JP2007114931A - Authentication apparatus - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an authentication apparatus capable of identifying and appropriately authenticating a plurality of people having similar appearances, such as family members. <P>SOLUTION: The growth rate of a human being in physical size is generally greater the younger he or she is. The growth rate slows down past 20 years of age and the distance Le between the outer corners of eyes has characteristics similar to those of a growth curve Ge. The authentication apparatus for performing face authentication then measures the distance (previous distance) Le between the outer corners of eyes at the time of registration from the feature data of a registered face during an authentication process and compares the previous distance with the present distance Le between the outer corners of the eyes. Thus when e.g. a son at the age of eleven who looks like his father requests collation, the previous distance Le30 of the father a year ago (30 years of age) is greater than the present distance Le 11 of the son, and hence it is possible to prevent the son from being authenticated as the father, by disabling the authentication if the previous distance is greater than the present distance. As a result, it is possible to identify and appropriately authenticate a plurality of people with similar appearances, such as family members. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to an authentication apparatus that performs authentication related to an authentication registrant and an authentication target person.

  In recent years, in the field of security and the like, there is an increasing demand for biometric authentication, particularly face authentication that excels in usability.

  As the face authentication technique, there is a technique disclosed in Patent Document 1, for example. In this technology, when registering facial feature data (registration data), the registration time is recorded, and the registration data whose elapsed time from the registration time to the authentication time is long is deleted. Therefore, it is possible to prevent the registration data from being enlarged.

JP 2004-54888 A

  However, since the face authentication technique of Patent Document 1 described above only deletes registration data that has become obsolete using information on registration time and importance, there is no direct contribution to improving authentication accuracy. In this case, it is difficult to discriminate a plurality of persons having a similar appearance like a family, and there is a possibility of authenticating a person to be authenticated who is not the person himself / herself.

  The present invention has been made in view of the above problems, and an object of the present invention is to provide an authentication apparatus that can identify and appropriately authenticate a plurality of persons having a similar appearance like a family.

  In order to solve the above-mentioned problems, the invention of claim 1 is an authentication device that performs authentication of identity between an authentication registrant and an authentication target person, and (a) predetermined biometric information relating to the authentication registrant 1st acquisition means which acquires as 1 biometric information, and registers with a recording means, (b) Measurement of the said authentication registrant at the 1st time when acquiring said 1st biometric information about a predetermined secular change parameter First measurement means for acquiring a value; (c) second acquisition means for acquiring the predetermined biological information relating to the person to be authenticated as second biological information; and (d) for the predetermined aging parameter, Second measurement means for acquiring a measurement value of the person to be authenticated at a second time point when acquiring second biological information; (e) the first biological information; the second biological information; For each measured value at the first time point and the second time point Hazuki, and a determining means for determining the identity.

  The invention according to claim 2 is the authentication apparatus according to claim 1, wherein the first acquisition means associates (a-1) age information about the authentication registrant with the first biological information. There is an age information recording means for recording in the recording means, and the determination means has (e-1) means for determining the identity based on the age information.

  According to a third aspect of the present invention, in the authentication apparatus according to the second aspect of the invention, the age information recording unit records the age information and the first biological information as a set on the same recording medium. Have means.

  According to a fourth aspect of the present invention, there is provided the authentication apparatus according to any one of the first to third aspects, wherein the first acquisition means includes (a-2) information on change characteristics relating to the predetermined aging parameter. Is recorded in the recording means in association with the first biological information, and the determination means is (e-2) means for determining the identity based on the information on the change characteristics. Have

  The invention according to claim 5 is the authentication apparatus according to claim 4, wherein the characteristic information recording means records the change characteristic information and the first biological information as a set on the same recording medium. Have means for

  The invention of claim 6 is the authentication apparatus according to any one of claims 1 to 5, wherein the predetermined secular change parameter is a parameter whose measured value increases with age, and the determination (E-3) If the measured value of the person to be authenticated obtained by the second measuring means is smaller than the measured value of the authentication registrant obtained by the first measuring means, the identity is determined. Has a means to deny.

  The invention according to claim 7 is the authentication apparatus according to any one of claims 1 to 6, wherein the first acquisition means includes: (a-3) time information relating to the first time point; Time information recording means for recording in the recording means in association with one biological information, and the determination means includes (e-4) time information recorded in the recording means and time information related to the second time point. And means for determining the identity.

