JP2001021309A - Individual body authentication method and individual person authentication method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To enable always accurate individual person authentication by setting a threshold value in accordance with individual characteristic. SOLUTION: In this individual person authentication method, unevenness information of wrinkles of an identical person identical finger to be authenticated is read m-times, each error is calculated from the combination mC2 in which arbitrary two are selected, similarity is calculated from a plurality of calculated errors, a characteristic curve of an FRR (principal rejection rate) and a characteristic curve of an FAR (others rejection rate) of the person are obtained from the similarity, and a threshold value peculiar to the individual person is set. The unevenness information of a finger of an individual person to be authenticated is compared with the unevenness information of a finger of the individual person which is previously registered, and individual person authentication is performed by using the threshold value peculiar to the individual person.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

The present invention relates to an individual authentication method for performing individual authentication and an individual authentication method for performing individual authentication.

[0002]

2. Description of the Related Art As a system for identifying an individual, for example, a biometric system, that is, a system called a biometric authentication system is known. This is a personal authentication technique using the physical characteristics of an individual, and is identified by fingerprints, finger joint wrinkles, iris, retina, ears, face, voice, and the like.

For example, Japanese Patent Application Laid-Open No. H10-26088 discloses a method for performing personal authentication based on wrinkles of a finger joint. As shown in FIG. 19, a plurality of linear output electrodes 2 which are long in a direction orthogonal to the longitudinal direction of the finger 1 to be collated by the person to be authenticated are arranged at predetermined intervals along the longitudinal direction of the finger 1. A linear electrode array 3 is provided, and a single input electrode 4 having a finger whose width in the longitudinal direction is wider than the output electrode 2 is provided at an end of the linear electrode array 3 along the longitudinal direction of the finger 1.
It has.

[0004] Each switch of the analog switch group 5 is connected to each output electrode 2, and each switch is sequentially turned on and off by a timing pulse from a timing pulse generator 6, and the finger 1 is moved to the linear electrode array. The oscillator 7 supplies an oscillation frequency signal of, for example, 1 MHz from the oscillator 7 to each output electrode 2 via each switch of the analog switch group 5 while being in contact with each output electrode 2 and input electrode 4 of the input electrode 3. The voltage signal from 4 is detected by a detection circuit 8
The A / D converter 9 converts the signal into a digital signal and outputs the digital signal.

In this publication, as shown in FIG. 20, a state in which the vicinity of the third joint of the finger 1 is brought into contact with the input electrode 4 is shown in FIG.
An operation of bringing the first joint portion and the second joint portion into contact with each output electrode 2 of the linear electrode array 3 is performed. By doing so, the capacitive coupling between the finger 1 and the output electrode 2 becomes strong in the part where the finger 1 and the output electrode 2 are in close contact with each other, as shown in FIG. Then, the value of the signal V (i) in that portion increases. Conversely, the finger 1 and the output electrode 2 are not in close contact with each other, such as the side wrinkles near the first joint in the figure (b) and the side wrinkles near the second joint in the figures (c1) and (c2) in the figures. In the part, the capacitive coupling between the finger 1 and the output electrode 2 is weakened,
Since the impedance increases, the value of the signal V (i) in that portion decreases. Thus, it is possible to obtain the wrinkle unevenness information of the finger. Then, by comparing and comparing the detected wrinkle unevenness information with the previously registered wrinkle unevenness information of an individual, personal authentication as to whether or not the user is an individual can be performed.

[0006]

In this publication,
At the time of comparison / matching, the position of the wrinkle unevenness information is aligned, the matching rate is calculated by an arbitrary method, the similarity is calculated, and it is determined whether the person is an individual or another person based on whether or not the similarity exceeds a set threshold. ing. Usually, the threshold is F
AR (False Acceptance Rate) and FRR (False Acceptance Rate)
Calculate se Rejection Rate) and set it to an arbitrary constant value. That is, the threshold is determined from the average FRR and the characteristic curve of FAR calculated from many samples. Note that FAR indicates the probability that another person will be mistakenly accepted as the person, and FRR indicates the probability that the person will be rejected. Therefore, FAR and FRR have contradictory characteristics. If one is strict, the other will be soft, and the characteristics can be changed depending on the setting of the threshold. Also, when analyzing and viewing FAR and FRR for each individual,
It may be far from the characteristic curve of FAR and FRR calculated from the whole, and the threshold calculated from the whole may be an inappropriate value for each individual.

