation, in other words, a segment pair with a large fluctuation in depth value in a three-dimensional blood vessel pattern, it is possible to suppress a decrease in the matching score due to a large positional deviation between the segments in the depth direction. It becomes possible to do.

  When the matching score in consideration of the blood vessel position variation rate is calculated, the matching unit 143 then adds the matching score to the final matching score (step S6107), and performs a process of calculating matching scores for all segment pairs. It is determined whether or not (step S6108).

  When there is a segment pair for which the process of calculating the matching score has not been performed (step S6108; No), the matching unit 143 cooperates with the blood vessel position variation rate calculating unit 145 to repeat the processes from step S6104. Then, when the process of calculating the matching score for all the segment pairs is performed (step S6108; Yes), the matching unit 143 next determines whether or not the final matching score is equal to or more than a threshold value (step S6109). ).

  When the final matching score is equal to or higher than the threshold value (step S6109; Yes), the matching unit 143 notifies the determining unit 144 that the matching has succeeded (step S6110), and ends the matching process between the matching data and the template (step S6110). return).

  On the other hand, when the final matching score is smaller than the threshold value (step S6109; No), the matching unit 143 next determines whether or not there is any unmatched template (step S6111). If there is a template that has not been collated (step S6111; Yes), the collation unit 143 cooperates with the simple alignment unit 141 and the blood vessel position variation rate calculation unit 145 to repeat the processing from step S6101. Then, when there is no template whose final matching score is equal to or higher than the threshold value even though all the templates have been matched (step S6111; No), the matching unit 143 notifies the determination unit 144 that the matching has failed (step S6111). (S6112), and the matching process ends.

  As described above, in the present embodiment, based on the blood vessel position variation rate according to the extraction stability of each segment in the three-dimensional blood vessel pattern, the variation of the position of the segment in the depth direction in the three-dimensional space is taken into consideration in the input living body. Check the data against the template. At this time, among the segment pairs of the segment of the input biometric data and the segment of the template, for the segment pair having a large blood vessel position variation rate, the matching score is calculated by increasing the tolerance for the displacement in the depth direction in the three-dimensional space. To do. Therefore, it is possible to suppress a decrease in the matching score of the segment pair that is a correct combination, although the extraction stability is low.

  In addition, in the present embodiment, among the segment pairs of the segment of the input biometric data and the segment of the template, for the segment pair having a small blood vessel position variation rate, the tolerance for positional deviation in the depth direction in the three-dimensional space is reduced. To calculate the matching score. Therefore, it is possible to prevent the matching score of a segment pair having a high extraction stability but not a correct combination from increasing.

  Therefore, according to the present embodiment, as in the first embodiment, it is possible to suppress the output of an incorrect matching result (authentication result), and the authentication accuracy is improved.

  The flowchart shown in FIG. 20 is merely an example of the matching process of the input biometric data and the template according to this embodiment. That is, the collation process performed by the biometric authentication device 1 according to the present embodiment may be a process in which a part of the process illustrated in FIGS. 16A and 20 is changed or omitted.

  Further, the biometric authentication device 1 according to each of the above embodiments can be realized by using, for example, a computer and a program executed by the computer. The biometric authentication device 1 realized by using a computer and a program will be described below with reference to FIG.

FIG. 21 is a diagram showing a hardware configuration of a computer.
As shown in FIG. 21, the computer 9 includes a processor 901, a main storage device 902, an auxiliary storage device 903, an input device 904, an output device 905, an input / output interface 906, a communication control device 907, and a medium. And a driving device 908. These elements 901 to 908 in the computer 9 are connected to each other by a bus 910, and data can be transferred between the elements.

  The processor 901 is a Central Processing Unit (CPU), a Micro Processing Unit (MPU), or the like. The processor 901 controls various operations of the computer 9 by executing various programs including an operating system. The processor 901 also creates, for example, the biometric data shown in FIGS. 5 to 8, FIG. 15, FIG. 16A and FIG. 16B, and FIGS. 17A to 17C, the process of registering the biometric data as a template, and the authentication. Perform processing, etc.