  Further, the invention of claim 8 is an authentication apparatus for performing authentication by identifying a corresponding registrant corresponding to an authentication target person from a plurality of authentication registrants, and (a) a predetermined number of each of the plurality of authentication registrants. First acquisition means for acquiring biological information as first biological information and registering it in the recording means; and (b) the first biological information at each first time point when acquiring the first biological information for a predetermined aging parameter. First measurement means for acquiring measured values of each of a plurality of authentication registrants; (c) second acquisition means for acquiring the predetermined biological information relating to the authentication subject as second biological information; Second measurement means for acquiring a measurement value of the person to be authenticated at a second time point when acquiring the second biological information for a predetermined secular change parameter; and (e) a second for each of the plurality of authentication registrants. 1 biological information and the second information And body information, based the on the respective measurement value at each first time point and the second point comprises identifying means for identifying the corresponding registrant.

  The invention according to claim 9 is the authentication apparatus according to the invention according to claim 8, wherein the first acquisition unit is configured to: (a-1) obtain age information regarding each of the plurality of authentication registrants in the first biometric information; And an age information recording unit that causes the recording unit to record in association with (1), and (e-1) a unit that identifies the corresponding registrant based on the age information.

  Further, the invention of claim 10 is the authentication device according to claim 9, wherein the age information recording means records the age information and the first biological information as a set on the same recording medium. Have means.

  The invention according to claim 11 is the authentication apparatus according to any one of claims 8 to 10, wherein the first acquisition means is (a-2) information on change characteristics relating to the predetermined aging parameter. Characteristic information recording means for causing the recording means to record the information in association with the first biological information, and the identification means (e-2) means for identifying the corresponding registrant based on the information of the change characteristic Have

  The invention according to claim 12 is the authentication apparatus according to claim 11, wherein the characteristic information recording unit records the change characteristic information and the first biological information as a set on the same recording medium. Have means for

  The invention according to claim 13 is the authentication device according to any one of claims 8 to 12, wherein the predetermined secular change parameter is a parameter whose measured value increases with age, and the identification (E-3) If the measured value of the person to be authenticated acquired by the second measuring means is smaller than the measured value of the authentication registrant acquired by the first measuring means, the authentication registrant And means for denying the identity of the person to be authenticated.

  The invention according to claim 14 is the authentication device according to any one of claims 8 to 13, wherein the first acquisition means includes: (a-3) time information relating to each first time point; Time information recording means for causing the recording means to record in association with the first biological information, wherein the identification means includes (e-4) time information recorded in the recording means and time information relating to the second time point. And means for identifying the corresponding registrant.

  In addition, according to a fifteenth aspect of the present invention, in the authentication device according to any one of the first to fourteenth aspects, the predetermined biological information is information relating to facial features.

  According to invention of Claim 1 thru | or 7 and Claim 15, in the 1st time about the 1st biometric information regarding an authentication registrant, the 2nd biometric information regarding an authentication object person, and a predetermined secular change parameter Based on the measurement value of the authentication registrant and the measurement value of the authentication target person at the second time point, the identity of the authentication registrant and the authentication target person is determined. As a result, a plurality of persons having a similar appearance like a family can be distinguished and appropriately authenticated.

  In particular, in the invention of claim 2, since the identity is determined based on the age information regarding the authentication registrant recorded in association with the first biological information, the accuracy of authentication can be improved.

  According to the invention of claim 3, since age information and first biological information are recorded as a set on the same recording medium, management of these information becomes easy.

  According to the fourth aspect of the present invention, since the identity is determined based on the change characteristic information related to the predetermined aging parameter recorded in association with the first biological information, the accuracy of authentication can be improved.

  According to the fifth aspect of the present invention, since the change characteristic information and the first biological information are recorded as a set on the same recording medium, management of the information becomes easy.

  According to the invention of claim 6, if the measured value of the person to be authenticated acquired by the second measuring means is smaller than the measured value of the authentication registrant acquired by the first measuring means, the identity is denied. Therefore, it is possible to appropriately distinguish a plurality of persons having approximate appearances.

  According to the seventh aspect of the present invention, the identity is determined based on the time information related to the first time point and the time information related to the second time point recorded in association with the first biological information. Accuracy can be improved.

  Further, according to the inventions of claims 8 to 15, the first biometric information about each of the plurality of authentication registrants, the second biometric information about the person to be authenticated, and the respective first aging parameters. Based on the measurement values of the plurality of authentication registrants at the time point and the measurement values of the authentication target person at the second time point, a corresponding registrant corresponding to the authentication target person is identified from the plurality of authentication registrants. As a result, a plurality of persons having a similar appearance like a family can be distinguished and appropriately authenticated.

  In the invention of claim 9, since the corresponding registrant is identified based on the age information regarding each of the plurality of authentication registrants recorded in association with the first biological information, the accuracy of authentication can be improved.

  According to the invention of claim 10, since the age information and the first biological information are recorded as a set on the same recording medium, management of these information becomes easy.

