JP2010049694A - Registration device and authentication device for individual authentication system using fingerprint information, computer-readable recording medium recording registration program for the system, and computer-readable recording medium recording authentication program for the system - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a registration device for an individual authentication system using fingerprint information, capable of improving the efficiency of individual authentication in individual authentication techniques by performing pattern discrimination by use of a lower part of a fingerprint image in registration and assigning a verification priority order according the type of pattern, thereby allowing verification using the verification priority order in verification. <P>SOLUTION: The device includes: a registered person fingerprint feature data extraction part 101 which extracts registrant fingerprint feature data from a registrant fingerprint image sampled by a registration fingerprint information input part 100; a registrant fingerprint classification information extraction part 102 which extracts registrant fingerprint classification information from the registrant fingerprint image sampled by the input part 100; a verification priority order determination processing part 310 which determines and outputs a verification priority order according to the registrant fingerprint classification information; and a verification data generation part 121 which writes verification data to a storage part 52 capable of storing data while associating the registrant fingerprint feature data, the registrant fingerprint classification information and the verification priority order. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  The present invention relates to a personal authentication system registration device using fingerprint information, a system authentication device, and a computer-readable recording medium on which the system registration program is recorded, which is suitable for personal authentication using, for example, a fingerprint pattern. The present invention also relates to a computer-readable recording medium on which the system authentication program is recorded.

  In recent years, personal authentication devices using biometric information such as fingerprints have been sold. This biometric information (biometric information, biometrics) uses, for example, fingerprint, palm print, finger shape, palm shape, voice, retina, iris, face image, dynamic signature, blood vessel pattern, keystroke, etc. And more reliable than passwords. Of biometric information, fingerprints are often used.

  In this personal authentication system using fingerprints, fingerprints are collated with samples by brute force. For example, a collation format (collation method) called a one-fingerprint-to-multiple registered fingerprint format (hereinafter referred to as one-to-many format) is used. It is used. This round-robin collation is a method of collating the registered collation data in order from the beginning, and if the person's data is accidentally at the top, the collation using a pattern is intentionally performed. It is expected that the processing will be completed quickly without doing so.

By the way, since this brute force verification method does not specify an individual using ID (Identification) or the like, a large amount of calculation is required until the person is authenticated at the time of verification.
Therefore, in order to avoid an enormous processing time, various methods for reducing the matching time in the one-to-many format by detecting the pattern type of the fingerprint and collating only the same pattern type can be considered. Yes.

  For example, Japanese Patent Laid-Open No. 7-29003 (hereinafter referred to as publicly known document 1) discloses a fingerprint collation apparatus that can shorten the collation processing time. The fingerprint collation apparatus disclosed in this publicly known document 1 extracts collation feature data from a fingerprint image and performs collation with respect to a plurality of fingerprints (hereinafter referred to as file fingerprints) registered in a file fingerprint data storage unit. The feature data for matching is extracted from a fingerprint image (hereinafter referred to as a search fingerprint), and the feature data for matching extracted from the search fingerprint is the file fingerprint of any file fingerprint registered in the file fingerprint data storage unit. It is determined whether or not the matching feature data has the highest similarity.

As a result, in the fingerprint collation, first, the additional information of the authentication fingerprint data is compared with the additional information of the registered fingerprint data for collation, and the file fingerprint in which the mismatch is detected is removed from the collation partner, and comprehensive collation is performed. Time has been shortened.
According to the fingerprint authentication method using this fingerprint pattern, if the fingerprint pattern for registration verification and the fingerprint pattern for authentication do not match, it is determined that the person is not the person and the process is completed. Further, according to this method, when collecting a fingerprint image, an angle between a finger and a fingerprint scanner (hereinafter sometimes abbreviated as a scanner), a finger position on the scanner, and the like are collected. Depending on the situation (hereinafter referred to as the fingerprint collection situation), a pattern different from the actual pattern may be determined. In this case, not only normal personal authentication cannot be performed, but there are cases where only a registered fingerprint of a pattern that does not match the collected pattern is collated. In addition, when there is no data matching the registration fingerprint group having the same pattern as the authentication fingerprint, a technique of brute force verification for a registration verification fingerprint having a pattern different from the pattern of the authentication fingerprint is also used. According to this method, there is no big difference from mere brute force verification.

  In addition, various methods for determining a fingerprint pattern have already been proposed. Japanese Laid-Open Patent Publication No. 9-44666 (hereinafter referred to as publicly known document 2) discloses a skin pattern and fingerprint pattern classification device for tracking and classifying skin pattern lines such as palm patterns and fingerprints. The skin pattern line tracking device disclosed in this known document 2 extracts a singular point that appears in a specific pattern part, and there is a characteristic line around the singular point (hereinafter referred to as a characteristic line). The feature line is traced, the fingerprint pattern feature is extracted, and the fingerprint pattern type is determined.

  FIG. 33 is a diagram for explaining the types of fingerprint patterns, showing five types of normal bow-shaped pattern α, protruding bow-shaped pattern β, right-flowing pattern γ, left-flowing pattern δ, and spiral pattern ε. Has been. In the classification shown in FIG. 33, the number of singular points (hereinafter referred to as core-type singular points) appearing in the semicircular portion is normally 0 for the ordinary arch pattern α, and the protruding arch pattern β , The right-flowing pattern γ and the left-flowing pattern δ are one, and the spiral pattern ε is two.

  Further, the number of singular points (hereinafter referred to as delta type singular points) appearing in the portion of the triangle also appears as many as the core type singular points. Fingerprints obtained by pressing a finger against a scanner or the like may be difficult to detect automatically because, for practical reasons, a fingerprint image cannot be obtained over a sufficiently wide range. For this reason, the known document 2 describes the case of fingerprint classification using core-type singular points.

  The types of right-flowing pattern γ, left-flowing pattern δ, and protruding bow pattern β are similar patterns that are difficult to classify these two types (deformation pattern γ, δ and protruding bow pattern β). There are many. Therefore, when distinguishing these two types of patterns, attention is paid to whether or not a contour line is included. In other words, the striated patterns γ and δ both include a scalloped line, whereas the protruding arched pattern β is formed from a arched line and does not include a scalloped line.

  FIG. 34 (a) is a diagram for explaining a vertical line. The ridges (curves forming a pattern) al-ar and bl-dr shown in FIG. 34A each have a horseshoe shape. And these ridges follow the two lines tracked from the horseshoe-shaped tip and the end flows to one side, the left-flowing line is roughly on the left side, Also, the right-flowing line (not shown) is approximately on the right side of both lines.

FIG. 34 (b) is a diagram for explaining the protruding arcuate line. In the ridges cl-cr and dl-dr shown in FIG. 34 (b), the lines traced starting from the arcuate protrusion are divided into the right side and the left side, respectively.
Generally, ridges are not necessarily continuous but have breaks. In this case, the pattern type is determined by appropriately complementing the discontinuity, tracking the ridge, and determining whether the tracked end matches the above-described conditions. Thus, the type of arcuate line or warp line is determined.

In the stag pattern shown in FIG. 34 (a), the core type singular point is located in a portion surrounded by a horseshoe shape. In other words, the contour pattern is located around the core type singular point. On the other hand, in the protruding arch pattern shown in FIG. 34 (b), the core-type singular point is surrounded by an arcuate line.
Therefore, the type of the contour pattern and the protruding arc pattern is determined based on whether a line surrounding the core type singular point is extracted and the line is an arcuate line or a curved line. Is done.

  Next, most spiral patterns can be divided into two types: core-type singularities and ridge flow around them. FIGS. 35A and 35B are diagrams for explaining a spiral pattern. In the pattern shown in FIG. 35A, two core-type singular points are connected via a ridge line or a valley line, and the ridges around the core form a ring. In the pattern shown in FIG. 35 (b), the line connected to the core-type singular point approximately forms a vortex.

Therefore, when the shape of a characteristic line (characteristic line) connected to the core type singular point cannot be classified into two types, it can be determined that the detected singular point is not a singular point due to a spiral pattern. So we can verify the spiral pattern.
FIG. 36 is a diagram showing an example of a contour pattern, in which two contour lines and two singular points respectively appear. The sword pattern shown in FIG. 36 has two stagnation lines, and a few people have it. The type in this case is determined by checking a characteristic line connected to a singular point, and can prevent misclassification into a spiral pattern.

  In addition to the above-mentioned known documents 1 and 2, many techniques for determining an accurate pattern from the input fingerprint image as much as possible have been proposed.

However, according to the prior art, when the type determination fails, there is a problem that authentication is not performed and the strict type determination is required again, which increases the amount of calculation.
In addition, when the check pattern shown in FIG. 36 is collated using the above-described determination method, the fingerprint is collated by strict determination, and therefore the type determination failure is not allowed. Also, if the area of the input fingerprint information is small, the pattern cannot be determined accurately, and a fingerprint image with a large area is required to make a precise classification determination. Therefore, according to the personal authentication using the pattern of the prior art, there is a problem that accurate personal authentication cannot be performed when the type determination fails.

  The present invention was devised in view of such problems, and in a personal authentication system, at the time of registration, a pattern is identified using a lower part of a fingerprint image, and a group is assigned to each pattern type to give a collation priority. Personal verification using fingerprint information that can be used to verify the identity even when the patterns do not match completely, and can greatly reduce the amount of calculation, by matching using the priority of the matching at the time of matching. It is an object of the present invention to provide a system registration device, a system authentication device, a computer readable recording medium recording the system registration program, and a computer readable recording medium recording the system authentication program.

For this reason, the registration device for personal authentication system using fingerprint information of the present invention uses a fingerprint information input unit for registration for collecting registrant fingerprint information in a system for performing personal authentication using fingerprint information. A registration device that obtains and writes verification data from input information, and a registrant fingerprint feature data extraction unit that extracts registrant fingerprint feature data from a registrant fingerprint image collected by a registration fingerprint information input unit; A registrant fingerprint classification information extraction unit that extracts registrant fingerprint classification information from a registrant fingerprint image collected by the registration fingerprint information input unit, and a verification that determines and outputs a collation priority according to the registrant fingerprint classification information The priority determination processing unit and the verification data are written in the storage unit that can hold the data by associating the registrant fingerprint feature data, the registrant fingerprint classification information, and the verification priority. It is characterized by being configured to include the over data generating unit.
In addition, an authentication apparatus for a personal authentication system using fingerprint information according to the present invention is a storage unit capable of holding data indicating whether or not a person to be verified is the person himself / herself in a system for performing personal authentication using fingerprint information. Authentication device for collation / determination using collation data written in the authentication object person from the fingerprint image of the authentication person collected by the authentication fingerprint information input unit for collecting the fingerprint information of the person to be authenticated An authentication target person fingerprint feature data extraction unit for extracting fingerprint feature data; an authentication target person fingerprint classification information extraction unit for extracting authentication target person fingerprint classification information from the authentication target person fingerprint image collected by the authentication fingerprint information input unit; A collation priority definition unit capable of outputting a collation priority for each collation data group classified based on the registrant fingerprint classification information, authentication target person fingerprint feature data, authentication target person identification Based on the classification information and the collation priority matches with the authentication target person fingerprint characteristic data and the registrant fingerprint characteristic data, is characterized in that it is configured to include a collation section that outputs a verification result.
Furthermore, a computer-readable recording medium on which a registration program for a personal authentication system using fingerprint information according to the present invention is recorded is obtained from a registrant computer used in a system for performing personal authentication using fingerprint information. Registrant fingerprint feature data extraction unit that extracts registrant fingerprint feature data from the registered registrant fingerprint image, and fingerprint categorization information extraction that extracts registrant fingerprint classification information from the registrant fingerprint image and extracts registrant fingerprint classification information A collation priority determination processing unit that determines and outputs a collation priority according to the registrant fingerprint classification information, collation data, registrant fingerprint feature data, registrant fingerprint classification information, and collation priority A program for making it function as a collation data generation unit that writes data to a storage unit that can hold data in association is recorded.
A computer-readable recording medium on which an authentication program for a personal authentication system using fingerprint information according to the present invention is recorded is a computer used in a system for authenticating an individual using fingerprint information. Authentication target person fingerprint feature data extraction unit for extracting authentication target person fingerprint feature data from the collected authentication target person fingerprint image, and authentication target person fingerprint classification information extraction for extracting authentication target person fingerprint classification information from the authentication target person fingerprint image A collation priority definition unit capable of outputting a collation priority for each collation data group classified based on registrant fingerprint classification information, authentication target person fingerprint feature data, authentication target person fingerprint classification information, and collation priority The authentication target person fingerprint feature data and the registrant fingerprint feature data are collated based on the ranking and function as a collation unit that outputs the collation result. It is characterized in that the program for is recorded.
Furthermore, a personal authentication system using fingerprint information related to the present invention is a system for performing personal authentication using fingerprint information, a storage unit that can hold data, and a registration unit for collecting registrant fingerprint information. Using a registration device that obtains verification data from information input using the fingerprint information input unit and writes it in the storage unit, and verification using the verification data written in the storage unit whether or not the person to be authenticated is the principal A registrant fingerprint feature data extraction unit that extracts a registrant fingerprint feature data from a registrant fingerprint image collected by the registration fingerprint information input unit, and a registration fingerprint information input. Registrant fingerprint classification information extraction unit that extracts registrant fingerprint classification information from the registrant fingerprint image collected by the registrant, collation data, registrant fingerprint feature data and registrant fingerprint classification information And a verification data generation unit to be written in the storage unit, and further, the authentication device uses the authentication target person fingerprint image collected by the authentication fingerprint information input unit for collecting the authentication subject person fingerprint information. Authentication target person fingerprint feature data extraction unit for extracting user fingerprint feature data, and authentication subject person fingerprint classification information extraction unit for extracting authentication subject person fingerprint classification information from the authentication subject person fingerprint image collected by the authentication fingerprint information input unit A collation priority definition unit capable of outputting a collation priority for each collation data group classified based on the registrant fingerprint classification information, authentication subject person fingerprint feature data, authentication subject person fingerprint classification information, and collation priority order Based on the authentication target person fingerprint feature data and the registrant fingerprint feature data, and a collation unit that outputs a collation result is characterized.