  The main storage device 902 includes a Read Only Memory (ROM) and a Random Access Memory (RAM) which are not shown. In the ROM of the main storage device 902, for example, a predetermined basic control program read by the processor 901 when the computer 9 is started is recorded in advance. Further, the RAM of the main storage device 902 is used by the processor 901 as a work storage area as needed when executing various programs. The RAM of the main storage device 902 can be used to store, for example, the biometric image 6, biometric data, the table 160 used for calculating the extraction stability, the matching score, the alignment parameter, and the like.

  The auxiliary storage device 903 is, for example, a non-volatile memory (including a Solid State Drive (SSD)) such as a flash memory or a Hard Disk Drive (HDD). The auxiliary storage device 903 can store various programs executed by the processor 901, various data, and the like. The auxiliary storage device 903 can be used, for example, for storing a program including biometric data creation processing, biometric data registration processing as a template, and authentication processing. Further, the auxiliary storage device 903 can be used to store, for example, the template database 150, the biometric image 6, biometric data, and the history of authentication processing.

  The input device 904 is, for example, a keyboard device, a touch panel device, or the like. When the operator (user) of the computer 9 performs a predetermined operation on the input device 904, the input device 904 transmits the input information associated with the operation content to the processor 901. The input device 904 can be used as the input device 2 of FIG. 1, for example.

  The output device 905 includes, for example, a display device such as a liquid crystal display device. The output device 905 can be used, for example, to display the operating state of the computer 9 and the results of the template registration process and the authentication process. That is, the output device 905 can be used as the display device 4 in FIG.

  The input / output interface 906 connects the computer 9 and other electronic devices. The input / output interface 906 includes, for example, a Universal Serial Bus (USB) standard connector and the like. Examples of electronic devices that can be connected to the computer 9 via the input / output interface 906 include the biometric image pickup device 3 and the authentication required device 5.

  The communication control device 907 is a device that connects the computer 9 to a communication network and controls various types of communication between the computer 9 and other electronic devices via the communication network. The communication control device 907 can be used, for example, to transmit a history of authentication processing results, templates registered in the template database 150, and the like to a predetermined management server (not shown). When the computer 9 and the management server are connected via the communication control device 907, for example, the template database 150 can be stored in the storage device of the management server.

  The medium driving device 908 reads out a program or data recorded in the portable storage medium 10 and writes data or the like stored in the auxiliary storage device 903 into the portable storage medium 10. As the medium driving device 908, for example, a memory card reader / writer compatible with one or more types of standards can be used. When a memory card reader / writer is used as the medium driving device 908, the portable storage medium 10 includes a memory card (flash memory) of a standard supported by the memory card reader / writer, for example, Secure Digital (SD) standard. ) Etc. are available. As the portable recording medium 10, for example, a flash memory having a USB standard connector can be used. The portable recording medium 10 includes a program including a process of creating the biometric data, a process of registering a template, and an authentication process, biometric data, a template database 150, a table 160 used for calculating extraction stability, an authentication result, and the like. It is available for memory.

  When the computer 9 is equipped with an optical disk drive that can be used as the medium driving device 908, various optical disks that can be recognized by the optical disk drive can be used as the portable recording medium 10. Optical discs that can be used as the portable recording medium 10 include, for example, Compact Disc (CD), Digital Versatile Disc (DVD), and Blu-ray Disc (Blu-ray is a registered trademark).

  When the computer 9 inputs an instruction to start the processing of FIG. 5, the processor 901 reads the program from the auxiliary storage device 903 and executes it. At this time, the processor 901 functions (operates) as the process selection unit 110, the image acquisition unit 120, the data creation unit 130, and the authentication unit 140. The RAM of the main storage device 902, the auxiliary storage device 903, and the like function as a storage unit that stores a biometric image and biometric data, a template database 150, a table 160 that is used to calculate extraction stability, and the like.