  According to the invention of claim 11, since the corresponding registrant is identified on the basis of the change characteristic information related to the predetermined aging parameter recorded in association with the first biological information, the accuracy of authentication can be improved.

  According to the twelfth aspect of the present invention, since the change characteristic information and the first biological information are recorded as a set on the same recording medium, management of the information becomes easy.

  According to the invention of claim 13, when the measured value of the person to be authenticated acquired by the second measuring means is smaller than the measured value of the authentication registrant acquired by the first measuring means, Since the identity regarding the person to be authenticated is denied, a plurality of persons having approximate appearances can be appropriately determined.

  According to the fourteenth aspect of the present invention, since the corresponding registrant is identified based on the time information related to the first time point and the time information related to the second time point recorded in association with the first biological information, the elapsed time Can improve the authentication accuracy.

  According to the fifteenth aspect of the present invention, since the predetermined biological information is information relating to facial features, appropriate authentication can be performed in face authentication with excellent usability.

<First Embodiment>
<Main components of the authentication system>
FIG. 1 is an external view showing a main configuration of an authentication system 1A according to the first embodiment of the present invention.

  The authentication system 1A includes an image input device 2 and a personal computer (hereinafter also referred to as “PC”) 3A, and functions as an authentication device that performs face authentication as biometric authentication.

  The image input device 2 includes two image capturing cameras (hereinafter also simply referred to as “cameras”) 21 and 22, and these cameras 21 and 22 are the authentication target person HMa who is the subject from different viewpoints. It is arranged so that the face can be photographed.

  When the face image of the person to be authenticated HMa is photographed by the image input device 2, the appearance information of the person to be authenticated HMa obtained by the photographing, that is, two face images are transmitted to the PC 3A via the communication line. Note that the image data communication method between the cameras 21 and 22 and the PC 3A is not limited to the wired method, and may be a wireless method.

  The PC 3A includes a main body unit 30, a display unit 31, and an operation input unit 32 including a keyboard and a mouse.

  The main body 30 has a CPU and a memory, and is a part that comprehensively controls each part of the PC 3A. Further, the main body unit 30 includes a storage unit configured by, for example, a hard disk drive (HDD), and stores an authentication database described later.

  FIG. 2 is a diagram illustrating a functional configuration of the authentication system 1A.

  The PC 3A includes a feature extraction unit 36 and an authentication processing unit 37A.

  The feature extraction unit 36 extracts a characteristic part of the face related to the authentication target person HMa (hereinafter also simply referred to as “characteristic part”) based on the image of the authentication target person HMa input from the image input device 2.

  The facial feature part may be eyes, nose or mouth, and can be detected by template matching using a standard template of the feature part, for example. Then, for the detected feature part, for example, as shown in FIG. 3, the coordinates of the feature points Q1 to Q7 of each part are calculated as feature data (information on facial features). The coordinates of these feature points Q1 to Q7 are based on the principle of triangulation based on stereo images obtained by the two cameras 21 and 22 and camera parameters indicating the positions and orientations of the cameras 21 and 22. The three-dimensional position of each feature point Q1-Q7 is calculated.

  Then, the feature data (predetermined biological information) acquired by the feature extraction unit 36 is registered in the authentication database 35 functioning as a recording unit or sent to the authentication processing unit 37A.

  The authentication processing unit 37A is configured mainly for face authentication, and has a function of authenticating each individual with a face image. That is, the degree of similarity between the feature data of the registrant (authentication registrant) registered in advance in the authentication database 35 and the feature data of the person to be authenticated HMa sent from the feature extraction unit 36 is calculated, and the face authentication process is performed. Done.

  For example, the above similarity is calculated based on the sum of the distances (square errors) from the feature point Q5 corresponding to the apex of the nose to the other feature points Q1 to Q4 and Q6 to Q7. Thereby, if it is the same person, it becomes possible to make similarity high.

  The registration operation of feature data in the authentication system 1A having the above-described functional configuration will be described below.

  FIG. 4 is a flowchart showing an operation of registering feature data in the authentication system 1A.

  First, an image of a person (registrant) who registers feature data is acquired by the image input device 2 and input to the PC 3A (step ST1).

  In step ST2, the feature extraction unit 36 extracts facial feature data based on the image input in step ST1. That is, the characteristic data (first biological information) regarding the registrant is acquired.

  In step ST3, the feature data extracted in step ST2 is registered in the authentication database 35 in association with the registrant ID information. This ID information includes information such as an ID number and a name.

  By using the authentication database 35 in which the feature data and ID information are registered in this way, the authentication system 1A performs a collation operation (1: 1) for determining the identity of the designated registrant and the person to be authenticated HMa. Authentication) and an identification operation (1: N authentication) for identifying registrants (corresponding registrants) corresponding to the authentication target person HMa from all registrants. These operations will be described below in order. .