[0007] Therefore, in the above-mentioned publication, if personal verification is performed simply using a threshold calculated using a large amount of data of many people, accurate personal authentication may not be performed.

Therefore, the invention according to claim 1 provides an individual authentication method in which a threshold can be set in accordance with the characteristics of an individual, thereby always performing accurate individual authentication. The invention according to claims 2 to 6 provides a personal authentication method that can set a threshold value according to the characteristics of an individual, and thereby can always perform accurate personal authentication.

[0009]

According to the first aspect of the present invention,
A threshold unique to an individual is calculated based on the biological information of the individual, and when the individual is authenticated using the threshold, the biological information of the individual to be authenticated is read a plurality of times, and a set of arbitrary two pieces of biological information among them is read. Some errors calculated by the combination are obtained by changing the combination, similarity is calculated from the obtained errors, a statistical value is calculated from the similarity, an individual-specific threshold is set, and the threshold is authenticated. There is an individual authentication method in which an individual to be associated is associated with biometric information registered in advance, biometric information for which authentication is required is compared with biometric information registered in advance, and individual authentication is performed using a set threshold.

According to a second aspect of the present invention, a threshold unique to an individual is calculated based on the individual's biometric information, and when the individual is authenticated using this threshold, the biometric information of the individual to be authenticated is read a plurality of times. Some of the errors calculated by combining any two of the biological information are obtained by changing the combination, a similarity is calculated from the obtained errors, and a statistical value is calculated from the similarity to obtain an individual. A specific threshold is set, this threshold is associated with biometric information registered in advance for an individual to be authenticated, biometric information with an authentication request is compared with biometric information registered in advance, and personal authentication is performed using the set threshold. The personal authentication method to be performed.

According to a third aspect of the present invention, in the personal authentication method according to the second aspect, the threshold value is set to a first fixed value when a statistical value obtained from a plurality of similarities exceeds an arbitrary upper limit set in advance. Is set to a second fixed value when the value falls below an arbitrary lower limit set in advance, and is set to a value calculated according to a specific function when the value falls between the upper limit and the lower limit.

According to a fourth aspect of the present invention, in the personal authentication method according to the second or third aspect, a statistical value is calculated by reading biometric information a plurality of times when registering biometric information of an individual to be authenticated. The invention according to claim 5 is the invention according to claim 2 or 3
In the personal authentication method described above, a plurality of pieces of biometric information that can be authenticated as a person are stored, and the stored biometric information is used for setting a threshold value of a corresponding individual. Claim 6
According to the present invention, in the personal authentication method according to any one of claims 2 to 5, the threshold used in the comparison is varied in an arbitrary range based on the individual-specific threshold calculated.

[0013]

Embodiments of the present invention will be described with reference to the drawings. In this embodiment, a description will be given of an embodiment in which the present invention is applied to a personal authentication apparatus for collating and authenticating finger wrinkle unevenness information as personal biometric information.

As shown in FIGS. 1A and 1B, a plurality of electrodes are provided along the longitudinal direction of the finger 11 of the subject so as to cover the first joint 11a and the second joint 11b of the finger. Output electrode 1
An electrode array 13 arranged as 2 is provided, and an input electrode 14 which is a single electrode is arranged at a predetermined interval on the outer side of the electrode array 13 at the base of the finger. That is,
When the input electrode 14 is brought into contact with the electrode array 13 along the longitudinal direction of the finger 11, the vicinity of the base of the finger 11 comes into contact. Each output electrode 12 of the electrode array 13 is arranged, for example, at a pitch of 0.2 mm, and has a total length of 40 mm and a length sufficient to cover the first joint 11 a and the second joint 11 b of the finger 11. ing.

A driving means for selectively supplying a signal of a predetermined frequency to each output electrode 12 of the electrode array 13 while the finger 11 is in contact with the electrode array 13 and the input electrode 14 is output. An electrode drive circuit 15 is provided, and a detection circuit 16 is provided as detection means for detecting a signal from the input electrode 14 while the finger 11 is in contact with the electrode array 13 and the input electrode 14.