  Note that the computer 9 that operates as the biometric authentication device 1 does not need to include all the elements 901 to 908 illustrated in FIG. 21, and it is possible to omit some elements depending on the application and conditions. For example, the computer 9 may be one in which the communication control device 907 and the medium driving device 908 are omitted.

The following supplementary notes will be disclosed with respect to the respective embodiments described above.
(Appendix 1)
A blood vessel image extracting unit that acquires a biometric image including an image of a blood vessel traveling in the living body, and extracts a two-dimensional blood vessel pattern image from the biometric image,
Based on the blood vessel width or the brightness at each position in the running direction of the blood vessel in the two-dimensional blood vessel pattern, the third order in which the blood vessel width or the brightness is added to the image of the two-dimensional blood vessel pattern as a dimension in the depth direction. A blood vessel pattern creation unit that creates an original blood vessel pattern,
A cross-sectional shape of luminance in the direction of the blood vessel width at each position in the running direction of the blood vessel in the image of the two-dimensional blood vessel pattern, and extraction stability of the blood vessel that represents the degree to which the blood vessel is extracted from the biometric image. A relationship with the cross-sectional shape of the brightness of the blood vessel, based on the stability determination unit that determines the extraction stability at each position in the traveling direction of the blood vessel,
An authentication unit that determines whether authentication is possible, based on biometric data including the three-dimensional blood vessel pattern and the extraction stability.
A biometric authentication device, comprising:
(Appendix 2)
The authentication unit is
The three-dimensional blood vessel pattern in the biometric data for registration that is created and registered based on the biometric image in advance, and the three-dimensional blood vessel pattern in the input biometric data created based on the newly acquired biometric image. Based on the degree of overlap with, and the extraction stability at each position in the traveling direction of the blood vessel, it is determined whether or not authentication is possible.
The biometric authentication device according to appendix 1, characterized in that.
(Appendix 3)
The authentication unit is
A registration unit that aligns the three-dimensional blood vessel pattern in the biometric data for registration and the three-dimensional blood vessel pattern in the input biometric data,
A matching reliability calculation unit that calculates matching reliability based on the extraction stability for each of the positions in the traveling direction of the blood vessel that are associated with each other in the two aligned blood vessel patterns;
A matching score is calculated based on the degree of overlap and the matching reliability for each position in the traveling direction of the blood vessel that is associated in the two aligned blood vessel patterns, and the matching score at each position in the blood vessel is calculated. A collation unit that calculates a value obtained by adding
Based on the value obtained by adding the matching score, a determination unit that determines whether the authentication is possible,
The biometric authentication device according to appendix 2, further comprising:
(Appendix 4)
The collation reliability calculation unit increases the collation reliability as the extraction stability is higher,
The matching unit increases the matching score as the matching reliability is higher,
The biometric authentication device according to appendix 3, characterized in that.
(Appendix 5)
The authentication unit is
A registration unit that aligns the three-dimensional blood vessel pattern in the biometric data for registration and the three-dimensional blood vessel pattern in the input biometric data,
A fluctuation rate calculation unit that calculates a blood vessel position fluctuation rate in the depth direction based on the extraction stability for each of the positions in the traveling direction of the blood vessel that are associated with each other in the two aligned blood vessel patterns,
A collation score is calculated based on the degree of overlap and the blood vessel position variation rate for each of the positions in the traveling direction of the blood vessel that are associated in the two aligned blood vessel patterns, and the collation of each position in the blood vessel is performed. A collation unit that calculates a value obtained by adding scores,
Based on the value obtained by adding the matching score, a determination unit that determines whether the authentication is possible,
The biometric authentication device according to appendix 2, further comprising:
(Appendix 6)
The fluctuation rate calculation unit lowers the blood vessel position fluctuation rate as the extraction stability is higher,
The matching unit increases the tolerance in the depth direction of the position of the blood vessel to be added when calculating the matching score as the blood vessel position variation rate is higher,
The biometric authentication device according to appendix 5, characterized in that.
(Appendix 7)
The authentication unit is
Of the respective positions in the traveling direction of the blood vessel in the three-dimensional blood vessel pattern in the biometric data for registration and the three-dimensional blood vessel pattern in the input biometric data, the extraction stability is higher than a predetermined threshold value. Based on the positional relationship between the higher positions, the two three-dimensional blood vessel patterns are aligned,
Based on the degree of overlap of the three-dimensional blood vessel pattern after alignment, and the extraction stability at each position in the traveling direction of the blood vessel, it is determined whether authentication is possible.
The biometric authentication device according to appendix 2, characterized in that.
(Appendix 8)
Computer
Acquiring a biometric image including an image of a blood vessel traveling in a living body, extracting a two-dimensional blood vessel pattern image from the biometric image,