  FIG. 5 is a flowchart showing a verification operation in the authentication system 1A.

  In step ST11, one registrant who wishes to collate is specified from among a plurality of registrants registered in the authentication database 35. The designation of the registrant is performed by the authentication person HMa operating the operation input unit 32A to input an ID number or name, or a personal identification number that only the person can know. The information such as the password is preferably recorded in the authentication database 35 in association with the registrant ID information.

  In step ST12, feature data associated with the registrant designated in step ST11 is read from the authentication database 35.

  In steps ST13 to ST14, operations similar to those in steps ST1 and ST2 are performed.

  In step ST15, the feature data (second biological information) of the person to be authenticated HMa extracted in step ST14 is compared with the feature data (first biological information) read in step ST12. Specifically, whether the designated registrant and the authentication target person HMa are the same person by calculating the similarity of each feature data in the authentication processing unit 37A and comparing it with a predetermined threshold value. Judge whether or not.

  FIG. 6 is a flowchart showing an identification operation in the authentication system 1A.

  In steps ST21 to ST22, operations similar to those in steps ST1 and ST2 are performed.

  In step ST23, the authentication processing unit 37A sequentially calculates the similarity between the feature data of the person to be authenticated HMa extracted in step ST22 and each feature data registered in the authentication database 35.

  In step ST24, ID information such as a name associated with the feature data having the highest similarity among the similarities with the feature data calculated in step ST23 is output to the display unit 31, for example. When displaying this ID information, in order to make it easy to understand who the identified person (registrant) is, for example, a thumbnail image of the registrant recorded in association with the feature data in the authentication database 35 is displayed. It is preferable to do this.

  As a usage form of the authentication system 1A capable of performing the above-described operation, for example, a case where a family uses the PC 3A jointly can be considered.

  In such a case, after first registering the feature data of all the family members in the authentication database 35, when each person starts using the PC 3A, the ID information of the use starter is input, or each member of the family member is registered. A person to be collated is specified by selecting an icon of a use starter from a plurality of icons to be represented. Then, when the verification operation for the designated person is performed in the authentication system 1A, and it is determined that the designated person and the use start person are the same person, that is, the person himself / herself, the setting previously registered by the person ( For example, the login process is performed when a personal wallpaper or the like is displayed on the display unit 31). When it is determined that the user is not the person (when authentication fails), it is preferable to perform a login process if the user is confirmed by, for example, inputting a personal identification number.

  As described above, in the case where the entire family uses one PC 3A, the appearances are often similar between parents and siblings, between siblings and sisters, etc., so these persons cannot be distinguished and authentication accuracy tends to be lowered. For example, when the son SN shown in FIG. 7 starts to use the PC 3A, the similarity between the son SN and the feature data of the father FT whose appearance is similar also increases. is there.

  Therefore, in the authentication system 1A according to the present embodiment, an authentication process for performing appropriate authentication by distinguishing a parent and child whose appearance is similar is possible, which will be described in detail below.

<Authentication process of authentication system 1A>
Growing children, etc., monotonically change the size and positional relationship of the entire face or part of the face as they age. For example, the distance Le between eyes (hereinafter also referred to as “inter-eye distance”) Le shown in FIG. 8, that is, the distance between the feature point Q1 and the feature point Q4 in FIG. 3 increases with age.

  Therefore, in the authentication system 1A of the present embodiment, the accuracy of authentication is improved by using the distance between eye corners (aging change parameter) Lb, whose measured value increases with age, for discrimination between parents and children.

  It is known that humans generally have a higher degree of growth in terms of body size as they are younger, and that the degree of growth slows down after the age of 20, but the above-mentioned distance Le between eyes is also shown in FIG. It has growth characteristics such as a growth curve Ge.

  For example, the distance between the corners Le30 of the father FT who was 30 years old one year ago and the distance Le10 between the corners of the son SN who was 10 years old was measured, and then the father FT who became 31 years old at the present time one year later The distance between the eye corners Le31 and the distance between the eye corners Le11 of the son SN who is 11 years old are measured. In this case, in the growing son SN, the distance between the corners of the eyes after one year is large, so Le10 <Le11. On the other hand, in the father FT whose physical growth has stopped substantially, the distance between the corners of the eyes after one year is substantially equal, so Le30≈Le31. Further, the distance Le11 between the eye corners of the son SN at the present time (11 years old) is smaller than the distance Le30 between the eye corners of the father FT one year ago (30 years old) by the distance Lf.