As shown in FIG. 2, the output electrode drive circuit 15 includes 200 2-input AND gates AND1, AND2.
AND2, AND3,..., AND200 are provided, and one input terminal of each of the AND gates AND1 to AND200 is connected to one of the input terminals of FIG.
The electrode selection signals SEL1, SEL2, SE as shown in FIG.
L3,... SEL200 are sequentially and alternately supplied, and a common carrier signal as shown in FIG. 3A is supplied to the other input terminal.

The carrier signal has a frequency of several hundred kHz to several MH.
The electrode selection signals SEL1, SEL2, and SEL2 are high-frequency signals having a frequency of about z, for example, a frequency of 2 MHz and a predetermined amplitude.
SEL3,... SEL200 are the AND gates AND1 to A
Each of the output electrodes 12 (1) at the timing of input to the ND200
12 to 200 (200). For example, as shown in FIG. 4, the electrode selection signal SEL1
Is supplied to the first output electrode 12 (1) as a drive signal at the timing when the signal is input to the AND gate AND1.

The detection circuit 16, as shown in FIG.
It comprises an amplifier 17, a detection circuit 18, an A / D converter 19, an oscillator 20, a control circuit 21, and an I / F (interface) circuit 22. After the weak signal from the input electrode 14 is amplified by the amplifier 17, the Detected by the detection circuit 18,
The detection output is converted into a digital signal by the A / D converter 19 and then supplied to the control circuit 21. The oscillator 20 supplies an operation clock to the detection circuit 18 and the A / D converter 19 to make the operation timing.

The control circuit 21 controls the I / F circuit 22 and transmits a digital signal from the A / D converter 19 to an external device 23 such as a personal computer (PC) via the I / F circuit 22. It has become.

In such a configuration, the electrode array 1
When the electrode selection signals SEL1 to SEL200 are sequentially and selectively supplied to the output electrodes 12 in a state where the finger 11 is in contact with the input electrode 3 and the input electrode 14, at a certain timing, as shown in FIG. A carrier signal flows from the location of the finger on that electrode, through the finger, to the input electrode 14. Then, impedance information on the carrier signal is extracted from the input electrode 14. The impedance information at this time includes the capacitance formed in the gap between the selected output electrode 12 and the finger 11 thereon, and the selected output electrode 12
The impedance of the finger 11 from the position of the upper finger 11 to the position of the finger 11 on the input electrode 14 and the finger 1 on the input electrode 14
It becomes a series component of the capacitance formed in the gap between the input electrode 1 and the input electrode 14.

The amplitude of the carrier signal rises and falls according to the impedance. A carrier signal from the input electrode 14 is detected by a detection circuit 16, amplified by an amplifier 17 in the detection circuit 16, then a carrier component is removed through a detection circuit 18, and is converted into a digital signal by an A / D converter 19. . Then, the external device 23 takes in the digital signal. From this digital signal, it is possible to obtain local wrinkle unevenness information of the finger 11.

FIG. 7 is a perspective view showing a state in which the finger 11 is placed on the sensor section composed of the electrode array 13 and the input electrode 14, and every time the finger 11 is placed on the sensor surface, the two-dimensional displacement of the sensor surface is changed. Pressing of the finger 11 on the sensor surface, rotation about the longitudinal direction of the finger 11 as an axis, deviation of the angle of the finger 11 in the longitudinal direction with respect to the direction in which the electrodes of the sensor section are arranged, and the like may occur, which may occur as an error. Is shown.

FIG. 8 shows a characteristic waveform of finger wrinkle unevenness information appearing in the carrier signal detected by the detection circuit 16. The signal level is greatly reduced in a portion where wrinkles are noticeable in the first joint portion and the second joint portion.

After such a signal waveform is captured as a digital signal by the external device 23, the signal is subjected to band-pass filter processing to perform digital processing for removing DC components and high-frequency components, and then the amplitude is normalized and finally personal information is obtained. Extract as wrinkle unevenness data. Then, the personal data is authenticated by comparing and comparing this data with the data of each personal data stored in advance. At the time of this comparison and collation, data alignment is performed. There are various methods for calculating an error between two data to be compared and matched. For example, after the alignment, the absolute value of the difference between the two data is simply calculated, and further, the accumulation of the square of the absolute value of the difference is calculated to calculate the error. Ask.