Based on the blood vessel width or the brightness at each position in the running direction of the blood vessel in the two-dimensional blood vessel pattern, the third order in which the blood vessel width or the brightness is added to the image of the two-dimensional blood vessel pattern as a dimension in the depth direction. The original blood vessel pattern is created, and the blood vessel is extracted from the biometric image and the cross-sectional shape of the luminance in the blood vessel width direction at each position in the traveling direction of the blood vessel in the image of the two-dimensional blood vessel pattern. Based on the relationship between the extraction stability of the blood vessel representing the degree and the cross-sectional shape of the brightness of the blood vessel, based on the determination, the extraction stability at each position in the traveling direction of the blood vessel,
Based on biometric data including the three-dimensional blood vessel pattern and the extraction stability, it is determined whether or not authentication is possible.
A biometric authentication method characterized by executing a process.
(Appendix 9)
In the process of determining whether the authentication is possible, the computer is
The three-dimensional blood vessel pattern in the biometric data for registration that is created and registered based on the biometric image in advance, and the three-dimensional blood vessel pattern in the input biometric data created based on the newly acquired biometric image. Based on the degree of overlap with, and the extraction stability at each position in the traveling direction of the blood vessel, it is determined whether or not authentication is possible.
9. The biometric authentication method according to appendix 8, which is characterized in that
(Appendix 10)
In the process of determining whether the authentication is possible, the computer is
Aligning the three-dimensional blood vessel pattern in the biometric data for registration and the three-dimensional blood vessel pattern in the input biometric data,
For each of the positions in the traveling direction of the blood vessel that are associated with each other in the two aligned blood vessel patterns, a matching reliability is calculated based on the extraction stability,
A matching score is calculated based on the degree of overlap and the matching reliability for each position in the traveling direction of the blood vessel that is associated in the two aligned blood vessel patterns, and the matching score at each position in the blood vessel is calculated. Calculate the value by adding
Based on the value obtained by adding the matching score, it is determined whether or not the authentication is possible.
10. The biometrics authentication method according to appendix 9, wherein the biometrics authentication method is performed.
(Appendix 11)
In the process of determining whether the authentication is possible, the computer is
Aligning the three-dimensional blood vessel pattern in the biometric data for registration and the three-dimensional blood vessel pattern in the input biometric data,
Calculating a blood vessel position variation rate in the depth direction based on the extraction stability for each of the positions in the traveling direction of the blood vessel that are associated with each other in the two blood vessel patterns that have been aligned,
A collation score is calculated based on the degree of overlap and the blood vessel position variation rate for each of the positions in the traveling direction of the blood vessel that are associated in the two aligned blood vessel patterns, and the collation of each position in the blood vessel is performed. Calculate the value with the added score,
Based on the value obtained by adding the matching score, it is determined whether or not the authentication is possible.
10. The biometrics authentication method according to appendix 9, wherein the biometrics authentication method is performed.
(Appendix 12)
In the process of determining whether the authentication is possible, the computer is
Of the respective positions in the traveling direction of the blood vessel in the three-dimensional blood vessel pattern in the biometric data for registration and the three-dimensional blood vessel pattern in the input biometric data, the extraction stability is higher than a predetermined threshold value. Based on the positional relationship between the higher positions, the two three-dimensional blood vessel patterns are aligned,
Based on the degree of overlap of the three-dimensional blood vessel pattern after alignment, and the extraction stability at each position in the traveling direction of the blood vessel, it is determined whether or not authentication is possible.
10. The biometrics authentication method according to appendix 9, wherein the biometrics authentication method is performed.
(Appendix 13)
Acquiring a biometric image including an image of a blood vessel traveling in the living body, extracting an image of a two-dimensional blood vessel pattern from the biometric image,
Based on the blood vessel width or the brightness at each position in the running direction of the blood vessel in the two-dimensional blood vessel pattern, the third order in which the blood vessel width or the brightness is added to the image of the two-dimensional blood vessel pattern as a dimension in the depth direction. The original blood vessel pattern is created, and the blood vessel is extracted from the biometric image and the cross-sectional shape of the luminance in the blood vessel width direction at each position in the traveling direction of the blood vessel in the image of the two-dimensional blood vessel pattern. Based on the relationship between the extraction stability of the blood vessel indicating the degree and the cross-sectional shape of the brightness of the blood vessel, the extraction stability at each position in the traveling direction of the blood vessel is determined,
Based on biometric data including the three-dimensional blood vessel pattern and the extraction stability, it is determined whether or not authentication is possible.
A biometric authentication program that causes a computer to perform processing.