  By using such a change in the distance between the corners of the corners of the eyes, the distance between the corners of the corners of the eyes (hereinafter referred to as “past distance”) Lp obtained based on the registered feature data and the distance between the corners of the corners of the eye measured at the present time. Comparing Ln (hereinafter also referred to as “current distance”) makes it possible to distinguish between the father FT and the son SN. That is, the measured value of the distance between the corners of the registrant obtained at the time of acquiring characteristic data for the registrant (first time point) and the authentication obtained at the time of acquiring characteristic data for the person to be authenticated HMa (second time point). By comparing the measured value of the distance between the corners of the subject's eyes, a plurality of persons having a similar appearance like a family can be discriminated.

  Specifically, when the past distance Lp> the current distance Ln, a contradiction occurs with respect to the growth curve Ge in which the inter-eye distance Le increases monotonously, so that the person to be collated (registrant) and the person to be authenticated The identity with HMa can be denied. For example, the measured value of the distance between the corners of the son SN (authenticated person) currently measured is smaller than the measured value of the distance between the corners of the father FT (registrant) measured in the past. It is possible to prevent the son SN from being authenticated as the father FT as not being the person himself. On the other hand, there is a certain time lapse from the point of registration of the feature data to the present, but if the past distance Lp≈the current distance Ln, the authentication target person HMa is not a growing son but a father FT. I can judge.

  From the above, by comparing the past inter-eye distance Lp acquired based on the feature data registered in the authentication database 35 with the current inter-eye distance Ln, it is possible to accurately discriminate even the parent and child whose feature data is approximate. It will be possible.

  Specifically, when comparing feature data in step ST15 of FIG. 5, the past inter-eye distance Lp obtained from the feature data in the authentication database 35 and the current data obtained from the feature data obtained in step ST14. The distance between eye corners Ln is compared. Thus, for example, when the son SN shown in FIG. 7 starts using PC3A, the father FT (FIG. 7) whose appearance is similar but past distance Lp> current distance Ln can be excluded. It is possible to prevent the processing from being performed.

  In the identification operation shown in the flowchart of FIG. 6, when the operation of step ST24 is performed, the past distance Lp obtained from the feature data in the authentication database 35 is compared with the current distance Ln. Thereby, for example, when the son SN starts using PC3A, even if the calculation result that the similarity of the father FT is the highest and the similarity of the son SN is the second highest in step ST23 of FIG. In ST24, since the father FT having the highest similarity but the past distance Lp> the current distance Ln can be excluded, the identification result of the son SN having the second highest similarity can be output, and appropriate login processing can be performed. .

  By the above-described operation of the authentication system 1A, in order to compare the past inter-eye distance Lp and the current inter-eye distance Ln at the time of registration of the feature data in the authentication process, a plurality of persons having a similar appearance like a family are obtained. It can be identified and properly authenticated.

Second Embodiment
The authentication system 1B according to the second embodiment of the present invention has a configuration similar to that of the authentication system 1A shown in FIGS. 1 and 2, but the configuration of the authentication processing unit of the PC is different.

  That is, the authentication processing unit 37B of the PC 3B in the authentication system 1B is configured to be able to perform an authentication process described below.

<Authentication processing of authentication system 1B>
In the authentication process in the authentication system 1A of the first embodiment, erroneous authentication between parents and children is prevented using the distance Le between eyes at the time of registration, but in the authentication system 1B of the second embodiment, By recording age information about the registrant in the authentication database 35, the authentication process can be performed with high accuracy. This authentication process will be described in detail.

  Regarding the distance between eye corners Le, by collecting a large number of samples and performing statistical processing, it is possible to know specific characteristics of the secular change of the eye corner distance Le.

  FIG. 10 is a diagram for explaining the characteristics of the secular change in the distance between the corners Le of the eyes. Here, FIG. 10A shows the secular change after the age of 10 and FIG. 10B shows the secular change after the age of 30.

  The change in the distance Le between the eyes after the age of 10 is monotonously increased according to the number of years elapsed as in the secular change characteristic Ge1 shown in FIG. On the other hand, regarding the change in the distance between the eye corners after the age of 30 years, the distance between the eye corners Le is almost a constant value regardless of the number of years elapsed, like the secular change characteristic Ge2 shown in FIG.

  By referring to such aging characteristics Ge1 and Ge2, for example, when a 10-year-old son SN and a 30-year-old father FT register feature data, each age (10 years and 30 years) is associated with the feature data. If this information is recorded in the authentication database 35, the eye-to-eye distance Le at the time of authentication (current time) can be estimated based on the recorded age and the number of elapsed years. If the time information related to the registration time is recorded in the authentication database 35 in association with the feature data, the number of years since registration can be easily calculated using the time information at the authentication time (current time). Then, by comparing the estimated distance between eye corners Le and the distance between eye corners Le measured at the present time, accurate authentication can be performed.