By the way, when personal authentication is performed using the wrinkle unevenness information perpendicular to the longitudinal direction of the finger 11, the input electrode 14
FRR (Fal
se Rejection Rate is calculated, and a threshold for collation is determined based on this, and a threshold is set for each individual.
The calculation of the FRR means that an error, that is, the inverse of the similarity is calculated from mC2 combinations of selecting arbitrary two pieces of data of the same finger taken a plurality of times, for example, the same person captured m times. I do.

The FAR (False Acceptance Rate) shown in FIG.
The characteristic curves of the FRR and the FRR are obtained by cumulatively expressing the result of the error calculation in a frequency distribution. The vertical axis indicates the probability, with the maximum value as 100%. The horizontal axis is a threshold value for discrimination, and is 8 bits and 256 from 0 to 255.
Expressed in stages. FIGS. 10 and 11 show characteristic curves of FRR and FAR of a person who was able to collect finger wrinkle unevenness information relatively similarly in a plurality of measurements. In the case of FIG. 10, the threshold value Th = 147, the FAR is 0.04%,
In the case of FIG. 11, the threshold value Th = 147 and the FAR is 3.97%.
It is.

On the other hand, FIG. 12 shows an example of a person who cannot always collect the wrinkle unevenness information of the finger in the same manner in a plurality of measurements. In this case, the threshold Th = 147 and the FA
R is as low as 0.07%, but FRR is as high as 25.5%. That is, the personal rejection rate increases.

As described above, the characteristic curve of the FRR is shown by the curve g in FIG.
As indicated by 1, it can be seen that the FRR has already become substantially 0 at a small threshold value, and the value falls sharply. In contrast,
When the error of the finger wrinkle unevenness information is large, the FRR characteristic curve gradually falls and does not drop to 0 as shown by curve g2 in FIG. Therefore, the average F
Even if a threshold for determining the characteristics of the sensor unit is determined from the characteristic curves of RR and FAR, there is a problem that the characteristic curves of FRR and FAR are different for each individual.

In the characteristic curves of FRR and FAR shown in FIG. 9, when the threshold value is small, the FRR (person rejection rate) becomes high, and in many cases, even a person cannot be accepted as a person. When the threshold value is large, the FAR (others acceptance rate) becomes high, and in many cases, another person is accepted as the person. In this case, the acceptance rate of the principal naturally increases.
If the user wants to use the application in a high security application, the threshold value needs to be set low. However, since the probability of the person being rejected increases, the usability is reduced. FIGS. 10 to 12 show that the characteristic curve of the FAR changes for each individual when the threshold value is set to the same value. For this reason, here, the wrinkle unevenness information of the finger is fetched a plurality of times for each individual, and based on this data, the error of the data is calculated for each individual, and the threshold value is determined. That is, as shown in FIG. 15, the external device 23 takes in the information of the wrinkle of the finger for each individual a plurality of times from the detection circuit 16 and filters the data by the filtering means 231 to remove the DC component and the high frequency component. After that, the data is normalized by the amplitude and normalization means 232 and stored in the dictionary data storage unit 233 as dictionary data. At the time of comparison and collation, the amplitude and the normalizing means 23 are used.
The data standardized in step 2 and the data in the dictionary data storage unit 233 are aligned by the alignment unit 234, and then compared and collated by the collation / discrimination unit 235 to perform personal authentication determination.

The horizontal axis in the characteristic curves of FAR and FRR can also be interpreted as an error between data, meaning that data is similar and the error is small as going to the left. The characteristic curves of FAR and FRR are obtained by expressing the result of error calculation of each data in a frequency distribution and expressing the results in a cumulative manner. The difference between the characteristic curves of FAR in FIGS. 10 to 12 is the difference in error between the data for each individual. Will be represented.

In the FRR characteristic curve of FIG. 13, the curve g1 has a higher autocorrelation value and a smaller error than the curve g2. The autocorrelation value indicates an average value of errors calculated by extracting two pieces of information from a plurality of pieces of wrinkle information of the self. Therefore, in the case of a person having a high autocorrelation value, the threshold used for discrimination is reduced, and when the error is smaller than a certain value, the threshold is fixed to a limit value, for example, 0.3. That is, if the error of the data captured a plurality of times for the same person and the same finger is small and the similarity is high, the threshold value of the individual is set low.