1 biometric authentication device 110 processing selection unit 120 image acquisition unit 130 data creation unit 131 blood vessel image extraction unit 132 blood vessel division unit 133 blood vessel pattern creation unit 133a blood vessel width extraction unit 133b depth value calculation unit 133c pattern three-dimensionalization unit 134 stability determination Part 134a Luminance section extraction part 134b Stability calculation part 135 Data output part 140 Authentication part 141 Simple alignment part 142 Collation reliability calculation part 143 Collation part 144 Judgment part 145 Blood vessel position variation rate calculation part 150 Template database 2 Input device 3 Living body Image capturing device 4 Display device 5 Authentication required device 6 Biometric image 602, 602a to 602i, 603, 603a, 603b, 604c to 604e Blood vessel (segment)
7 virtual three-dimensional space 9 computer 901 processor 902 main storage device 903 auxiliary storage device 904 input device 905 output device 906 input / output interface 907 communication control device 908 medium drive device 10 portable recording medium

Claims (7)

A blood vessel image extracting unit that acquires a biometric image including an image of a blood vessel traveling in the living body, and extracts a two-dimensional blood vessel pattern image from the biometric image,
Based on the blood vessel width or the brightness at each position in the running direction of the blood vessel in the two-dimensional blood vessel pattern, the third order in which the blood vessel width or the brightness is added to the image of the two-dimensional blood vessel pattern as a dimension in the depth direction. A blood vessel pattern creation unit that creates an original blood vessel pattern,
A cross-sectional shape of luminance in the direction of the blood vessel width at each position in the running direction of the blood vessel in the image of the two-dimensional blood vessel pattern, and extraction stability of the blood vessel that represents the degree to which the blood vessel is extracted from the biometric image. A relationship with the cross-sectional shape of the brightness of the blood vessel, based on the stability determination unit that determines the extraction stability at each position in the traveling direction of the blood vessel,
An authentication unit that determines whether authentication is possible, based on biometric data including the three-dimensional blood vessel pattern and the extraction stability.
A biometric authentication device, comprising: The authentication unit is
The three-dimensional blood vessel pattern in the biometric data for registration that is created and registered based on the biometric image in advance, and the three-dimensional blood vessel pattern in the input biometric data created based on the newly acquired biometric image. Based on the degree of overlap with, and the extraction stability at each position in the traveling direction of the blood vessel, it is determined whether or not authentication is possible.
The biometric authentication device according to claim 1, wherein The authentication unit is
A registration unit that aligns the three-dimensional blood vessel pattern in the biometric data for registration and the three-dimensional blood vessel pattern in the input biometric data,
A matching reliability calculation unit that calculates matching reliability based on the extraction stability for each of the positions in the traveling direction of the blood vessel that are associated with each other in the two aligned blood vessel patterns;
A matching score is calculated based on the degree of overlap and the matching reliability for each position in the traveling direction of the blood vessel that is associated in the two aligned blood vessel patterns, and the matching score at each position in the blood vessel is calculated. A collation unit that calculates a value obtained by adding
Based on the value obtained by adding the matching score, a determination unit that determines whether the authentication is possible,
The biometric authentication device according to claim 2, further comprising: The authentication unit is
A registration unit that aligns the three-dimensional blood vessel pattern in the biometric data for registration and the three-dimensional blood vessel pattern in the input biometric data,