  Specifically, first, a data table in which the aging characteristics (for example, FIG. 10A and FIG. 10B) of the inter-eye distance Le are stored in the authentication database 35 and managed. In addition, it is preferable to set the data table of the secular change characteristic in consideration of the past degree of growth in addition to general changes obtained from a large number of samples. Next, when the feature data is registered in the authentication database 35 in association with the ID information in step ST3 of FIG. 4, the age at the time of registration and the registration date and time are registered in the authentication database 35. When comparing feature data in step ST15 of FIG. 5, the age at the time of registration and the registration date / time are read from the authentication database 35, and the time difference between the registration date / time and the current time is added to the age at the time of registration. While calculating an age, based on the data table of the secular change characteristic preserve | saved at the authentication database 35, the distance between eye corners (henceforth "estimated distance") corresponding to the present age is estimated. Then, the estimated distance Lk and the current distance Ln are compared. In this comparison, since the comparison is made with the distance between the eye corners (estimated distance) Lk that is close to the actual situation, the accuracy is improved compared to the first embodiment that simply compares with the past distance Lp.

  Thus, for example, when the son SN shown in FIG. 7 starts using PC3A, the father FT (FIG. 7) whose appearance is approximate but estimated distance Lp> current distance Ln can be excluded, so login as father FT The processing can be prevented from being performed.

  Further, if the above-described authentication process is applied to the identification operation shown in the flowchart of FIG. 6, the accuracy in identifying the registrant corresponding to the authentication target person from a plurality of registrants is improved.

  The operation of the authentication system 1B described above provides the same effects as those of the first embodiment. Further, to record the age information of the registrant at the time of registration of the feature data, and to estimate the current inter-eye distance Le from this age information (and the elapsed time since registration) and the aging characteristics of the inter-eye distance Le. As a result, the accuracy of authentication can be improved.

  In the authentication system 1B, it is not essential to record the feature data and the age information as a set on one recording medium (authentication database 35), and the feature data and the age information are recorded on separate recording media. You may do it.

  In the authentication process of the authentication system 1B, it is not essential to record the age information in association with the feature data, and a data table (change characteristic information) representing the secular change from the registered age and the feature data are set as a set. It may be recorded on the same recording medium (for example, the authentication database 35). Also in this case, since the current distance between the eye corners Le can be estimated based on the data table associated with the feature data, the authentication accuracy can be improved.

<Third Embodiment>
The authentication system 1C according to the third embodiment of the present invention has a configuration similar to that of the authentication system 1A shown in FIGS. 1 and 2, but the configuration of the authentication processing unit of the PC is different.

  That is, the authentication processing unit 37C of the PC 3C in the authentication system 1C is configured to be able to perform an authentication process described below.

<Authentication process of authentication system 1C>
In the second embodiment described above, for example, the distance between the corners of the eye estimated from the distance between the corners Lp of the past based on the data table showing the transition of the distance between the corners Le of the eyes such as the secular change characteristics Ge1 and Ge2 (FIG. 10). (Estimated distance) Authentication is rejected when the current distance between the eye corners (current distance) Ln is smaller than Lk, but the human growth speed varies from person to person. It does not necessarily grow along the growth curve shown in FIG.

  Therefore, in the authentication system 1C of the third embodiment, even when the estimated distance Lk> the current distance Ln, the authentication is not immediately rejected, but this rejection condition is relaxed so that more appropriate authentication processing can be performed. To do. Specifically, even when the estimated distance Lk> the current distance Ln, the evaluation function is set so that the degree of similarity is relatively small compared to the case where the estimated distance Lk ≦ the current distance Ln. The authentication process is performed with the similarity obtained in step (hereinafter also referred to as “corrected similarity”) as the final similarity. This process will be specifically described.

  FIG. 11 is a diagram showing a flow of authentication processing in the authentication system 1C.

  First, the face image Ja of the authentication registrant (hereinafter referred to as “Mr. A”) is acquired by the image input device 2, and the feature data of Mr. A extracted by the feature extraction unit 36 based on this face image Ja and the current time Ta is associated and registered in the authentication database 35 (step SP11). On the other hand, the age of Mr. A is also registered in the authentication database 35 in parallel with the registration operation in step SP11 (step SP12).

  Next, when the person to be authenticated (hereinafter referred to as “Mr. B”) wishes to collate with Mr. A at the time Tb when the time has elapsed from the time Ta, the feature extraction unit 36 obtains the feature data of Mr. B. Extract (step SP21). Then, based on the feature data extracted in step SP21, the distance between B's eyes (current distance) Ln at the present time is measured (step SP22).