When the maximum value of the autocorrelation value is indicated by 100, the threshold value is uniformly set to 0.3 for a person whose autocorrelation value exceeds an arbitrary value, for example, 85. In addition, a person whose autocorrelation value is below an arbitrary value, for example, 60, sets the threshold value to 0.7 uniformly. Other persons whose autocorrelation values are in the range of 60 to 85 set the threshold according to a specific function. FIG. 14 shows the setting contents. Note that the functions need not necessarily be continuous. In addition, although a straight-line approximation is shown here, a higher-order curve may be used, or a jump threshold may be set according to the value of the autocorrelation value.

As described above, the threshold value for determination is set to F
By setting according to RR, the characteristics of the sensor unit can be controlled more strictly. Therefore, the wrinkle unevenness information of the finger is
For example, a person who can easily change the position of the finger with respect to the sensor portion, a person who cannot stably place the finger at the same position, a person who cannot properly adjust the pressing force of the finger on the sensor surface, etc.
By setting a large threshold value for people of the type whose characteristic curve of R does not rise sharply, it is possible to set such that even if FAR is slightly sacrificed, the FRR does not become extremely large.

For those who can stably acquire the same finger wrinkle unevenness information every time, the characteristic curve of FRR falls sharply, so the threshold value is set small. As a result, the region where the characteristic curves of FRR and FAR cross can be reduced, and the device can be used under conditions of high performance.

In this embodiment, the wrinkle unevenness information of the finger of each individual is previously taken in plural times and stored in the dictionary data storage unit as data for collation, and a threshold value is set based on the stored data. Then, the data of the wrinkle unevenness information of the finger taken in at the time of authentication is compared with the data stored in the dictionary data storage unit, and personal authentication is performed using a preset threshold value, but this is not necessarily limited to this. Not something.

For example, every time the individual is authenticated, the data of the unevenness information of the wrinkles of the finger of the individual is stored in the dictionary data storage unit, and the similarity and FRR are calculated each time and a new threshold value is set. However, the new threshold may be used for the next personal authentication. In this way, every time personal authentication is performed, the threshold value of the person can be determined.
Also, the data of the finger wrinkle unevenness information of an arbitrary number of times for each individual who has recently been authenticated is stored in the dictionary data storage unit, the similarity and the FRR are calculated, and a new threshold value to be used next time is set. Is also good.

Further, the user may be allowed to adjust the threshold in an arbitrary range based on the threshold of the individual for whom the threshold has been obtained. For example, a user setting screen is provided on the display device of the external device 23, and “LOW” as shown in FIG.
A “MID”, “HIGH” collation threshold selection screen is displayed, and the user operates the screen with a mouse or the like to “LOW”,
One of “MID” and “HIGH” is selected and a threshold is selected. In this case, "MID" sets an original threshold obtained in advance, "LOW" sets a threshold obtained by adding a certain value to the original threshold, and "HIGH" sets a threshold obtained by subtracting a certain value from the original threshold. Set. Adding a constant value to the original threshold value lowers the collation level, and conversely, subtracting a constant value from the original threshold value increases the collation level.

For example, FRR, FA as shown in FIG.
In the case of an individual having an R characteristic curve, the MID threshold Th
Is set to “112”, but this threshold value can be slightly reduced or increased by the user's selection operation. For example, if "HIGH" is selected, the threshold value becomes smaller and F
The RR is larger, which increases security.

Further, as shown in FIG.
In the case of an individual having the characteristic curve of, the threshold value Th of the MID is set to “200”, and this threshold value can be slightly reduced or increased by the user's selection operation. As described above, since the user can adjust the obtained threshold value in an arbitrary range, it is possible to set whether the user slightly weakens or hardens personal authentication. In the above-described embodiment, the wrinkle unevenness information of the finger is used as the biometric information of the individual. However, the present invention is not limited to this, and may be a fingerprint of the finger.

[0040]

According to the first aspect of the present invention, a threshold can be set in accordance with the characteristics of an individual, thereby providing an individual authentication method capable of always performing accurate individual authentication. Further, according to the second to sixth aspects of the present invention, the threshold can be set in accordance with the characteristics of the individual, thereby providing a personal authentication method capable of always performing accurate personal authentication.

[Brief description of the drawings]

FIG. 1 is a block diagram including a sensor unit according to an embodiment of the present invention.

FIG. 2 is a circuit configuration diagram of an output electrode drive circuit in the embodiment.

FIG. 3 is a signal waveform diagram showing an input signal from an output electrode drive circuit in the embodiment.

FIG. 4 is a signal waveform diagram showing an example of an input signal from an output electrode drive circuit in the embodiment.