Further, the registration device includes a collation priority order determination processing unit that determines and outputs a collation priority according to the registrant fingerprint classification information, and the collation data generation unit includes the collation data as registrant fingerprint feature data. The registrant fingerprint classification information and the collation priority order may be associated and written to the storage unit.
The registrant fingerprint feature data extraction unit of the registration device and the authentication subject fingerprint feature data extraction unit of the authentication device are shared, and the registrant fingerprint classification information extraction unit of the registration device and the authentication subject fingerprint classification of the authentication device The information extraction unit may be shared.

  The personal authentication system registration / authentication method using fingerprint information related to the present invention includes a first extraction step of extracting registrant fingerprint feature data from a collected registrant fingerprint image, and extraction at the first extraction step. Registrant fingerprint feature data registration step of writing the registered registrant fingerprint feature data in a storage unit capable of holding the data, a second extraction step of extracting registrant fingerprint classification information from the registrant fingerprint image, and a second extraction step Step for registering registrant fingerprint classification information that writes the registrant fingerprint classification information extracted in step 1 to the storage unit, and for collation data to be written to the storage unit in association with the registrant fingerprint feature data and registrant fingerprint classification information A data registration step, a verification data extraction step for extracting verification data from the storage unit, and an authentication target person fingerprint feature data from the collected authentication target person fingerprint image. A third extraction step for extracting data, a fourth extraction step for extracting authentication subject person fingerprint classification information from the authentication subject person fingerprint image, and a collation priority for each collation data group classified based on the registrant fingerprint classification information Collation / determination for collating / determining whether or not the person to be authenticated is the person based on the collation priority order defining step for outputting the rank, the authentication target person fingerprint feature data, the authentication person fingerprint classification information and the collation priority order It is characterized by comprising steps.

  The determination method for a personal authentication system using fingerprint information related to the present invention includes a center detection step for detecting a fingerprint center coordinate from an authentication subject fingerprint image, and a division for dividing the authentication subject fingerprint image into a plurality of unit blocks. A recognition step for recognizing a central unit block having a fingerprint center coordinate detected in the center detection step among a plurality of unit blocks divided in the division step, and a central unit block recognized in the recognition step The block group generation step for generating a block group having a plurality of unit blocks from the portion on the opposite side of the fingerprint tip side, and each of the plurality of unit blocks constituting the block group generated in the block group generation step, Unit block flow direction judgment to determine local fingerprint wave flow direction A method is characterized in that it is constructed to include a fingerprint classification information judgment step of judging the fingerprint classification information based on the determined flow direction of the fingerprint wave in unit block flow direction determination step.

According to the personal authentication system registration device using the disclosed fingerprint information and the computer-readable recording medium on which the registration program for the system is recorded, it is not necessary to calculate the priority of collation when inputting the authentication data. There is an advantage that the amount can be reduced.
Furthermore, according to the authentication apparatus for personal authentication system using the disclosed fingerprint information and the computer-readable recording medium in which the authentication program for the system is recorded, it is possible to greatly reduce the calculation effort and perform efficient collation.
According to the personal authentication system using fingerprint information related to the present invention, the registration / authentication method for the system, and the determination method for the system, the following effects or effects can be obtained.
(1) The personal authentication system obtains verification data from information input using a storage unit capable of holding data and a registration fingerprint information input unit for collecting registrant fingerprint information, and writes the data to the storage unit. The registration device is collected by the registration fingerprint information input unit, and includes a registration device and an authentication device that verifies / determines whether or not the person to be authenticated is the identity by using the verification data written in the storage unit. A registrant fingerprint feature data extraction unit that extracts registrant fingerprint feature data from the registered registrant fingerprint image, and a registrant fingerprint classification that extracts registrant fingerprint classification information from the registrant fingerprint image collected by the registration fingerprint information input unit An information extraction unit; and a collation data generation unit that writes the collation data to the storage unit in association with the registrant fingerprint feature data and the registrant fingerprint classification information. Authentication target person fingerprint feature data extraction unit for extracting authentication target person fingerprint feature data from the authentication target person fingerprint image collected by the authentication fingerprint information input unit for collecting information, and authentication fingerprint information input unit Authentication target person fingerprint classification information extraction unit for extracting authentication target person fingerprint classification information from the authentication target person fingerprint image, and output of collation priority for each collation data group classified based on registrant fingerprint classification information Based on the verification priority definition unit, the authentication target person fingerprint feature data, the authentication target person fingerprint classification information, and the verification priority order, the authentication target person fingerprint feature data and the registrant fingerprint characteristic data are verified, and a verification result is output. Since it is configured with a collation unit, the strictness of pattern determination can be relaxed, which increases the processing speed of pattern determination and is simple Ri compared to matching quick can be authenticated with the person.

  (2) The registration device includes a collation priority order determination processing unit that determines and outputs a collation priority according to the registrant fingerprint classification information, and the collation data generation unit converts the collation data into the registrant fingerprint feature data. Can be configured so that the registered fingerprint classification information and the collation priority are written to the storage unit in association with each other. In this way, it is not necessary to calculate the collation priority when inputting authentication data. There is an advantage that the amount can be reduced.

(3) The registrant fingerprint feature data extraction unit of the registration device and the authentication subject fingerprint feature data extraction unit of the authentication device are shared, and the registrant fingerprint pattern extraction unit of the registration device and the authentication subject fingerprint of the authentication device The classification information extraction unit may be shared and configured, and there is an advantage that the cost of the apparatus can be reduced.
(4) According to the personal authentication system registration / authentication method using fingerprint information related to the present invention, the first extraction step of extracting registrant fingerprint feature data from the collected registrant fingerprint image, and the first extraction A registrant fingerprint feature data registration step of writing the registrant fingerprint feature data extracted in the step into a storage unit capable of holding the data, a second extraction step of extracting registrant fingerprint classification information from the registrant fingerprint image, A registrant fingerprint classification information registration step for writing the registrant fingerprint classification information extracted in the second extraction step into the storage unit, and a collation data as a storage unit in association with the registrant fingerprint feature data and the registrant fingerprint classification information The verification data registration step for writing to the verification data, the verification data extraction step for extracting the verification data from the storage unit, and the authentication target person identification from the collected authentication target person fingerprint image A third extraction step for extracting feature data, a fourth extraction step for extracting authentication subject person fingerprint classification information from the authentication subject person fingerprint image, and collation for each collation data group classified based on the registrant fingerprint classification information A collation priority order defining step for outputting a priority order, and collation / determination based on the authentication subject person fingerprint feature data, the authentication subject person fingerprint classification information, and the collation priority order for collation / determination Since the determination step is configured, there is an advantage that even if the input image is not completely reproduced and the pattern cannot be accurately determined, it can be determined as a pattern having a high collation priority.

  (5) According to the determination method for a personal authentication system using fingerprint information related to the present invention, a center detection step for detecting a fingerprint center coordinate from the authentication subject person fingerprint image, and the authentication subject person fingerprint image as a plurality of unit blocks A recognition step for recognizing a central unit block having the fingerprint center coordinates detected in the center detection step among the plurality of unit blocks divided in the division step, and a recognition step A block group generation step for generating a block group having the plurality of unit blocks from a portion on the opposite side of the fingerprint tip side with respect to the recognized central unit block, and the block group generated in the block group generation step. For each of the plurality of unit blocks constituting the unit block, a local fingerprint wave flow direction is determined. Since it comprises a block flow direction determining step and a fingerprint classification information determining step for determining fingerprint classification information based on the flow direction of the fingerprint wave determined in the unit block flow direction determining step, it is not necessarily a pattern. Even if they do not match, the user authentication can be ensured and unnecessary collation can be suppressed to the minimum. Therefore, there is an advantage that the labor of calculation can be largely omitted and efficient collation is possible. Furthermore, a desired strictness can be set in collation, and precise collation and collation giving priority to speed can be properly used as necessary.

  (6) The determination unit is configured to determine the identity / non-identity in consideration of the collation priority, and is configured to determine the identity / non-identity in consideration of the authentication target person fingerprint classification information. Or the matching rate output from the collating unit may be configured to determine the identity / non-identity as the collation result. In this way, the user can quickly perform without performing redundant collation. Has the advantage of authenticating the person.

  (7) Both the registrant fingerprint classification information extraction unit and the authentication target person fingerprint classification information extraction unit use the portion of the fingerprint image opposite to the fingerprint tip side with respect to the fingerprint center of the fingerprint image. It may be configured to extract the target person fingerprint classification information, and in this way, the direction in the specific area below the fingerprint center is focused, so the input image is incomplete and accurate Even if a pattern cannot be identified, there is an advantage that it can be compared with a pattern having a high priority. Furthermore, the fingerprint image can be collated even in a small area, and the strictness can be relaxed and the processing speed until obtaining the pattern determination result is improved as compared with the pattern determination method that has already been proposed. There are advantages.

  (8) When the determination unit cannot determine the person / non-person until the collation operation at the predetermined collation priority is completed, the collation unit restricts the user authentication at the collation priority lower than the predetermined collation priority. In this way, when the collation priority is low and the collation priority of the pattern having a low hit rate is reached, the personal authentication is interrupted and efficient operation can be performed.

(9) Since the authentication device is configured to include an authentication subject person fingerprint feature data extraction unit, an authentication subject person fingerprint classification information extraction unit, a collation priority definition unit, and a collation unit, the calculation effort is reduced. It can be omitted greatly, and efficient collation can be performed.
(10) The collation priority definition unit may be configured to hold data associating the registrant fingerprint classification information with the collation priority as collation data. In this way, the input image Even if the pattern is not completely reproduced and the pattern cannot be accurately discriminated, it can be collated with the collation data having a high collation priority.

  (11) The storage unit is configured to hold the verification data for each classification type of the registrant fingerprint classification information, or the storage unit includes the registrant fingerprint feature data storage unit and the registrant fingerprint classification information. In this way, even if the patterns do not match, authentication can be ensured and unnecessary verification can be suppressed to a minimum, greatly increasing the amount of calculation. Can be reduced.

  (12) According to the personal authentication system using biometric information related to the present invention, the registration device includes a registrant biometric information feature data extraction unit, a registrant bioclassification information extraction unit, and a collation data generation unit. In addition, since the authentication device is configured to include an authentication subject biometric information feature data extraction unit, an authentication subject biometric classification information extraction unit, a collation priority order definition unit, and a collation unit, collation priority Since the collation is performed in consideration of the order, there is an advantage that the biometric information input as the biometric information for authentication and the biometric information of the registration data for collation can be surely collated even if they do not completely match.

  (13) The biometric information feature data extraction unit of the registration device and the authentication device may be shared, and the biometric classification information extraction unit of the registration device and the authentication device may be shared. For example, the registration and authentication input devices can be used together, and for example, collation can be performed without adding a new device for inputting seal data.

(14) The registration device may include a biometric information input unit for registration, a registrant biometric information feature data extraction unit, a registrant bioclassification information extraction unit, and a collation data generation unit. In this way, even when the patterns do not necessarily match, personal authentication can be ensured and unnecessary collation can be suppressed to a minimum.
(15) The authentication device may include an authentication subject biometric information feature data extraction unit, an authentication subject biometric classification information extraction unit, a collation priority definition unit, a collation unit, and a determination unit. In this way, it is possible to quickly authenticate the person as compared with mere brute force verification.

  (16) The program recorded on the computer-readable recording medium of the present invention causes the computer to function as a registrant fingerprint feature data extraction unit, a fingerprint classification information extraction unit, and a collation data generation unit. The authentication target person fingerprint feature data extraction unit, the authentication target person fingerprint classification information extraction unit, the collation priority definition unit, and the collation unit function so that they can be easily installed on the computer, and thus use fingerprint information. Therefore, there is an advantage that a system for performing personal authentication can be constructed efficiently.