  On the other hand, by subtracting the time Ta from the time Tb, the elapsed time from the point of registration of the feature data of Mr. A is calculated (step SP13). Next, based on the elapsed time calculated in step SP13 and the age of Mr. A at the time of registration registered in the authentication database 35, the age of Mr. A at time Tb is calculated (step SP14). Then, based on a data table indicating the secular change of the inter-eye distance Le such as the secular change characteristics Ge1 and Ge2 (FIG. 10), the inter-eye distance (estimated distance) Lk of Mr. A corresponding to the time Tb is estimated ( Step SP15).

  Next, the process of reading the feature data of Mr. A from the authentication database 35 (step SP16) is performed, and then the read feature data of Mr. A, the estimated distance Lk estimated in step SP15, and extraction in step SP21. Based on the characteristic data of Mr. B and the current distance Ln measured at step SP22, the similarity (corrected similarity) between Mr. A and Mr. B is calculated (step SP30). In this case, the modified similarity Z is calculated based on the following equation (1).

Z = N + k × f (Lk, Ln) (1):
In the above equation (1), N is a similarity based only on Mr. A's feature data and Mr. B's feature data, and K is a predetermined coefficient (a coefficient obtained from empirical information). F (Lk, Ln) is a function that outputs (Lk−Ln) when Lk> Ln, and outputs H × (Ln−Lk) (where H> 1) when Lk ≦ Ln. . With this function, the similarity in the case of Lk> Ln with respect to the case of Lk ≦ Ln can be relatively reduced.

  Thereby, the comparison operation between Mr. A and Mr. B can be appropriately performed by comparing the modified similarity Z calculated in the above step SP30 with the predetermined threshold value α.

  When the above processing is applied to the identification operation shown in the flowchart of FIG. 6, the above modified similarity Z is calculated for all registrants registered in the authentication database 35 and corresponds to the largest modified similarity Z. ID information of the registrant to be output is output.

  By the operation of the authentication system 1C described above, the same effects as those of the first embodiment are obtained. Furthermore, since the similarity is corrected using the above equation (1), more appropriate authentication can be performed.

<Modification>
In each of the above-described embodiments, it is not essential to obtain a measurement value of the distance between the corners of the eyes in order to discriminate a parent and child, etc., and a secular change parameter (for example, a palm shape, etc.) that changes in size, shape, and color (Data) measurement values may be acquired.

  Moreover, it is not essential to measure the absolute value of the distance between the corners of the eyes, and a relative value may be measured. For example, a child's face is characterized by a smaller ratio of the vertical distance between the eyes and mouth relative to the distance between the corners of the eyes than an adult's face. Therefore, by acquiring a relative measurement value of the ratio of the vertical distance between the eyes and the mouth with respect to the lateral distance between the outer corners of the eyes, it may be used for discrimination of the parent and child.

  Further, the measurement values such as the distance between the corners of the eyes include not only values obtained by measurement but also those calculated as feature data.

  In each of the above embodiments, it is not essential to measure the distance between the corners of the eyes by stereo viewing using two cameras. For example, the distance between the corners of the eyes can be measured by an active method using a light cutting method. Also good. In this way, if the three-dimensional coordinate information is acquired using stereo vision or an active method, the absolute value of the face size and shape can be acquired, so that the influence of measurement errors due to face orientation and the like can be eliminated.

  In each of the embodiments described above, it is not essential to perform the authentication process with a personal computer, and another processing device that can perform the authentication process may be provided outside the personal computer.

It is an external view which shows the principal part structure of 1 A of authentication systems which concern on 1st Embodiment of this invention. It is a figure which shows the function structure of 1 A of authentication systems. It is a figure which shows the feature points Q1-Q7 regarding the feature part of a face. It is a flowchart which shows the operation | movement which registers the feature data in the authentication system 1A. It is a flowchart which shows the collation operation | movement in 1 A of authentication systems. It is a flowchart which shows the identification operation | movement in 1 A of authentication systems. It is a figure explaining that the appearance approximates in father FT and son SN. It is a figure which shows the distance Le between eyes. It is a figure for demonstrating the change of the distance Le between eyes according to age. It is a figure for demonstrating the characteristic of the secular change in the distance Le between eyes. It is a figure which shows the flow of the authentication process in the authentication system 1C which concerns on 3rd Embodiment of this invention.