FIG. 5 is a block diagram illustrating a configuration of a detection circuit in the embodiment.

FIG. 6 is a diagram for explaining contact of a finger with a sensor unit and a signal flow in the embodiment.

FIG. 7 is a view for explaining various states when a finger is placed on the sensor unit in the embodiment.

FIG. 8 is a view showing a characteristic waveform of finger wrinkle unevenness information according to the embodiment;

FIG. 9 is a diagram for explaining FAR and FRR.

FIG. 10 shows an individual FAR and FRR in the embodiment.
The figure which shows the characteristic curve example of.

FIG. 11 shows the FAR and FRR of the individual in the embodiment.
The figure which shows the characteristic curve example of.

FIG. 12 shows an individual FAR and FRR according to the embodiment.
The figure which shows the characteristic curve example of.

FIG. 13 is a characteristic curve of FRR in the embodiment;
The figure which shows the relationship between an autocorrelation value and a threshold value.

FIG. 14 is a diagram showing a setting example of a threshold according to the embodiment;

FIG. 15 is a functional block diagram of an external device according to the embodiment.

FIG. 16 is a diagram showing an example of a user setting screen according to another embodiment of the present invention.

FIG. 17 shows an individual FAR and FRR according to the embodiment.
FIG. 6 is a diagram showing a relationship between an example of the characteristic curve of FIG.

FIG. 18 is an individual FAR and FRR according to the embodiment.
FIG. 6 is a diagram showing a relationship between an example of the characteristic curve of FIG.

FIG. 19 is a block diagram showing a conventional example.

FIG. 20 is a view for explaining the finger unevenness information detecting operation of the conventional example.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 11 ... Finger 12 ... Output electrode 14 ... Input electrode 16 ... Detection circuit 23 ... External device

Continued on the front page (72) Inventor Kenichi Ide 70, Yanagicho, Saiwai-ku, Kawasaki-shi, Kanagawa F-term in Toshiba Yanagimachi Plant (reference) 2E250 BB00 DD08 2F063 AA50 BA29 BD05 BD06 BD20 CA40 HA04 HA09 KA02 KA05 LA03 LA06 LA11 LA14 LA18 LA19 LA23 LA29 5B043 AA09 BA02 GA01

Claims (6)

[Claims] 1. A threshold unique to an individual is calculated based on the biological information of the individual, and when the individual is authenticated using the threshold, the biological information of the individual to be authenticated is read a plurality of times, and any two of the two are read out. Some errors calculated by combining the two pieces of biological information are obtained by changing the combination, similarity is calculated from the obtained errors, a statistical value is calculated from the similarity, and a threshold value unique to the individual is set. The threshold value is associated with biometric information registered in advance for an individual to be authenticated, biometric information with an authentication request is compared with biometric information registered in advance, and individual authentication is performed using the set threshold value. Individual authentication method. 2. Calculating a threshold peculiar to an individual based on the biometric information of the individual, and when authenticating the individual using the threshold, reading the biometric information of the individual to be authenticated a plurality of times, and selecting any two of them. Some errors calculated by combining the two pieces of biological information are obtained by changing the combination, similarity is calculated from these obtained errors, a statistical value is calculated from the similarity, and a threshold unique to an individual is set. The threshold value is associated with biometric information registered in advance for an individual to be authenticated, biometric information having an authentication request is compared with biometric information registered in advance, and personal authentication is performed using the set threshold value. Personal authentication method. 3. The threshold value is set to a first fixed value when a statistical value obtained from a plurality of similarities exceeds an arbitrary upper limit set in advance, and is set to a first fixed value when the statistical value falls below an arbitrary lower limit set in advance. 3. The personal authentication method according to claim 2, wherein the value is set to a fixed value, and when the value is between the upper limit value and the lower limit value, the value is calculated according to a specific function. 4. The personal authentication method according to claim 2, wherein the statistical value is calculated by reading biometric information a plurality of times when registering biometric information of an individual to be authenticated. 5. The personal authentication method according to claim 2, wherein a plurality of pieces of biometric information that can be authenticated as the individual are stored, and the stored biometric information is used for setting a threshold value of the corresponding individual. . 6. The personal authentication method according to claim 2, wherein the threshold value used in the comparison can be changed within an arbitrary range based on the calculated individual-specific threshold value.