1 is a configuration diagram of a personal authentication system using fingerprint information according to a first embodiment of the present invention. It is a block diagram of a registration device concerning a 1st embodiment of the present invention. It is a block diagram of the authentication device concerning a 1st embodiment of the present invention. (A) is a figure for demonstrating extraction of the fingerprint feature data based on 1st Embodiment of this invention, (b) is a figure which shows the fingerprint wave in a unit block. It is a figure for demonstrating extraction of the other fingerprint feature data based on 1st Embodiment of this invention. (A)-(d) is a figure for demonstrating the direction which concerns on 1st Embodiment of this invention, respectively. It is a flowchart which shows the data registration procedure for collation which concerns on 1st Embodiment of this invention. It is a flowchart which shows the collation information generation procedure which concerns on 1st Embodiment of this invention. It is a flowchart which shows the outline of the collation procedure which concerns on 1st Embodiment of this invention. It is a flowchart which shows the collation procedure which concerns on 1st Embodiment of this invention. It is a flowchart which shows the other collation procedure which concerns on 1st Embodiment of this invention. It is a figure which shows an example of the fingerprint image determined as a right flow type. It is a flowchart of the procedure determined as the right flow type which concerns on 1st Embodiment of this invention. It is a flowchart of the procedure determined as the right flow type which concerns on 1st Embodiment of this invention. It is a figure which shows an example of the fingerprint image determined to a left flow type. It is a flowchart of the procedure determined with the left flow type which concerns on 1st Embodiment of this invention. It is a flowchart of the procedure determined with the left flow type which concerns on 1st Embodiment of this invention. (A), (b) is a figure which shows an example of the fingerprint image determined as a waveform. It is a flowchart of the procedure determined with the waveform which concerns on 1st Embodiment of this invention. (A), (b) is a figure which shows an example of the fingerprint image determined as a vortex type | mold both. It is a flowchart of the procedure determined with the vortex type which concerns on 1st Embodiment of this invention. It is a flowchart which shows the example of a definition of the right flow type | mold collation priority based on 1st Embodiment of this invention. It is a flowchart which shows the example of a definition of the left flow type | mold collation priority based on 1st Embodiment of this invention. It is a flowchart which shows the example of a definition of the waveform collation priority based on 1st Embodiment of this invention. It is a flowchart which shows the example of a definition of the vortex-type collation priority based on 1st Embodiment of this invention. It is a block diagram of the authentication apparatus which concerns on the 1st modification of 1st Embodiment of this invention. It is a block diagram of the registration apparatus which concerns on the 2nd modification of 1st Embodiment of this invention. It is a block diagram of the authentication apparatus which concerns on the 2nd modification of 1st Embodiment of this invention. It is a flowchart which shows the example of a definition of the waveform collation priority based on the 2nd modification of 1st Embodiment of this invention. It is a flowchart which shows the authentication procedure which concerns on the 2nd modification of 1st Embodiment of this invention. It is a block diagram of the registration apparatus in the personal authentication system using the biometric information which concerns on 2nd Embodiment of this invention. It is a block diagram of the authentication apparatus in the personal authentication system using the biometric information which concerns on 2nd Embodiment of this invention. It is a figure for demonstrating the kind of fingerprint pattern. (A) is a figure for demonstrating a warp-like line, (b) is a figure for demonstrating a protrusion arcuate line. (A), (b) is a figure for demonstrating a spiral pattern, respectively. It is a figure which shows an example of a stripe pattern.

Embodiments of the present invention will be described below with reference to the drawings.
(A) Description of First Embodiment of the Present Invention FIG. 1 is a configuration diagram of a personal authentication system using fingerprint information according to a first embodiment of the present invention. A personal authentication system (hereinafter sometimes abbreviated as “system”) 49 shown in FIG. 1 performs personal authentication using fingerprint information, and includes a storage unit 52, a registration device 50, and an authentication device 51. Configured.

  Here, the storage unit 52 holds collation data (collation fingerprint element or collation information) for each classification type of fingerprint classification information (fingerprint pattern information), and can hold registrant fingerprint feature data. A fingerprint feature data storage unit (registrant fingerprint feature data storage unit) 111 and a fingerprint pattern information storage unit (registrant fingerprint classification information storage unit) 112 capable of holding registrant fingerprint pattern information are included. The verification data is data to be referred to at the time of personal authentication, and is a combination of fingerprint feature data and a pattern type, and the classification type represents a classification category. The function of the storage unit 52 is realized by, for example, a hard disk. In addition, the information held as the feature data includes the position and number of the feature points of the fingerprint or the relative relationship between the feature points, and the registered data is compared by comparing whether these pieces of information are equal. And whether or not the data input for authentication is equal.

As will be described later, both the fingerprint feature data storage unit 111 and the fingerprint pattern information storage unit 112 are shared with the authentication device 51. In the following description, the storage unit 52 may be referred to as a database 52.
Further, the registration device 50a, the authentication devices 51a and 51b, and the systems 49a and 49b shown in FIG. 1 are a modification device and a system of the registration device 50, the authentication device 51, and the system 49, respectively. This will be described in detail in a first modification and a second modification of the first embodiment.

  Furthermore, the registration device 50 obtains collation data from information input using a fingerprint information input unit (registration fingerprint information input unit) 100 (see FIG. 2) for collecting registrant fingerprint information, and stores it. 52 is written. Then, the authentication device 51 performs collation / determination using the collation data written in the storage unit 52 as to whether or not the person to be authenticated is the person himself / herself.

  FIG. 2 is a block diagram of the registration device 50 according to the first embodiment of the present invention. In addition to the fingerprint information input unit 100, the registration device 50 shown in FIG. 2 includes a fingerprint feature data extraction unit (registrant fingerprint feature data extraction unit) 101 and a fingerprint pattern information extraction unit (registrant fingerprint classification information extraction unit). 102 and a collation data generation unit 121.

The function of the fingerprint information input unit 100 is realized by, for example, a fingerprint scanner (not shown). The fingerprint information input unit 100 not only collects biometric information of a registrant as a fingerprint for registration, but also is used to input an authentication fingerprint different from registration data, as will be described later. It is done.
The fingerprint feature data extraction unit 101 extracts registrant fingerprint feature data from the registrant fingerprint image collected by the fingerprint information input unit 100, and the fingerprint pattern information extraction unit 102 collects by the fingerprint information input unit 100. The registrant fingerprint image information (registrant fingerprint classification information) is extracted from the registrant fingerprint image, and the collation data generation unit 121 uses the registrant fingerprint feature data, the registrant fingerprint pattern information, and the collation data. Are stored in the storage unit 52 in association with each other.

2 having the same reference numerals as those described above have the same or similar functions as those described above, and redundant description thereof will be omitted.
Thus, a registrant fingerprint image is collected by the fingerprint information input unit 100 according to an instruction of a control unit (not shown), and a registrant fingerprint feature is extracted from the collected registrant fingerprint image by the fingerprint feature data extraction unit 101. Data is extracted. Also, the fingerprint pattern information extraction unit 102 extracts registrant fingerprint pattern information from the collected registrant fingerprint images, and classifies the pattern types.

  The collation data generating unit 121 associates (combines) the registrant fingerprint feature data with the registrant fingerprint pattern information based on the extracted collation feature data and the fingerprint pattern, and at the time of personal authentication. The verification data to be used is generated, and the verification data is written in the fingerprint feature data storage unit 111 and the fingerprint pattern information storage unit 112, respectively.

  Note that the two types of fingerprint feature data storage unit 111 and fingerprint pattern information storage unit 112 for collation shown in FIG. 2 are each configured to hold information individually. Since this is one-to-one information linked to each other, this function can be realized by using a single storage device. The use of a single storage device can also be applied to each modification of the first embodiment described later and the system in the second embodiment.

  FIG. 3 is a block diagram of the authentication device 51 according to the first embodiment of the present invention. In addition to the fingerprint information input unit 100, the authentication device 51 shown in FIG. 3 includes a fingerprint feature data extraction unit (authentication subject person fingerprint feature data extraction unit) 101 and a fingerprint pattern information extraction unit (authentication subject person fingerprint classification information extraction). Part) 102, a collation priority definition part 113, a fingerprint information collation part (collation part) 114, and a determination part 115.

  This fingerprint feature data extraction unit 101 extracts authentication subject person fingerprint feature data from the authentication subject person fingerprint image collected by the authentication fingerprint information input unit 100 for collecting the subject person fingerprint information. The pattern information extraction unit 102 extracts authentication subject person fingerprint pattern information (authentication subject person fingerprint classification information) from the authentication subject person fingerprint image collected by the authentication fingerprint information input unit 100.

In addition, the fingerprint feature data extraction unit 101 and the fingerprint pattern information extraction unit 102 extract the registrant fingerprint pattern information by using the part of the fingerprint center of the registrant fingerprint image that is opposite to the fingerprint tip side. It has become.
FIG. 4A is a view for explaining extraction of fingerprint feature data according to the first embodiment of the present invention. The fingerprint image shown in FIG. 4A (hereinafter sometimes referred to as an image) is divided into 40 parts vertically and horizontally to form 1600 unit blocks (small blocks). Here, the number of pixels of one unit block is 8 pixels × 8 pixels. The reason for this number of pixels is to make it possible to determine the local fingerprint wave flow direction by each unit block alone.

  FIG. 4B shows a fingerprint wave in the unit block. The unit block shown in FIG. 4B is 8 pixels × 8 pixels, and the fingerprint wave is displayed at the image level. The direction of the fingerprint wave flow is also called a direction, and the local direction of the fingerprint wave flow means the direction in the unit block. An example of the direction is as shown in FIGS. In the following description, the directions of right flow, left flow, wave, vortex, horizontal, and vertical are referred to as right flow system, left flow system, wave system, vortex system, horizontal system, and vertical system, respectively. is there.

When the number of pixels of the input image is, for example, 320 pixels × 320 pixels and the fingerprint pattern is clearly expressed in the image, the input image has 8 pixels × 8 pixels. Divided into unit blocks. Further, even if one unit block is used alone, the local direction is determined.
Further, two core blocks are specified for the unit block located below the fingerprint center (fingerprint center coordinates) from the 1600 unit blocks shown in FIG.

  The core block is a group of unit blocks and is also called a block group (block group). FIG. 4 (a) shows two core blocks, as if two windows were provided. The pattern determination is determined based on the direction presence ratio of each unit block constituting the core block.

Thereby, the fingerprint pattern is determined based on the core block.
In the following description, these core blocks are referred to as a block group A and a block group B, respectively. In addition, for a unit block having a fingerprint center coordinate (hereinafter referred to as a center unit block or a center block), a plurality of unit blocks are specified from a portion on the opposite side to the fingerprint front end side, and a core block is generated. .

  Furthermore, in the following description, it is assumed that an input image of 320 pixels × 320 pixels is divided into unit blocks of 8 pixels × 8 pixels and pattern determination is performed. Note that if the fingerprint pattern is not clearly shown in the image, or based on the size of the image or the shape of the fingerprint in the image, the size of one unit block is changed by the number of pixels that can identify the direction. it can. Therefore, an image of 320 pixels × 320 pixels can be divided into unit blocks having a divisor of 320 (2, 4, 5, 10, etc.) according to the fingerprint collection situation.

That is, the division number is changed according to the number of pixels of the input image from the fingerprint information input unit 100. For example, when the input image is 300 pixels × 300 pixels, it cannot be divided by 8 pixels × 8 pixels. Accordingly, in this case, the image is divided into 900 unit blocks each having 10 pixels × 10 pixels, and the pattern is determined.
The fingerprint pattern information extraction unit 102 (see FIG. 2) selects a block group according to the pattern type. The fingerprint pattern information extraction unit 102 uses different block groups, for example, when determining a wave pattern and when determining a vortex pattern. These will be described later with reference to FIGS. 19 and 21.

Next, FIG. 5 is a diagram for explaining extraction of other fingerprint feature data according to the first embodiment of the present invention. The image shown in FIG. 5 shows a block group A and a block group B that are used when the pattern type is discriminated to be a vortex type. In this case as well, the part below the fingerprint center is used. .
Here, the reason why the lower part is used is that the upper part from the fingerprint center is similar in any pattern type and is not suitable for pattern discrimination. Therefore, even if a fingerprint image of a large area that is indispensable for pattern classification cannot be collected, it is possible to determine the pattern by using only the part below the center of the fingerprint.

  Therefore, when the system 49 determines the pattern, first, the fingerprint pattern information extraction unit 102 detects the fingerprint center coordinates from the image (center detection step), and divides the image into, for example, 1600 unit blocks (division step). ) Of the divided 1600 unit blocks, the center unit block having the detected fingerprint center coordinates is recognized (recognition step). Further, the fingerprint pattern information extraction unit 102 generates a block group having a plurality of unit blocks from the portion on the opposite side of the fingerprint tip side with respect to the recognized central unit block (block group generation step), and the generated block For each of the plurality of unit blocks constituting the group, the local fingerprint wave flow direction (local direction) is determined (unit block flow direction determination step), and the fingerprint is determined based on the determined fingerprint wave flow direction. Pattern information is determined (fingerprint classification information determination step).

Note that the fingerprint feature data extraction methods shown in FIG. 4A and FIG. 5 are also used in modifications of the first embodiment and the second embodiment described later.
In this way, the pattern type is determined by focusing on the direction in the specific area (core block) below the center of the fingerprint. Therefore, the input image is not completely reproduced and the pattern is accurately reproduced. Even if it cannot be discriminated, there is an advantage that it can be collated with a pattern having a high priority.

  In determining the fingerprint pattern information, a block group to be used for the determination is selected according to the positional relationship between the block group A and the block group B shown in FIG. 4A, and each unit included in the block group is selected. The flow direction of the fingerprint wave of the block is classified into four types: right flow system, left flow system, vertical system and horizontal system. In addition, unit blocks in which no fingerprint image exists are excluded.

Furthermore, the direction types detected from the fingerprint wave state of the unit block (8 pixels × 8 pixels) will be described with reference to FIGS.
6 (a) to 6 (d) are diagrams for explaining the directions according to the first embodiment of the present invention, respectively, showing the directions of the vertical system, the horizontal system, the right flow system, and the left flow system, respectively. ing. These direction patterns (20 patterns) shown in FIGS. 6A to 6D are used to classify the flow direction of fingerprint waves.

Then, the fingerprint pattern of the input image is compared with these direction patterns, and an approximate pattern type is discriminated by paying attention to a fingerprint wave in a specific area below the fingerprint center.
In addition, the collation using this pattern is used also in each modification of 1st Embodiment mentioned later and 2nd Embodiment.

  2 and 3, the registrant fingerprint feature data extraction unit of the registration device 50 and the authentication subject fingerprint feature data extraction unit of the authentication device 51 are shared, and the registrant fingerprint pattern information extraction of the registration device 50 is also performed. The authentication target person fingerprint pattern information extraction unit of the authentication device 51 is shared. This sharing makes it possible to reduce the cost of the system 49.

  Furthermore, the collation priority order definition unit 113 (see FIG. 3) can output the collation priority order for each collation data group classified based on the registrant fingerprint pattern information. (Omitted). Further, the collation priority order definition unit 113 holds data associating the registrant fingerprint pattern information with the collation priority order as collation data.