Explanation of symbols

1A-1C Authentication System 2 Image Input Devices 3A-3C Personal Computer (PC)
31 Display part 35 Authentication database 36 Feature extraction part 37A-37C Authentication process part HMa Authentication subject Le Distance between eye corners Q1-Q7 Feature point

Claims (15)

An authentication device that performs authentication of identity between an authentication registrant and an authentication target person,
(a) first acquisition means for acquiring predetermined biometric information relating to the authentication registrant as first biometric information and registering it in a recording means;
(b) first measurement means for acquiring a measurement value of the authentication registrant at a first time point when acquiring the first biological information for a predetermined aging parameter;
(c) second acquisition means for acquiring the predetermined biological information related to the person to be authenticated as second biological information;
(d) second measurement means for acquiring a measurement value of the person to be authenticated at a second time point when acquiring the second biological information for the predetermined aging parameter;
(e) a determination unit that determines the identity based on the first biological information, the second biological information, and each measurement value at the first time point and the second time point;
An authentication device comprising: The authentication device according to claim 1,
The first acquisition means includes
(a-1) Age information recording means for causing the recording means to record information relating to the age of the authentication registrant in association with the first biological information;
Have
The determination means includes
(e-1) means for determining the identity based on the age information;
An authentication apparatus comprising: The authentication device according to claim 2,
The age information recording means includes
Means for recording the age information and the first biological information as a set on the same recording medium;
An authentication apparatus comprising: The authentication device according to any one of claims 1 to 3,
The first acquisition means includes
(a-2) characteristic information recording means for causing the recording means to record information on change characteristics relating to the predetermined aging parameter in association with the first biological information;
Have
The determination means includes
(e-2) means for determining the identity based on the information of the change characteristics;
An authentication apparatus comprising: The authentication apparatus according to claim 4,
The characteristic information recording means includes:
Means for recording the change characteristic information and the first biological information as a set on the same recording medium;
An authentication apparatus comprising: The authentication device according to any one of claims 1 to 5,
The predetermined aging parameter is a parameter whose measured value increases with age,
The determination means includes
(e-3) means for denying the identity when the measured value of the person to be authenticated acquired by the second measuring means is smaller than the measured value of the authentication registrant acquired by the first measuring means ,
An authentication apparatus comprising: The authentication device according to any one of claims 1 to 6,
The first acquisition means includes
(a-3) time information recording means for causing the recording means to record time information related to the first time point in association with the first biological information;
Have
The determination means includes
(e-4) means for determining the identity based on time information recorded in the recording means and time information related to the second time point;
An authentication apparatus comprising: An authentication device that performs authentication by identifying a corresponding registrant corresponding to an authentication target person from a plurality of authentication registrants,
(a) first acquisition means for acquiring predetermined biological information relating to each of the plurality of authentication registrants as first biological information and registering it in a recording means;
(b) first measurement means for acquiring measured values of each of the plurality of authentication registrants at each first time point when acquiring the first biological information for a predetermined aging parameter;
(c) second acquisition means for acquiring the predetermined biological information related to the person to be authenticated as second biological information;
(d) second measurement means for acquiring a measurement value of the person to be authenticated at a second time point when acquiring the second biological information for the predetermined aging parameter;
(e) Identifying the corresponding registrant based on the first biometric information, the second biometric information, and the measured values at the first time point and the second time point for each of the plurality of authentication registrants Identification means for
An authentication device comprising: The authentication device according to claim 8, wherein
The first acquisition means includes
(a-1) Age information recording means for causing the recording means to record information relating to age of each of the plurality of authentication registrants in association with the first biological information;
Have
The identification means includes
(e-1) means for identifying the corresponding registrant based on the age information;
An authentication apparatus comprising: The authentication device according to claim 9,
The age information recording means includes
Means for recording the age information and the first biological information as a set on the same recording medium;
An authentication apparatus comprising: The authentication device according to any one of claims 8 to 10,
The first acquisition means includes
(a-2) characteristic information recording means for causing the recording means to record information on change characteristics relating to the predetermined aging parameter in association with the first biological information;
Have
The identification means includes
(e-2) means for identifying the corresponding registrant based on the information of the change characteristic;
An authentication apparatus comprising: The authentication device according to claim 11,
The characteristic information recording means includes:
Means for recording the change characteristic information and the first biological information as a set on the same recording medium;
An authentication apparatus comprising: The authentication device according to any one of claims 8 to 12,
The predetermined aging parameter is a parameter whose measured value increases with age,
The identification means includes
(e-3) If the measured value of the authentication subject acquired by the second measuring means is smaller than the measured value of the authentication registrant acquired by the first measuring means, the authentication registrant and the authentication Means to deny identity with the subject,
An authentication apparatus comprising: The authentication device according to any one of claims 8 to 13,
The first acquisition means includes
(a-3) time information recording means for causing the recording means to record time information relating to each first time point in association with the first biological information;
Have
The identification means includes
(e-4) means for identifying the corresponding registrant based on time information recorded in the recording means and time information relating to the second time point;
An authentication apparatus comprising: The authentication device according to any one of claims 1 to 14,
The authentication apparatus according to claim 1, wherein the predetermined biological information is information relating to facial features.

Images