A fluctuation rate calculation unit that calculates a blood vessel position fluctuation rate in the depth direction based on the extraction stability for each of the positions in the traveling direction of the blood vessel that are associated with each other in the two aligned blood vessel patterns,
A collation score is calculated based on the degree of overlap and the blood vessel position variation rate for each of the positions in the traveling direction of the blood vessel that are associated in the two aligned blood vessel patterns, and the collation of each position in the blood vessel is performed. A collation unit that calculates a value obtained by adding scores,
Based on the value obtained by adding the matching score, a determination unit that determines whether the authentication is possible,
The biometric authentication device according to claim 2, further comprising: The authentication unit is
Of the respective positions in the traveling direction of the blood vessel in the three-dimensional blood vessel pattern in the biometric data for registration and the three-dimensional blood vessel pattern in the input biometric data, the extraction stability is higher than a predetermined threshold value. Based on the positional relationship between the higher positions, the two three-dimensional blood vessel patterns are aligned,
Based on the degree of overlap of the three-dimensional blood vessel pattern after alignment, and the extraction stability at each position in the traveling direction of the blood vessel, it is determined whether authentication is possible.
The biometrics authentication system according to claim 2, wherein Computer
Acquiring a biometric image including an image of a blood vessel traveling in a living body, extracting a two-dimensional blood vessel pattern image from the biometric image,
Based on the blood vessel width or the brightness at each position in the running direction of the blood vessel in the two-dimensional blood vessel pattern, the third order in which the blood vessel width or the brightness is added to the image of the two-dimensional blood vessel pattern as a dimension in the depth direction. The original blood vessel pattern is created, and the blood vessel is extracted from the biometric image and the cross-sectional shape of the luminance in the blood vessel width direction at each position in the traveling direction of the blood vessel in the image of the two-dimensional blood vessel pattern. Based on the relationship between the extraction stability of the blood vessel representing the degree and the cross-sectional shape of the brightness of the blood vessel, based on the determination, the extraction stability at each position in the traveling direction of the blood vessel,
Based on biometric data including the three-dimensional blood vessel pattern and the extraction stability, it is determined whether or not authentication is possible.
A biometric authentication method characterized by executing a process. Acquiring a biometric image including an image of a blood vessel traveling in a living body, extracting a two-dimensional blood vessel pattern image from the biometric image,
Based on the blood vessel width or the brightness at each position in the running direction of the blood vessel in the two-dimensional blood vessel pattern, the third order in which the blood vessel width or the brightness is added to the image of the two-dimensional blood vessel pattern as a dimension in the depth direction. The original blood vessel pattern is created, and the blood vessel is extracted from the biometric image and the cross-sectional shape of the luminance in the blood vessel width direction at each position in the traveling direction of the blood vessel in the image of the two-dimensional blood vessel pattern. Based on the relationship between the extraction stability of the blood vessel representing the degree and the cross-sectional shape of the brightness of the blood vessel, based on the determination, the extraction stability at each position in the traveling direction of the blood vessel,
Based on biometric data including the three-dimensional blood vessel pattern and the extraction stability, it is determined whether or not authentication is possible.
A biometric authentication program that causes a computer to perform processing.