Therefore, even if the input image is not completely reproduced and the pattern cannot be accurately determined, it can be compared with the matching data having a high matching priority.
Then, collation priority is given to the classified patterns, and collation is sequentially performed until matching data for matching is found. Therefore, even when the patterns do not necessarily match, the identity authentication can be performed reliably, and unnecessary collation can be suppressed to the minimum, so that the labor of calculation can be greatly reduced and efficient collation is possible.

  Then, according to a program provided in the fingerprint information matching unit 114 described below, the matching priority is assigned to each pattern type, such as the right flow type is the first and the vortex type is the second. The collation priority order defined by the above is held in the collation priority order definition unit 113. Details of the method for assigning the collation priority will be described later with reference to FIGS.

As a result, collation can be performed using the collation priority without performing brute force at the time of collation, so that the amount of calculation can be greatly reduced.
Then, the fingerprint information collating unit 114 collates the authentication subject person fingerprint feature data with the registrant fingerprint feature data based on the authentication subject person fingerprint feature data, the authentication subject person fingerprint pattern information, and the collation priority order, and outputs a collation result. Is. The fingerprint information matching unit 114 authenticates the person when the matching rate between the authentication target person's fingerprint feature data and the fingerprint feature data held as matching data exceeds a predetermined matching rate. .

  In addition, when collating fingerprint feature data using the collation priority order definition, the fingerprint information collation unit 114 sequentially collates the collation data in the order of the patterns having the high collation priority. Further, the fingerprint information collating unit 114 can interrupt the personal authentication when collation is performed for a pattern having a low collation priority and a low matching rate.

Further, by changing a predetermined matching rate defined in advance, it becomes possible to set a desired strictness or accuracy, and it is possible to selectively use a precise collation and a collation giving priority to speed as necessary.
The determination unit 115 (see FIG. 3) determines the person / non-person based on the collation result. The determination unit 115 determines the person / non-person in consideration of the fingerprint pattern information of the authentication subject extracted by the fingerprint pattern information extraction unit 102. Further, the determination unit 115 determines the identity / non-identity as a result of matching the matching rate output from the fingerprint information matching unit 114. In other words, the determination unit 115 functions as an individual determination unit that determines whether or not the user is a person based on the matching rate calculated by the fingerprint feature data matching unit 114.

If the determination unit 115 cannot determine the person / non-person until the collation operation at the predetermined collation priority is completed, the fingerprint information collation unit 114 performs the personal authentication at the collation priority lower than the predetermined collation priority. To come to a limit.
As a result, when the collation priority is low and the collation priority of the pattern having a low hit rate is reached, the personal authentication is interrupted and efficient operation can be performed.

The collation priority order definition unit 113 is realized by a memory, a hard disk, and the like. The functions of the fingerprint information collation unit 114 and the determination unit 115 are executed by a CPU (Central Processing Unit) that executes a program stored in the hard disk. It is realized by doing.
In addition, in FIG. 3, those having the same reference numerals as those described above are the same or have the same function, and thus further description is omitted.

  In constructing the system 49, the system operator prepares a desired hardware device and also prepares a software program (program) using a personal computer or a workstation. In addition, the system operator may construct a system in another place using a hard disk (recording medium) of a personal computer or workstation in which the program is installed. In addition, the system operator may distribute a recording medium such as a hard disk (not shown) or a CD-ROM (Compact Disk-Read Only Memory) in which the program is recorded. That is, the system operator may handle a personal computer or a workstation so as to function as the registration device 50 or the authentication device 51.

  For example, the program recorded in the computer-readable recording medium of the present invention using, for example, a hard disk or a CD-ROM authenticates the computer used in the system 49b from the authentication target person fingerprint image collected from the authentication target person. Authentication subject person fingerprint feature data extraction unit 101 that extracts subject person fingerprint feature data, authentication subject person fingerprint pattern information extraction unit 102 that extracts authentication subject person fingerprint pattern information from the authentication subject person fingerprint image, and registrant fingerprint pattern information Collation priority definition unit 113 capable of outputting a collation priority for each collation data group classified based on authentication target person fingerprint feature data, authentication subject person fingerprint pattern information, and collation priority order based on collation priority order Functions as a collation unit 114 that collates the feature data with the registrant fingerprint feature data and outputs a collation result. It is of the order.

The functions of the authentication target person fingerprint feature data extraction unit 101, the authentication target person fingerprint pattern information extraction unit 102, the collation priority order definition unit 113, and the collation unit 114 are the hard disk of the computer (not shown). The program stored in () is executed by the CPU of the computer.
And the method of installing the program on each of a personal computer or a workstation is such as CD-ROM, CD-R (CD-Recordable), CD-RW (CD-Rewritable), floppy disk, etc. of those computers. Use a reader to install on a hard disk, or download and install from another network.

Note that both the computer-readable recording medium on which the program is recorded and the installation method are used in the same manner in each modified example of the first embodiment and the second embodiment to be described later.
Therefore, by using a computer-readable recording medium in which this program is recorded, it is possible to easily install the computer and to efficiently construct a system for performing personal authentication using fingerprint information.

  As a result, the fingerprint feature data extraction unit 101 extracts the authentication target person's fingerprint feature data from the fingerprint collected by the fingerprint information input unit 100, and the fingerprint pattern information extraction unit 102 extracts the authentication target person from the collected fingerprint. Fingerprint pattern information is extracted, and both the data and information are input to the fingerprint information matching unit 114. Further, the collation priority order defining unit 113 outputs the collation priority order for each collation data group classified based on the registrant fingerprint pattern information.

  The fingerprint information matching unit 114 reads data and information from each of the fingerprint feature data storage unit 111 and the fingerprint pattern information storage unit 112, and the data, information, authentication target person fingerprint feature data, and authentication target The person to be authenticated is collated based on the person's fingerprint pattern information and collation priority order, a collation result (match rate) is output, and the determination unit 115 determines the person / non-person based on the collation result.

As described above, when registering the fingerprint feature data, the collation priority order to be collated is defined, the fingerprint feature data and the fingerprint pattern information are held in separate storage devices, and collation is performed while referring to the collation priority order definition.
The user of the system 49 performs registration / authentication as follows. That is, the personal authentication system registration / authentication method using fingerprint information related to the present invention is in the system 49 for performing personal authentication using fingerprint information.

First, registrant fingerprint feature data is extracted from the collected registrant fingerprint image (first extraction step), and the registrant fingerprint feature data extracted in the first extraction step stores data 52 that can hold data. (Registrant fingerprint characteristic data registration step).
Then, registrant fingerprint pattern information is extracted from the registrant fingerprint image (second extraction step), and the registrant fingerprint pattern information extracted in this second extraction step is written in the storage unit 52 (registrant fingerprint classification information). Registration step).

  On the other hand, the collation data is written in the storage unit 52 in association with the registrant fingerprint feature data and the registrant fingerprint pattern information (collation data registration step), and the collation data is extracted from the storage unit 52 (collation Data extraction step), authentication target person fingerprint feature data is extracted from the collected authentication target person fingerprint image (third extraction step).

  Subsequently, authentication target person fingerprint pattern information is extracted from the authentication target person fingerprint image (fourth extraction step), and a verification priority is output for each verification data group classified based on the registrant fingerprint pattern information (verification). (Priority definition step), based on the authentication subject person fingerprint feature data, the authentication subject person fingerprint pattern information, and the collation priority order, it is collated / determined whether or not the authentication subject person is the person (verification / verification). Judgment step).

This personal authentication system registration / authentication method is used in the same way in each of the modifications of the first embodiment and the second embodiment described later.
In this way, since the authentication device 51 performs collation in consideration of collation priority for each pattern in personal authentication using a fingerprint pattern, the pattern of fingerprint information input as an authentication fingerprint and the collation registration data Even if the pattern does not completely match, it can be verified reliably. In addition, since the collation priority is considered in this way, strict pattern determination processing is not necessary.

The personal authentication system registration / authentication method using fingerprint information related to the present invention with the above-described configuration will be described in detail with reference to FIGS.
The registration / authentication method related to the present invention is a method in which a fingerprint inputted within a determinable range can be obtained even if a fingerprint image of a large area necessary for strict determination of a fingerprint pattern is not obtained in personal authentication using a fingerprint. The pattern type is determined from the state of the image as it is, and even if the fingerprint of the registrant's fingerprint does not match the fingerprint of the person to be authenticated due to the input state when the fingerprint is collected, it is given for each pattern type. According to the collation priority order, the personal authentication processing time is shortened as much as possible.

FIG. 7 is a flowchart showing a collation data registration procedure according to the first embodiment of the present invention. First, fingerprint information is input (step A1), fingerprint feature data is extracted from the fingerprint feature data storage unit 111 (step A2), and fingerprint pattern information is extracted from the fingerprint pattern information storage unit 112 (step A3). .
FIG. 8 is a flowchart showing a collation information generation procedure according to the first embodiment of the present invention. When fingerprint information is input (step A10), fingerprint feature data is generated (step A11), and there are five types of patterns of the input fingerprint: right flow system, left flow system, wave system, vortex system, and undecidable. (Step A12). Then, collation data in which the fingerprint feature data and the pattern information are combined is generated (step A13), and the obtained collation data is registered in the database (storage unit) 52 (step A14).

FIG. 9 shows an outline of the verification procedure, and FIGS. 10 and 11 show details of the verification procedure.
FIG. 9 is a flowchart showing an outline of a collation procedure according to the first embodiment of the present invention. When fingerprint information is input (step B1), fingerprint feature data is generated (step B2), and a pattern is determined (step B3). Then, the collation priority order by pattern is referred to, the collation order is determined using the pattern information (step B4), and the fingerprint is collated (step B5).

FIG. 10 is a flowchart showing a verification procedure according to the first embodiment of the present invention, and shows the execution of verification in the authentication device 51 (see FIG. 3).
First, fingerprint information for authentication is collected (step B11), fingerprint feature data is extracted (step B12), and a pattern of an input fingerprint is extracted (step B13), and a right flow type, a left flow type, and a wave type are extracted. , Vortex, and undecidable, and are held as personal authentication elements. In addition, the collation priority order is referred to by the collation priority order definition unit 113 in the order of the highest possibility of determination for each characteristic of the pattern determination process defined in advance (step B14).

Further, the collation priority order may be determined at the time of collation by using the collation priority order judging method shown in FIGS. 25 to 30 described later from the fingerprint image pattern inputted for authentication.
Next, n of the fingerprint data groups for collation belonging to the pattern having the highest collation priority indicated by the collation priority order are stored in a database (DB; fingerprint feature data storage unit 111 and fingerprint pattern) in which the collation feature data is written. From the storage unit 52 including the information storage unit 112) (step B15) or the processing of step B16 to step B20 is repeated. Here, n represents an integer of 0 or more, and represents the number of a plurality of registered fingerprint data that one collation fingerprint data group has. The fingerprint feature data storage unit 111 and the fingerprint pattern information storage unit 112 are referred to in step B15 and step B22 described later, respectively.

  For example, in a for loop statement, the initial value is set to 0 (step B16). If the number of loops is less than n in step B17, the Yes route is passed and the i-th (i is 0 or more) in step B18. n) or less matching feature data is compared with the authentication fingerprint feature data. Then, the database 52 is referred to as to whether or not the feature points match (step B19). When matching fingerprint data is found, it passes through the Yes route and is authenticated (step B23). As a result, “authentication” is output.

In step B19, while there is no collation fingerprint data matching the fingerprint collation data group of the pattern having the highest collation priority, the value of i is counted up through the No route (step B20). The processing from step B17 is repeated.
When the loop process is completed in step B17, the route No is passed, and in step B21, it is determined whether or not the next collation priority is present. Here, if there is a pattern having the next collation priority, it passes through the Yes route, and in step B22, the fingerprint feature data group of the next pattern having the higher collation priority is stored in the fingerprint feature data storage unit 111 and the fingerprint pattern information storage. After obtaining n items (n is the same as n described above) from the part 112 (see circled numbers 1 and 2 in the figure), the processing from step B16 is repeated again. In Step B21, when there is no pattern having the next collation priority, the route is No and the person is not authenticated (Step B24).

  Further, in each of step B15 and step B22, data that matches the pattern indicated by the priority order is collectively read into the memory 52 from the database 52 in which the verification data is stored, and the authentication subject person's fingerprint pattern and its reading are read. The data is verified. For example, if there are 100 pieces of collation data matching the pattern indicated by the priority order being processed, these 100 cases are collectively read into the memory and sequentially collated in the memory.

In this way, the fingerprint feature data group for matching that matches the pattern is acquired from the database 52 in which the fingerprint feature data for matching is written, and the database 52 is referred to each time at the time of authentication, and compared with the fingerprint feature data for authentication. The
In addition, the matching registration data for the pattern with high matching priority and the matching registration data for the pattern with low matching priority both have data matching the authentication data, and the matching registration with low matching priority. If the data matching rate exceeds the matching rate of the matching registration data having a high collation priority, whether or not the user is the person is authenticated based on the collation priority level and the matching rate at this time. That is, one of them is determined as the principal.

  For example, the matching rate condition is 70%, the matching rate of the registration verification data having the first priority pattern is 75%, and the matching rate of the registration verification data having the fifth priority pattern matching rate. Is 75% or more, it is very unlikely that the verification registration data will be assigned to the pattern with the fifth priority. Therefore, the registration verification data having the pattern with the first priority is authenticated with the person. be able to.

In this way, personal authentication can be performed based on the high collation priority and the matching rate.
FIG. 11 is a flowchart showing another verification procedure according to the first embodiment of the present invention. The collation procedure shown in FIG. 11 refers to collation data matching the pattern indicated by the priority order from the database 52 each time collation is performed.
First, the processing from step B11 to step B14 is the same as that shown in FIG. Next, in step B15a, the number of feature data of the pattern having the highest collation priority is obtained from the fingerprint feature data storage unit 111 and the fingerprint pattern information storage unit 112. Then, it is determined whether or not n is 0 (step B27). When n is 0, it is determined whether or not there is a next priority in step B28 through the Yes route. Passing the Yes route, in step B29, n number of feature data of the pattern having the next collation priority order is acquired from the database 52, and the process returns to step B27.
If n is not 0 in step B27, the route goes through No route and proceeds to step B30. If there is no next priority in step B28, the route goes through No route and the process of step B38 described later is performed.

  In step B30, the initial value of the loop is set to 1. In step B31, the i-th data that matches the priority pattern being processed is stored in the database (the storage unit including the fingerprint feature data storage unit 111 and the fingerprint pattern information storage unit 112). 52). N of the fingerprint data groups for matching belonging to the pattern having the highest matching priority are acquired from the database (DB) 52.

  Subsequently, in step B32, the i-th matching feature data and the authentication fingerprint feature data are verified, and in step B33, it is checked whether or not the feature points match. The user is authenticated (step B39). On the other hand, if they do not match in step B33, the route passes through the No route, and in step B34, it is checked whether or not collation has been completed for all n cases. In B35, after being incremented, the processing from Step B31 is repeated. Here, n is the same as described above.

  In step B34, if all n cases have been verified, the route passes through the No route, and in step B36, it is checked whether there is a next priority level. In step B38, the process is terminated without authenticating the user. On the other hand, if there is the next priority in step B36, the Yes route is passed, and in step B37, n fingerprint feature data numbers of the next matching priority pattern are obtained from the database 52, and n is 0 again. It is determined whether or not (step B40). When n is 0, the route proceeds to step B36 through the Yes route. On the other hand, if n is not 0, the route is No, i is set to 1 (step B41), and the process returns to step B31.

Therefore, data to be collated is acquired from the database 52 every time collation is performed. Any of the procedures shown in FIGS. 10 and 11 can be changed as appropriate by changing the processing speed, accuracy, and the like.
10, “feature data group” is acquired from the database 52, and “feature data count” is acquired from the database 52 in steps B15a and B37 illustrated in FIG.

Hereinafter, with reference to the drawings, an operation for classifying each fingerprint type will be described.
First, a pattern determination example in which the fingerprint image example shown in FIG. 12 is determined to be the right flow type will be described with reference to FIGS. 13 and 14. FIG. 12 is a diagram showing an example of a fingerprint image determined to be the right flow type. FIGS. 13 and 14 are both procedures for determining the right flow type (steps C1 to C14) according to the first embodiment of the present invention. It is a flowchart.

In step C1 shown in FIG. 13, fingerprint information (fingerprint image) is input, and the fingerprint center coordinates are detected. Next, in step C2, the entire image is divided into a plurality of blocks of 8 pixels × 8 pixels, and in step C3, the block where the detected fingerprint center coordinate is located is recognized as the center block.
Further, in step C4, the right-flow pattern determination is performed by using a block group (hereinafter referred to as processing) in which blocks from the center block to the bottom of the image are 8 blocks wide (all blocks including the lower block). A block group in the right half of the processing block group (hereinafter referred to as block group A) and a block group in the left half of the processing block group (hereinafter referred to as block group B). Is done. Further, the positional relationship between the block group A and the block group B at that time is as shown in FIG.

Subsequently, in step C5, the fingerprint wave flow directions of the blocks constituting the block group recognized as the target of the determination process are classified into four types: right flow system, left flow system, vertical system, and horizontal system ( 6 (a) to 6 (d)), and blocks having no fingerprint image are excluded from processing targets.
In step C6, if many of the blocks belonging to the block group A (for example, 80%) are in the right flow system, the vertical block in the block group B is determined to be the right flow system, passes the Yes route, and in step C7. The vertical block of the block group B is classified as the right flow system. If it is not determined in step C6 that the flow is rightward, the route passes No route, and the process proceeds to step C8 shown in FIG.

  In Step C8, if many of the blocks belonging to the block group B (for example, 80%) are the left flow system, the Yes route is passed, and in Step C9, the vertical blocks in the block group A are classified as the left flow system. The If it is not determined in step C8 that the flow is left-handed, the route passes No route and proceeds to step C10. Although 80% is defined as the threshold value, it is not limited to this value, and can be set to a desired value when the strictness of determination is changed.

Therefore, even if the block is determined to be vertical, it is classified as right flow or left flow as much as possible. Therefore, even if the determined direction is an oblique finger input angle in the fingerprint collection situation, In the case of the right flow type and the left flow type, it is avoided that the vertical type is erroneously determined.
At this time, since the direction of each block located in the processing block group is determined as a determination factor, in step C10, the right flow ratio of the block group A and the right flow ratio of the block group B are added to obtain the total right flow value. Is recognized as a maximum of 200%. Similarly, the left flow ratio of the block group A and the left flow ratio of the block group B are added, and the total left flow value is recognized as the maximum (MAX) 200%.

Then, from the recognized total right flow value and total left flow value, it is determined in step C11 to step C14 whether or not the input fingerprint is a right flow type.
First, obviously, a right flow ratio is determined to be a right flow type. For example, if the value obtained by subtracting the total left flow value from the total right flow value is 100% or more (step C11), it passes through the Yes route and is determined to be the right flow type (step C14). Since the right-flow type is obviously selected here, the threshold is set as high as possible.

Next, it determines about what was not determined with the right flow in the determination. That is, if the No route in Step C11 is passed, and if the total right flow value is 40% or more and the total left flow value is 15% or less in Step C12, the Yes route is passed and the right flow type is determined. (Step C14).
Furthermore, if the determination condition is not satisfied in step C12, the route passes through the NO route. In step C13, it is determined that the flow is not the right flow type, and the process ends.

In the determinations shown in FIGS. 13 and 14, the pattern cannot be determined when the fingerprint state such as finger roughening or the minimum fingerprint image necessary for the determination cannot be obtained. In this case, the pattern determination is impossible and a separate group is set, and the collation priority is assigned.
In this way, it is determined from the ratio between the total right flow value and the total left flow value whether or not it is right flow. For example, depending on the fingerprint collection situation, it is actually a right flow type pattern and the total right flow Even if the value becomes smaller, it is determined as the right flow type.

In addition, calculation efficiency is improved by eliminating the need for strict determination.
Next, a pattern determination example in which the fingerprint image example shown in FIG. 15 is determined to be the left flow type will be described with reference to FIGS. 16 and 17. FIG. 15 is a diagram showing an example of a fingerprint image determined to be the left flow type. FIGS. 16 and 17 are both procedures for determining the left flow type according to the first embodiment of the present invention (steps D1 to D14). It is a flowchart.

  The procedure from the input of fingerprint information to the calculation of the total right flow value and the total left flow value is the same as the above-described right flow determination. That is, fingerprint information is input, fingerprint center coordinates are detected (step D1), the entire image is divided into a plurality of blocks of 8 pixels × 8 pixels (step D2), and the block where the detected fingerprint center coordinates are located Is recognized as a central block (step D3).

Further, the determination is made using the right half block group A of the processing block group in which the blocks from the center block to the left and right of the image are 8 blocks wide, and the left half block group B of the processing block group. (Step D4).
Next, the flow direction of the fingerprint wave of each block constituting the block group recognized as the determination processing target is classified into four types of a right flow system, a left flow system, a vertical system, and a horizontal system. Blocks that do not exist are excluded from processing targets (step D5).

  If many of the blocks belonging to the block group A (for example, 80%) are the right flow system, the vertical block of the block group B is determined to be the right flow system (step D6). The vertical block is classified as a right flow system (step D7). If it is not determined in step D6 that the flow is rightward, the route proceeds through step No8 to step D8 shown in FIG.

In this step D8, if many of the blocks belonging to the block group B (for example, 80%) are the left flow system, the Yes route is passed, and in step D9, the vertical blocks in the block group A are classified as the left flow system. If the left flow system is not determined in step D8, the route passes No route and proceeds to step D10.
Further, in step D10, the right flow ratio of the block group A (right flow system ratio) and the right flow ratio of the block group B are added, and the total right flow value is recognized as a maximum of 200%. Similarly, the left flow ratio of the block group A and the left flow ratio of the block group B are added, and the total left flow value is recognized as a maximum of 200%.

  Then, from the recognized total right flow value and total left flow value, it is determined in step D11 to step D14 whether or not the input fingerprint is a left flow type. First, in Step D11, if the value obtained by subtracting the total right flow value from the total left flow value through the Yes route is 100% or more, it is determined as the left flow type (Step D14). That is, obviously, a left flow ratio (a ratio of the left flow system) is determined to be a left flow type. Here, the threshold value is set to a large value in order to select a left-flow type.

  Next, if the value obtained by the subtraction in step D11 is smaller than 100%, the route that has passed through the No route is determined, and in step D12, it is determined that the left flow type is not determined. If the total left flow value is 40% or more and the total right flow value is 15% or less, it passes through the Yes route and is determined to be the left flow type in step D14. Otherwise, it passes through the No route. In Step D13, the left flow type is not determined.

In the determination shown in FIGS. 16 and 17, the pattern cannot be determined when the fingerprint state such as finger roughening or the minimum fingerprint image necessary for the determination cannot be obtained. In this case, the pattern determination is impossible and a separate group is set, and the collation priority is assigned.
In this way, it is determined from the ratio of the total right flow value and the total left flow value whether or not it is left flow. For example, depending on the fingerprint collection situation, the left flow value is actually left and the total left flow value is small. Even if it becomes, it is determined that it is a left flow type. In addition, calculation efficiency is improved by eliminating the need for strict determination.

  Next, a pattern determination example in which the fingerprint image example shown in FIGS. 18A and 18B is determined to be corrugated will be described with reference to FIG. FIGS. 18A and 18B are diagrams showing an example of a fingerprint image determined to be corrugated, and FIG. 19 is a flowchart of a procedure for determining corrugation according to the first embodiment of the present invention. The procedure for classifying the directions of the block group A and the block group B into four types of right flow system, left flow system, vertical system, and horizontal system from the input of fingerprint information is the same process as the right flow determination. .

  That is, fingerprint information is input, fingerprint center coordinates are detected (step E1), the entire image is divided into a plurality of unit blocks each consisting of 8 pixels × 8 pixels (step E2), and the unit where the detected fingerprint center coordinates are located The block is recognized as a central block (central unit block) (step E3). Further, a block group A in the right half of the processing block group in which unit blocks from the center block to the left and right are 8 blocks (8 unit blocks) in width and a block group B in the left half of the processing block group. It is determined by using (step E4).

Subsequently, the direction of the fingerprint wave flow of each block (unit block) constituting the block group recognized as a determination processing target is classified into four types: right flow system, left flow system, vertical system, and horizontal system. The unit block having no fingerprint image is excluded from the processing target (step E5).
In step E6, if many (eg, 70%) of the unit blocks belonging to the block group A are the left flow system, and many (eg, 70%) of the unit blocks belonging to the block group B are the right flow system. , The Yes route is passed and it is determined to be corrugated (step E8). In step E6, if not, the route is No, and in step E7, the waveform is not determined.

Here, the determination threshold is 70%. This value is defined in consideration of the fact that the direction of the unit block located vertically below the central block is often a horizontal system. Further, when changing the strictness of determination, it can be set to a desired value.
In the determination shown in FIG. 19, the pattern cannot be determined when the fingerprint state such as rough finger or the minimum fingerprint image necessary for the determination cannot be obtained. In this case, the pattern determination is impossible and a separate group is set, and the collation priority is assigned.

As described above, in the wave type determination, if the right flow rate of the block group A and the left flow rate of the block group B are substantially equal, the wave system is determined, and if the wave system is determined, the determination result is obtained. Is determined to be corrugated, and if the determination condition is not satisfied, the processing is terminated as not corrugated.
Next, a pattern determination example in which the fingerprint image example shown in FIGS. 20A and 20B is determined to be a vortex type will be described with reference to FIG. FIGS. 20A and 20B are diagrams showing an example of a fingerprint image determined as a vortex, and FIG. 21 is a flowchart of a procedure for determining a vortex according to the first embodiment of the present invention. The process from the input of fingerprint information to the detection of the center block is the same process as the right flow type determination.

That is, fingerprint information is input, fingerprint center coordinates are detected (step F1), the entire image is divided into a plurality of unit blocks of 8 pixels × 8 pixels (step F2), and the unit where the detected fingerprint center coordinates are located The block is recognized as a central block (step F3).
Then, in step F4, the block group used for the vortex pattern determination is moved 8 blocks (unit blocks) below the center block, and 5 unit blocks moved to the left and right from that position. Eight neighboring unit blocks located in the vicinity thereof are recognized as processing blocks. The recognized right processing block group is referred to as a block group A, and the left processing block group is referred to as a block group B (see FIG. 5).

Further, in step F5, the direction of fingerprint wave flow of each unit block constituting the block group recognized as a determination processing target is classified into four types: right flow system, left flow system, vertical system, and horizontal system. This classification is the same process as the right flow type determination.
That is, in step F6, if many of the unit blocks belonging to the block group A (for example 70%) are in the left flow system and many of the unit blocks belonging to the block group B (for example 70%) are in the right flow system. , The Yes route is passed, and it is determined as a vortex type (step F8). In Step F6, otherwise, the No route is passed, and in Step F7, the vortex type is not determined.

Here, the determination threshold is 70%. Note that the left flow rate of the block group A and the right flow rate of the block group B are not necessarily 100% depending on individual fingerprints. Therefore, the threshold value is defined in consideration of them, and can be set to a desired value when the strictness of determination is changed.
In the determination shown in FIG. 21, the pattern cannot be determined when the fingerprint state such as rough finger or the minimum fingerprint image necessary for the determination cannot be obtained. In this case, the pattern determination is impossible and a separate group is set, and the collation priority is assigned.

Thus, in the vortex type determination, if most of the direction of the block group A is the left flow and most of the direction of the block group B is the right flow, it can be determined as the vortex type.
In this way, if the vortex system can be determined, the determination result is a vortex type, and if the determination condition is not satisfied, the process ends as not being a vortex type.

Next, an example of collation priority definition will be described with reference to FIGS.
FIG. 22 is a flowchart showing a definition example of the right flow type collation priority order according to the first embodiment of the present invention, and the collation priority order definition when the right flow type pattern determination examples shown in FIGS. 13 and 14 are used. An example of right flow type collation priority definition in the unit 113 is shown. In the following description, in the drawings, the first collation priority to the fifth collation priority may be displayed as priority 1 to priority 5, respectively.

If it is determined in step G1 shown in FIG. 22 that the fingerprint data for authentication is the right flow type (see FIGS. 13 and 14), the pattern type having the first matching priority among the registered data for matching is the same type. The right-hand flow type. Further, the second collation priority is a vortex type (step G2).
The reason for this is that in the determination of the vortex pattern shown in FIGS. 20 (a) and 20 (b), when the block group used for determination cannot be determined as the vortex pattern, This is because there is a high probability that the vortex pattern will be the right-flowing pattern.

Further, the third collation priority is a waveform (step G3). This is because in the right flow pattern determination, the flow of the right flow pattern that is input obliquely cannot be accurately recognized, and there is a possibility that the waveform is determined as a wave pattern in the waveform determination shown in FIG.
Next, fingerprints that are determined to be undecidable (non-determinable type) are defined in the fourth collation priority order (step G4), and the fingerprint characteristic data for collation of the left flow pattern is defined in the fifth collation priority order. (Step G5). This is because the characteristics of the left flow type determination shown in FIGS. 16 and 17 are very unlikely to determine the right flow type fingerprint as the left flow type.

FIG. 23 is a flowchart showing a definition example of the left flow type collation priority order according to the first embodiment of the present invention, and the collation priority order definition when the left flow type pattern determination examples shown in FIGS. 16 and 17 are used. An example of left flow matching priority definition in the unit 113 is shown.
In step G11 in FIG. 23, if the authentication fingerprint pattern is the left flow pattern at the time of the left flow pattern determination shown in FIGS. 16 and 17, the first verification priority in the verification registration data. The pattern pattern to be collated is the same left flow type. Further, the second collation priority is a vortex type (step G12). The reason for this is that, in the vortex type determination shown in FIG. 21, when the block group used for the determination cannot be determined as a vortex type, in the left flow type pattern determination, the vortex pattern that has flowed to the left is the left flow type. This is because the probability of doing this is high.

Further, the third collation priority is a waveform (step G13). This is because, in the left flow type determination process, the flow of the left flow type fingerprint pattern input obliquely cannot be accurately recognized, and may be wave-shaped in the determination.
Next, the fingerprint that is determined to be unidentifiable is defined as the fourth collation priority (step G14). Further, the fingerprint characteristic data for collation of the right flow pattern is defined in the fifth collation priority (step G15). This is because the right flow type determination process has a very low possibility of determining the left flow fingerprint as the right flow type.

FIG. 24 is a flowchart showing a definition example of the waveform collation priority order according to the first embodiment of the present invention. When the waveform pattern determination example shown in FIG. 19 is used, the waveform pattern in the collation priority order definition unit 113 is used. An example of collation priority definition is shown.
In the waveform pattern determination, if the pattern of the authentication fingerprint is determined to be a waveform pattern, the pattern pattern to be compared with the highest priority among the verification registration data is the same waveform pattern (step G21).

  Here, since the wave pattern is determined to be the right flow type or the left flow type depending on the fingerprint collecting situation, the next collation priority is the right flow type or the left flow type. Therefore, in step G22, the collation priority is determined depending on which ratio of the left flow type and the right flow type is greater for the fingerprint pattern input for authentication. When it is determined that the fingerprint pattern for authentication is a wave pattern, or when there are many left flow types in the processing block, the Yes flow from Step G22 is followed, and the left flow type is registered for verification as the second verification priority. Reference is made from the fingerprint data (step G23), and a right-flow type fingerprint is referenced from the registered fingerprint data for collation as the third collation priority (step G24).

In Step G22, when there are many right flow types in the processing block, the No flow is passed through the No route, and the right flow type fingerprint is referred to from the registered fingerprint data for verification as the second verification priority (Step G27). As the third collation priority, the left-flow type fingerprint is referred to from the collation registered fingerprint data (step G28).
Next, in step G25, the fingerprint that cannot be determined is defined as the fourth collation priority, and in step G26, the vortex-type fingerprint is defined as the fifth collation priority. This is because the characteristic of the vortex pattern determination process is very unlikely to determine a wave fingerprint as a vortex.

FIG. 25 is a flowchart showing a definition example of the vortex matching priority order according to the first embodiment of the present invention. When the vortex pattern determination example shown in FIG. An example of collation priority definition is shown.
In the vortex pattern determination, when it is determined that the pattern of the fingerprint for authentication is a vortex pattern, the pattern pattern that is collated with the highest priority among the collation registration data is the same vortex pattern (step G31).

  Here, since the vortex pattern may be determined to be the right flow type or the left flow type depending on the flow of the center of the vortex and the fingerprint collecting situation, the right flow type or the left flow type is used as the next collation priority. It is said. That is, in step G32, the collation priority is determined depending on which ratio of the left flow type or the right flow type of the fingerprint input for authentication is larger, and there are many left flow types in the direction near the central block. In this case, the left flow fingerprint is referred to from the registered fingerprint data for collation as the second collation priority order through the Yes route (step G33), and the right flow fingerprint as the third collation priority order is registered fingerprint data for collation. (Step G34).

  In Step G32, if there are many right flow types in the direction near the center block, the right flow type fingerprint is referred to from the registered fingerprint data for verification as the second verification priority through the No route (Step G37). Further, as the third collation priority, the left-flow type fingerprint is referred from the collation registered fingerprint data (step G38).

Next, the fingerprint that cannot be determined is defined as the fourth collation priority (step G35). Further, the characteristic of the wave pattern determination is that the possibility that the vortex fingerprint is determined to be a wave pattern is very low. Therefore, the fingerprint characteristic data for wave pattern verification is set to the fifth matching priority (step G36). ).
In step G32, it is possible to determine the collation priority order depending on which of the left flow type and the right flow type the direction near the central block of the fingerprint inputted for authentication is larger.

  In this way, in the personal authentication system 49 using fingerprints using fingerprint patterns, the fingerprint feature data is extracted from the collected fingerprint image at the time of registration of the verification data, and is classified into a pattern type different from the actual pattern. The pattern is determined while allowing the pattern to be verified, and at the time of verification, verification is performed based on the verification priority defined for each pattern based on the pattern determined from the authentication input fingerprint.

In this way, the user can quickly authenticate himself / herself without redundant verification.
In addition, since the pattern determination method of the present embodiment uses only a specific area in the lower part from the fingerprint center, even if the fingerprint image has a small area, pattern determination is possible although it is not strict. Even if this pattern determination is determined to be a different type from the actual pattern, the collation priority is defined in the order of the highest possibility of determination for each pattern, so it must be classified into a type different from the actual pattern. Is acceptable.

Therefore, compared to the personal authentication method using fingerprints that has already been proposed, the strictness of pattern determination can be relaxed, so that the processing speed of pattern determination can be improved and verification can be performed. By collating according to priority, it is possible to quickly authenticate the person as compared with mere brute force collation.
In addition, when there are a large number of collation persons, the larger the number of persons, the shorter the total collation time can be achieved by using the determination method related to the present invention.

(A1) Description of First Modified Example of First Embodiment of the Invention In the first modified example, verification using the modified device of the authentication device 51 will be described. Similarly to the personal authentication system 49 shown in FIG. 1, the personal authentication system 49a using the fingerprint information according to the first modification also uses the fingerprint information to authenticate the individual. The storage unit 52, the registration device 50 and an authentication device 51a.

  FIG. 26 is a block diagram of an authentication device 51a according to a first modification of the first embodiment of the present invention. The authentication device 51a shown in FIG. 26 is used for collation / determination using the collation data written in the storage unit 52 as to whether or not the person to be authenticated is the person himself / herself, in order to collect registrant fingerprint information. The collation data is obtained from the storage unit 52 based on the information input using the fingerprint information input unit 100 of the above and collated. The authentication device 51a includes a second fingerprint information matching unit 114a. The second fingerprint information matching unit 114a includes a matching priority reference unit 201, a matching data selection unit 221, and a fingerprint information matching unit 114. It is composed.

  Here, the collation priority reference unit 201 reads (refers to) the collation priority stored in the collation priority definition unit 113, and the collation data selection unit 221 includes the collation priority reference unit 201. Are connected to the fingerprint feature data storage unit 111 and the fingerprint pattern information storage unit 112 of the storage unit 52, and data for verification is selected.

In addition, what is shown in FIG. 26 and has the same code | symbol as what was mentioned above has the same thing as what was mentioned above, or has the same function, The duplicate description is abbreviate | omitted.
As a result, in response to an instruction from a control unit (not shown), a fingerprint image is collected by the fingerprint information input unit 100, and the fingerprint feature data is extracted by the fingerprint feature data extraction unit 101. Fingerprint feature data to be used is generated, and this fingerprint feature data is input to the fingerprint information matching unit 114.

  On the other hand, the fingerprint image output from the fingerprint information input unit 100 is extracted and classified by the fingerprint pattern information extraction unit 102, and the classified data is input to the collation priority order reference unit 201. In the collation priority order reference unit 201, the collation priority order of the fingerprint pattern for authentication is referred to from the collation priority order definition unit 113, and in the collation data selection unit 221, the fingerprint feature data storage unit 111 and the fingerprint pattern storage unit 112 is referred to, and collation data is acquired.

  Then, the fingerprint information collation unit 114 collates the obtained authentication fingerprint feature data and the collation fingerprint feature data for each pattern, and the determination unit 115 determines whether the person is the person based on the matching rate (collation result). It is determined whether or not. Here, when there is no data matching the fingerprint fingerprint matching data of the pattern indicated by the collation priority, it is determined that the person is not the person, and the determination unit 115 requests the fingerprint information verification unit 114 to perform verification again.

  Then, when the fingerprint information collation unit 114 is requested to collate with the next fingerprint feature data from the determination unit 115, the fingerprint feature collation feature data having the next collation priority order is converted into the collation fingerprint feature data storage unit. The data is read from 111 and collation with the pattern according to collation priority is repeated. At this time, collation fingerprint data for the pattern that matches the pattern determined by the determination unit 115 is read from the database (storage unit) 52, and the read data is compared with the input data. Is done.

If no matching data is found in this determination, the registered fingerprint data for collation indicated by the next collation priority is referred to, for example, until the end of the third collation priority or until a certain collation priority. Compared with fingerprint data for authentication.
In this way, regardless of the state of the input fingerprint image, strict pattern determination processing becomes unnecessary by pattern determination in the determinable range in the image. Also, at the time of collation, by collating using the collation priority for each fingerprint pattern, the process of specifying an individual such as ID input becomes unnecessary. Therefore, the processing speed of personal authentication using a one-to-many fingerprint that requires individual identification is improved.

(A2) Description of Second Modification of First Embodiment of the Present Invention In the first embodiment and the first modification of the first embodiment, the collation priority order is given by the collation priority order defining unit 113, and collation At that time, the collation priority order was read out and referred to. In the method of the second modified example, when creating collation data, a parameter capable of determining collation priority is added to the collation data, and the person is authenticated using the collation data at the time of collation. ing.

Similarly to the personal authentication system 49 shown in FIG. 1, the personal authentication system 49b using the fingerprint information according to the second modification performs personal authentication using the fingerprint information. 50a and an authentication device 51b.
FIG. 27 is a block diagram of a registration device 50a according to a second modification of the first embodiment of the present invention. 27. The collation priority order determination processing unit 310 of the registration device 50a shown in FIG. 27 determines and outputs the collation priority order according to the registrant fingerprint pattern information, and the collation data generation unit 121 uses the collation data. Are registered in the storage unit 52 in association with the registrant fingerprint feature data, the registrant fingerprint pattern information, and the collation priority.

Thus, a registrant fingerprint image is collected by the fingerprint information input unit 100 according to an instruction of a control unit (not shown), and a registrant fingerprint feature is extracted from the collected registrant fingerprint image by the fingerprint feature data extraction unit 101. Data is extracted and fingerprint feature data used for matching is generated.
In addition, the fingerprint pattern information extraction unit 102 extracts registrant fingerprint pattern information from the collected registrant fingerprint image, and the matching priority determination processing unit 310 responds to the extracted registrant fingerprint pattern information. The collation priority order is determined, and the collation data generation unit 121 associates the registrant fingerprint feature data, the registrant fingerprint pattern information with the collation priority order in the collation data generation unit 121, and the fingerprint feature data storage unit 111 It is written in the fingerprint-like information storage unit 112.

In addition, what is shown in FIG. 27 and has the same reference numeral as that described above has the same or similar function as that described above, and redundant description thereof will be omitted.
Note that the two types of fingerprint feature data storage unit 111 for comparison and the fingerprint pattern information storage unit 112 shown in FIG. 27 can be used to manage information as a single storage unit 52 in addition to holding information individually. Therefore, the fact that these two types of information are one-to-one linked information can be used.

  In constructing the system 49b, the system operator prepares a desired hardware device and also prepares a software program and distributes a recording medium such as a hard disk or a CD-ROM in which the program is recorded. Therefore, the system operator causes a personal computer or a workstation to function as the registration device 50a.

  The program recorded on the computer-readable recording medium of the present invention is a registrant fingerprint feature data for extracting registrant fingerprint feature data from a registrant fingerprint image collected from a registrant. The extraction unit 101 extracts the registrant fingerprint pattern information from the registrant fingerprint image, and extracts the registrant fingerprint pattern information. The collation priority is determined according to the registrant fingerprint pattern information. A collation priority order determination processing unit 310 to output, and a collation data generation unit that writes the collation data to the storage unit 52 that can hold the data by associating the registrant fingerprint feature data, the registrant fingerprint pattern information, and the collation priority. It is for functioning as 121.

The functions of the registrant fingerprint feature data extraction unit 101, fingerprint pattern information extraction unit 102, and collation data generation unit 121 are respectively the programs stored in the hard disk of the computer, and the CPU of the computer. It is realized by executing.
Therefore, by using a computer-readable recording medium in which this program is recorded, it is possible to easily install the computer and to efficiently construct a system for performing personal authentication using fingerprint information.

Next, FIG. 28 is a block diagram of an authentication device 51b according to a second modification of the first embodiment of the present invention. The authentication device 51b shown in FIG. 28 does not have the collation priority definition unit 113 (see FIG. 3), and the fingerprint information collation unit 114 uses the collation priority stored in the storage unit 52 based on the collation data. Direct reference.
Further, what is shown in FIG. 28 and has the same reference numeral as that described above has the same or similar function as that described above, and redundant description thereof will be omitted.

  As a result, in the authentication device 51b, the authentication target person fingerprint feature data extraction unit 101 extracts the authentication target person fingerprint from the authentication target person fingerprint image collected by the authentication fingerprint information input unit for collecting the authentication target person fingerprint information. The feature data is extracted, and the fingerprint pattern information extraction unit 102 extracts the authentication target person fingerprint pattern information from the authentication target person fingerprint image collected by the fingerprint information input unit 100.

Further, the fingerprint information collation unit 114 collates the authentication target person based on the collation priority order included in the collation data held in the storage unit 52, and outputs a collation result.
With reference to FIGS. 29 and 30, an example in which the above-described configuration is used to register and authenticate the collation priority order of patterns during registration in this modification will be described in detail. FIG. 29 is a flowchart showing a procedure for registering collation data according to the second modification of the first embodiment of the present invention.

  First, a registered fingerprint for collation is input (step H1), a pattern is determined (step H2), a collation priority order of the determined pattern is determined (step H3), and fingerprint feature data is extracted (step S3). H4), the information is registered in the database 52 using the three types of information: pattern type, collation priority, and fingerprint feature data (step H5).

  FIG. 30 is a flowchart showing an authentication procedure according to the second modification of the first embodiment of the present invention. When the authentication fingerprint is input (step H11), the pattern is determined (step H12), and the collation priority K (K represents a natural number) is set to 1 (step H13). Then, the same pattern as the determined pattern is collated with the registered data for collation having the Kth priority (step H14), and it is checked whether there is matching data (step H15).

  Here, if there is matching data, the Yes route is passed and the authentication is successful and the process ends (step H19), but if there is no matching data, the No route is passed and K is incremented (step H16). It is checked whether or not the value of K exceeds the maximum priority (step H17). If it exceeds, the Yes route is passed and authentication fails (step H18). As described above, the processing from step H14 is repeated.

Accordingly, when inputting a registered fingerprint for verification, in addition to the pattern information and fingerprint feature data, verification priority information is given to the pattern in advance to generate verification data, and the fingerprint pattern information extraction unit 102 The pattern type is extracted, and the collation priority is referenced from the collation data.
The collation priority definition method described with reference to FIGS. 22 to 26 can also be executed in the same manner.

As described above, since the fingerprint information collation unit 114 does not need to calculate the collation priority when the authentication data is input, the calculation amount can be greatly reduced.
In addition, since the collation priority is not referred to from the collation priority definition unit 113 during collation and the collation priority is obtained from the fingerprint pattern input for authentication, the efficiency of personal authentication is improved.

Therefore, in this way, it is not required that the registered fingerprint pattern for verification and the fingerprint pattern for authentication match, and when these patterns do not match, they are collated using the collation priority for each pattern and match. Personal authentication using fingerprints is possible in descending order of probability, and rapid personal authentication is possible.
(B) Description of Second Embodiment of the Invention Both the registration device 50 (50a) and the authentication device 51 (51a, 51b) described above collect fingerprints as biometric information. This authentication information is not limited to fingerprints. For example, palm print, finger print, palm print, voice (voice print), retina, iris, face image, dynamic signature, blood vessel pattern, keystroke, etc. It can also be applied to a system that collects and authenticates biometrics.

  FIG. 31 is a block diagram of a registration device in a personal authentication system using biometric information according to the second embodiment of the present invention. The registration device 60 shown in FIG. 31 obtains collation data from information input using a registration biometric information input unit for collecting biometric information of a registrant, and writes it in the storage unit 61. Biometric information input unit (registration biometric information input unit) 60a, biometric information feature data extraction unit (registrant biometric information feature data extraction unit) 60b, biometric classification information extraction unit (registrant bioclassification information extraction unit) 60c, And a collation data generation unit 60d.

Here, the function of the biometric information input unit 60a is realized by a scanner capable of collecting a shape such as a palm print or a measurement device (not shown).
The biometric information feature data extraction unit 60b extracts registrant biometric information feature data from the biometric information image of the registrant collected by the biometric information input unit 60a. The biometric classification information extraction unit 60c is a biometric information input unit. The registrant biometric classification information is extracted from the biometric information image of the registrant collected by 60a, and the collation data generation unit 60d uses the registrant biometric information feature data, the registrant biometric classification information, and the collation data. Are stored in the storage unit 61 in association with each other.

  Further, the storage unit 61 shown in FIG. 31 includes a biometric information feature data storage unit (registrant biometric information feature data storage unit) 61a that can hold registrant biometric information feature data and a biometric information that can hold registrant biometric classification information. An information storage unit (registrant biometric information storage unit) 61b is provided, and its function is realized by, for example, a memory (not shown). This storage unit 61 is also connected to the authentication device 70 as described later. It is designed to be shared.

  FIG. 32 is a block diagram of an authentication apparatus in a personal authentication system using biometric information according to the second embodiment of the present invention. The authentication device 70 shown in FIG. 32 is used for collation / determination using the collation data written in the storage unit 61 as to whether or not the person to be authenticated is the person himself / herself. In addition, a biometric information feature data extraction unit (authentication subject biometric information feature data extraction unit) 60b, a biometric classification information extraction unit (authentication subject biometric classification information extraction unit) 60c, a collation priority order definition unit 70a, and biometric information A collation unit (collation unit) 70b and a determination unit 70c are provided. In FIG. 32, those having the same reference numerals as those described above have the same or similar functions, and therefore redundant description is omitted.

  Here, the biometric information feature data extraction unit 60b extracts the verification subject biometric information feature data from the biometric information image of the authentication subject collected by the biometric information input unit 60a for collecting the biometric information of the authentication subject. The biometric classification information extraction unit 60c extracts the authentication subject biometric classification information from the biometric information image of the authentication subject collected by the biometric information input unit 60a. And shared.

That is, the registrant biometric information feature data extraction unit of the registration device 60 and the authentication subject biometric information feature data extraction unit of the authentication device 70 are shared, and the registrant biometric classification information extraction unit and the authentication device 70 of the registration device 60 are used. The authentication subject biometric classification information extraction unit is commonly used.
Moreover, the collation priority order definition part 70a can output collation priority for every data group for collation classified based on registrant biometric classification information, and the biometric information collation part 70b is authentication subject person biometric information. The authentication subject person fingerprint feature data and the registrant fingerprint feature data are collated based on the feature data, the authentication subject biometric classification information and the collation priority order, and a collation result is output.

  Therefore, a personal authentication system using biometric information related to the present invention is a system for authenticating an individual using biometric information, and a storage unit 61 that can hold data, and for collecting biometric information of a registrant. The registration device 60 that obtains the verification data from the information input using the biometric information input unit 60a for registration and writes it in the storage unit 61, and whether or not the person to be authenticated is the person is written in the storage unit 61. The authentication apparatus 70 is configured to be collated / determined using the collation data.

  With this configuration, the biometric information feature data extraction unit 60b extracts the biometric information feature data of the authentication subject from the biometric information collected by the biometric information input unit 60a, and the biometric classification information extraction unit 60c Authentication subject biometric classification information is extracted from the collected biometric information, and both the data and information are input to the biometric information matching unit 70b. Further, the collation priority order definition unit 70a outputs the collation priority order for each collation data group classified based on the registrant biometric classification information.

  Then, the biometric information matching unit 70b reads data and information from each of the biometric information feature data storage unit 61a and the biometric information storage unit 61b, and the data, information, authentication subject biometric information feature data, and authentication. Based on the subject biometric classification information and the collation priority, the subject of authentication is collated and a collation result is output. Further, the determination unit 70c determines the person / non-person based on the collation result.

Thus, in registering biometric information feature data, collation priority orders to be collated are defined, biometric information feature data and biometric information are held in separate storage devices, and collation is performed while referring to collation priority orders. The
Further, as described above, since the authentication device 70 uses biometric information to perform collation in consideration of the collation priority order, the biometric information input as the biometric information for authentication and the biometric information of the registration data for collation are complete. Even if it does not match, it can be verified reliably.

Furthermore, since both the registration and authentication input devices can be used, verification can be performed without adding a new device for inputting seal stamp data.
(C) Others The present invention is not limited to the above-described embodiments and modifications thereof, and various modifications can be made without departing from the spirit of the present invention.

For example, a known technique can be used as a technique for collecting a fingerprint image, extracting fingerprint feature data from the collected fingerprint image, and detecting a fingerprint center coordinate.
The definition examples shown in FIGS. 27 to 30 show the collation priorities specialized when the procedures shown in FIGS. 13, 14, 17, 21, and 22, respectively, are used, and have already been proposed. It does not indicate the collation priority common to the pattern determination. Therefore, the superiority of the present invention is not impaired at all even by performing the classification by providing a pattern type different from the pattern type described above.

  Further, the fifth collation priority order defined in FIGS. 27 to 30 is very unlikely to be determined from the procedures shown in FIGS. 13, 14, 17, 21, and 22, respectively. For this reason, when there is no pattern that matches the matching registration data up to the fourth matching priority, it is possible to omit the matching with the matching registration data of the pattern indicated by the matching priority and reject the personal authentication. .

When defining the collation priority from the characteristics of the pattern judgment procedure using the pattern judgment procedure other than the right flow type, wave type, vortex type, right flow type, etc. It is also possible to use the pattern determination or collation priority definition process.
(D) Appendix (Appendix 1) In a system for authenticating an individual using fingerprint information,
A storage that can hold data;
A registration device that obtains verification data from information input using a registration fingerprint information input unit for collecting registrant fingerprint information and writes the data to the storage unit;
And an authentication device for verifying / determining whether or not the person to be authenticated is the person using the verification data written in the storage unit,
The registration device
A registrant fingerprint feature data extraction unit that extracts registrant fingerprint feature data from the registrant fingerprint image collected by the registration fingerprint information input unit;
A registrant fingerprint classification information extraction unit that extracts registrant fingerprint classification information from the registrant fingerprint image collected by the registration fingerprint information input unit;
A collation data generation unit that writes the collation data to the storage unit in association with the registrant fingerprint feature data and the registrant fingerprint classification information;
further,
The authentication device
An authentication subject fingerprint feature data extraction unit that extracts authentication subject person fingerprint feature data from the authentication subject person fingerprint image collected by the authentication fingerprint information input unit for collecting the authentication subject person fingerprint information;
An authentication subject fingerprint classification information extraction unit that extracts authentication subject person fingerprint classification information from the authentication subject person fingerprint image collected by the authentication fingerprint information input unit;
A collation priority definition unit capable of outputting a collation priority for each collation data group classified based on the registrant fingerprint classification information;
A verification unit that compares the authentication target person fingerprint feature data with the registrant fingerprint feature data based on the authentication target person fingerprint feature data, the authentication target person fingerprint classification information, and the verification priority, and outputs a verification result; A personal authentication system using fingerprint information, characterized in that it is configured.

(Appendix 2) The authentication device
The personal authentication system using fingerprint information according to appendix 1, characterized in that it comprises a determination unit for determining the identity / non-identity based on the collation result.
(Supplementary note 3)
The personal authentication system using fingerprint information according to appendix 2, wherein the personal / non-person is determined in consideration of the collation priority.

(Supplementary note 4)
Supplementary Note 2 or Supplementary Note 3, wherein the authentication subject person fingerprint classification information extraction unit extracts the authentication subject person fingerprint classification information and is configured to determine the person / non-person. Personal authentication system using fingerprint information.
(Supplementary note 5)
Using the fingerprint information according to any one of appendix 2 to appendix 4, wherein the match rate output from the collation unit is configured to determine the person / non-person as the collation result Personal authentication system.

(Appendix 6) The registrant fingerprint classification information extraction unit
Using the fingerprint information according to appendix 1, wherein the registrant fingerprint classification information is extracted by using a portion opposite to the fingerprint tip side with respect to the fingerprint center of the registrant fingerprint image. Personal authentication system.
(Supplementary Note 7) The authentication subject person fingerprint classification information extraction unit
The fingerprint information as set forth in appendix 1, wherein the authentication target person fingerprint classification information is extracted using a portion of the fingerprint center of the authentication subject person fingerprint image opposite to the front end side of the fingerprint. The personal authentication system used.

(Supplementary Note 8) If the determination unit cannot determine the person / non-person until the collation work in the predetermined collation priority is completed,
The fingerprint information according to any one of appendix 1 to appendix 7, wherein the collation unit is configured to limit personal authentication at a collation priority lower than the predetermined collation priority. There was a personal authentication system.

(Supplementary Note 9) The registration apparatus includes a collation priority order determination processing unit that determines and outputs a collation priority according to the registrant fingerprint classification information,
The verification data generation unit
The fingerprint information according to appendix 1, wherein the verification data is configured to write the registrant fingerprint feature data, the registrant fingerprint classification information, and the verification priority in association with each other in the storage unit. Personal authentication system using

(Appendix 10) The authentication device
An authentication subject fingerprint feature data extraction unit that extracts authentication subject person fingerprint feature data from the authentication subject person fingerprint image collected by the authentication fingerprint information input unit for collecting the authentication subject person fingerprint information;
An authentication subject fingerprint classification information extraction unit that extracts authentication subject person fingerprint classification information from the authentication subject person fingerprint image collected by the authentication fingerprint information input unit;
The authentication target person fingerprint feature data and the registrant fingerprint feature data are collated based on the collation priority order included in the collation data, and a collation unit that outputs the collation result is provided. A personal authentication system using fingerprint information according to appendix 7, which is a feature.

(Supplementary Note 11) The collation priority definition part
The fingerprint information according to any one of appendix 1 to appendix 8, characterized in that data relating the registrant fingerprint classification information and the collation priority is stored as the collation data. Personal authentication system using
(Supplementary note 12)
The personal authentication system using fingerprint information according to any one of appendix 1 to appendix 11, wherein the collation data is stored for each classification type of the registrant fingerprint classification information .

(Supplementary note 13)
A registrant fingerprint feature data storage unit capable of holding the registrant fingerprint feature data;
13. A personal authentication system using fingerprint information according to appendix 12, characterized by comprising a registrant fingerprint classification information storage unit capable of holding the registrant fingerprint classification information.

(Supplementary Note 14) The registrant fingerprint feature data extraction unit of the registration device and the authentication subject fingerprint feature data extraction unit of the authentication device are shared,
Supplementary notes 1 to 8, 10, 11 characterized in that the registrant fingerprint classification information extraction unit of the registration device and the authentication subject fingerprint classification information extraction unit of the authentication device are shared. A personal authentication system using the fingerprint information according to any one of the above.

(Supplementary Note 15) In a system for authenticating an individual using biometric information,
A storage that can hold data;
A registration device that obtains verification data from information input using a registration biometric information input unit for collecting biometric information of a registrant and writes the data to the storage unit;
And an authentication device for verifying / determining whether or not the person to be authenticated is the person using the verification data written in the storage unit,
The registration device
A registrant biometric information feature data extraction unit for extracting registrant biometric information feature data from a biometric information image of the registrant collected by the biometric information input unit for registration;
A registrant bioclassification information extraction unit that extracts registrant bioclassification information from the biometric information image of the registrant collected by the biometric information input unit for registration;
A collation data generation unit that writes the collation data to the storage unit in association with the registrant biometric feature data and the registrant bioclassification information;
further,
The authentication device
An authentication subject biometric information feature data extraction unit for extracting authentication subject biometric information feature data from the authentication subject biometric information image collected by the authentication biometric information input unit for collecting biometric information of the authentication subject; ,
An authentication subject biometric classification information extraction unit that extracts authentication subject biometric classification information from the authentication subject biometric information image collected by the authentication biometric information input unit;
A collation priority definition unit capable of outputting a collation priority for each collation data group classified based on the registrant biometric classification information;
A verification unit for verifying the authentication target person fingerprint feature data and the registrant fingerprint feature data based on the authentication target person biometric information feature data, the authentication target biometric classification information, and the verification priority, and outputting a verification result; A personal authentication system using biometric information, characterized by comprising:

(Supplementary Note 16) While the registrant biometric information feature data extraction unit of the registration device and the authentication subject biometric information feature data extraction unit of the authentication device are shared,
The individual using biometric information according to appendix 15, wherein the registrant biometric classification information extracting unit of the registration device and the authentication subject biometric classification information extracting unit of the authentication device are shared. Authentication system.

(Supplementary Note 17) In a system for performing personal authentication using fingerprint information,
A first extraction step of extracting registrant fingerprint feature data from the collected registrant fingerprint image;
A registrant fingerprint feature data registration step of writing the registrant fingerprint feature data extracted in the first extraction step into a storage unit capable of holding data;
A second extraction step of extracting registrant fingerprint classification information from the registrant fingerprint image;
A registrant fingerprint classification information registration step of writing the registrant fingerprint classification information extracted in the second extraction step into the storage unit;
A collation data registration step of writing collation data into the storage unit in association with the registrant fingerprint feature data and the registrant fingerprint classification information;
A collation data extracting step for extracting the collation data from the storage unit;
A third extraction step of extracting authentication target person fingerprint feature data from the collected authentication object person fingerprint image;
A fourth extraction step of extracting authentication subject person fingerprint classification information from the authentication subject person fingerprint image;
A collation priority definition step for outputting a collation priority for each collation data group classified based on the registrant fingerprint classification information;
A verification / determination step is configured to verify / determine whether or not the authentication target person is the person based on the authentication target fingerprint feature data, the authentication target fingerprint classification information, and the verification priority. A registration / authentication method for a personal authentication system using fingerprint information.

(Supplementary Note 18) In a system for performing personal authentication using fingerprint information,
A center detection step for detecting the fingerprint center coordinates from the fingerprint image of the person to be authenticated;
A division step of dividing the authentication subject fingerprint image into a plurality of unit blocks;
A recognition step for recognizing a central unit block having the fingerprint center coordinates detected in the center detection step among the plurality of unit blocks divided in the division step;
A block group generation step of generating a block group having the plurality of unit blocks from a portion on the opposite side of the fingerprint tip side with respect to the central unit block recognized in the recognition step;
A unit block flow direction determination step for determining a local fingerprint wave flow direction for each of the plurality of unit blocks constituting the block group generated in the block group generation step;
Using fingerprint information, comprising: fingerprint classification information determination step for determining fingerprint classification information based on the fingerprint wave flow direction determined in the unit block flow direction determination step Judgment method for personal authentication system.

(Supplementary note 19) In a system for performing personal authentication using fingerprint information,
A registration device for obtaining and writing verification data from information input using a registration fingerprint information input unit for collecting registrant fingerprint information,
A registrant fingerprint feature data extraction unit that extracts registrant fingerprint feature data from the registrant fingerprint image collected by the registration fingerprint information input unit;
A registrant fingerprint classification information extraction unit that extracts registrant fingerprint classification information from the registrant fingerprint image collected by the registration fingerprint information input unit;
A collation priority determination processing unit that determines and outputs a collation priority according to the registrant fingerprint classification information;
A collation data generation unit is provided that writes the collation data into a storage unit that can hold the data by associating the registrant fingerprint feature data, the registrant fingerprint classification information, and the collation priority. A registration apparatus for a personal authentication system using fingerprint information.

(Supplementary Note 20) In a system for authenticating an individual using fingerprint information,
An authentication apparatus for verifying / determining whether or not the person to be authenticated is a person using verification data written in a storage unit capable of holding data,
An authentication subject fingerprint feature data extraction unit that extracts authentication subject person fingerprint feature data from the authentication subject person fingerprint image collected by the authentication fingerprint information input unit for collecting the authentication subject person fingerprint information;
An authentication subject fingerprint classification information extraction unit that extracts authentication subject person fingerprint classification information from the authentication subject person fingerprint image collected by the authentication fingerprint information input unit;
A collation priority definition unit capable of outputting a collation priority for each collation data group classified based on the registrant fingerprint classification information;
A verification unit that compares the authentication target person fingerprint feature data with the registrant fingerprint feature data based on the authentication target person fingerprint feature data, the authentication target person fingerprint classification information, and the verification priority, and outputs a verification result; An authentication device for a personal authentication system using fingerprint information, characterized in that it is configured.

(Supplementary Note 21) A computer used in a system for performing personal authentication using fingerprint information,
A registrant fingerprint feature data extraction unit that extracts registrant fingerprint feature data from the registrant fingerprint image collected from the registrant;
A fingerprint classification information extraction unit that extracts registrant fingerprint classification information from the registrant fingerprint image and extracts the registrant fingerprint classification information;
A collation priority determination processing unit that determines and outputs a collation priority according to the registrant fingerprint classification information;
A program for causing the collation data to function as a collation data generation unit that writes the registrant fingerprint feature data, the registrant fingerprint classification information, and the collation priority in a storage unit that can hold data is recorded. Computer-readable recording medium.

(Supplementary Note 22) A computer used in a system for performing personal authentication using fingerprint information,
An authentication subject fingerprint feature data extraction unit that extracts the authentication subject fingerprint feature data from the authentication subject fingerprint image collected from the authentication subject;
An authentication target person fingerprint classification information extraction unit that extracts authentication target person fingerprint classification information from the authentication target person fingerprint image;
A collation priority definition unit capable of outputting a collation priority for each collation data group classified based on the registrant fingerprint classification information;
Functions as a collation unit that collates the authentication subject fingerprint feature data with the registrant fingerprint feature data based on the authentication subject fingerprint feature data, the authentication subject fingerprint classification information, and the collation priority, and outputs a collation result A computer-readable recording medium on which a program for causing the program to be recorded is recorded.

49, 49a, 49b Personal authentication system 50, 50a, 60 Registration device 51, 51a, 51b, 70 Authentication device 52, 61 Storage unit (database)
60a biometric information input unit 60b registrant biometric information feature data extraction unit 60c registrant biometric classification information extraction unit 60d collation data generation unit 61a biometric information feature data storage unit 61b biometric information storage unit 70b biometric information collation unit 70c, 115 determination unit DESCRIPTION OF SYMBOLS 100 Fingerprint information input part 101 Fingerprint feature data extraction part 102 Fingerprint pattern information extraction part 111 Fingerprint feature data storage part 112 Fingerprint pattern information storage part 113 Collation priority definition part 114 Fingerprint information collation part 114a 2nd fingerprint information collation part 121 For collation Data generation unit 132 Feature data generation unit 133, 70a Fingerprint pattern classification unit 201 Collation priority order reference unit 221 Collation data selection unit 300a Biometric information input unit 310 Collation priority order determination processing unit

Claims (4)

In a system that authenticates an individual using fingerprint information,
A registration device for obtaining and writing verification data from information input using a registration fingerprint information input unit for collecting registrant fingerprint information,
A registrant fingerprint feature data extraction unit that extracts registrant fingerprint feature data from a registrant fingerprint image collected by the registration fingerprint information input unit;
A registrant fingerprint classification information extraction unit that extracts registrant fingerprint classification information from the registrant fingerprint image collected by the registration fingerprint information input unit;
A collation priority determination processing unit that determines and outputs a collation priority according to the registrant fingerprint classification information;
A collation data generation unit is provided that writes the collation data into a storage unit that can hold the data by associating the registrant fingerprint feature data, the registrant fingerprint classification information, and the collation priority. A registration apparatus for a personal authentication system using fingerprint information. In a system that authenticates an individual using fingerprint information,
An authentication apparatus for verifying / determining whether or not the person to be authenticated is a person using verification data written in a storage unit capable of holding data,
An authentication subject fingerprint feature data extraction unit that extracts authentication subject person fingerprint feature data from the authentication subject person fingerprint image collected by the authentication fingerprint information input unit for collecting the authentication subject person fingerprint information;
An authentication subject fingerprint classification information extraction unit that extracts authentication subject person fingerprint classification information from the authentication subject person fingerprint image collected by the authentication fingerprint information input unit;
A collation priority definition unit capable of outputting a collation priority for each collation data group classified based on the registrant fingerprint classification information;
A verification unit that compares the authentication target person fingerprint feature data with the registrant fingerprint feature data based on the authentication target person fingerprint feature data, the authentication target person fingerprint classification information, and the verification priority, and outputs a verification result; An authentication device for a personal authentication system using fingerprint information, characterized in that it is configured. A computer used in a system that performs personal authentication using fingerprint information.
A registrant fingerprint feature data extraction unit that extracts registrant fingerprint feature data from the registrant fingerprint image collected from the registrant;
A fingerprint classification information extraction unit that extracts registrant fingerprint classification information from the registrant fingerprint image and extracts the registrant fingerprint classification information;
A collation priority determination processing unit that determines and outputs a collation priority according to the registrant fingerprint classification information;
Fingerprint information for causing the collation data to function as a collation data generation unit that writes the registrant fingerprint feature data, the registrant fingerprint classification information, and the collation priority in a storage unit that can hold data. A computer-readable recording medium on which a registration program for a personal authentication system used is recorded. A computer used in a system that performs personal authentication using fingerprint information.
An authentication subject fingerprint feature data extraction unit that extracts the authentication subject fingerprint feature data from the authentication subject fingerprint image collected from the authentication subject;
An authentication target person fingerprint classification information extraction unit that extracts authentication target person fingerprint classification information from the authentication target person fingerprint image;
A collation priority definition unit capable of outputting a collation priority for each collation data group classified based on the registrant fingerprint classification information;
Functions as a collation unit that collates the authentication subject fingerprint feature data with the registrant fingerprint feature data based on the authentication subject fingerprint feature data, the authentication subject fingerprint classification information, and the collation priority, and outputs a collation result A computer-readable recording medium on which an authentication program for a personal authentication system using fingerprint information is recorded.

Images