JP2009245231A - Individual authentication method and individual authentication system using bending and stretching movement measurement of finger - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a new biometric individual authentication method and biometric individual authentication system, having pretense resistance. <P>SOLUTION: In this individual authentication method, as information unique to a user, attention is paid to a bending and stretching movement of a finger from many human movements. A joint of the human finger is a hinge joint of one degree of freedom, and further, a tendon is present between the first joint and the second joint, and both movements are linked, so that reproducibility of the bending and stretching movement of the finger is extremely high. A time-series change of a silhouette shape in the bending and stretching movement of the finger is defined as a characteristic amount by use of a concept of curvature, and authentication information is generated from the characteristic amount. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  The present invention relates to a personal authentication technique, and more particularly to a biometric authentication technique having impersonation resistance.

  Conventionally, various studies have been made on personal authentication techniques, and in recent years, biometric authentication using biometric features unique to the person as authentication information has been adopted in many situations. However, recently, a problem of “spoofing” by a malicious third party who artificially created a model finger or the like based on a user's residual fingerprint has been pointed out. With regard to this point, focusing on human movements that are difficult to steal compared to still two-dimensional image information such as fingerprints, the use of feature quantities extracted from the movements as authentication information has been studied. For example, research has been conducted to generate authentication information for identifying an individual from a feature amount of human movement related to walking, keystrokes, signatures, and the like.

However, all of these movements have low reproducibility, and it has been said that it is difficult for a system using these as authentication information to achieve high authentication accuracy. For example, Japanese Laid-Open Patent Publication No. 2002-142255 (Patent Document 1) discloses a biometric authentication system that performs personal authentication using lip movement. However, what kind of lip movement is used as authentication information? It is not disclosed in detail, and it cannot be inferred whether the reproducibility of the lip movement is sufficient for use as authentication information.
JP 2002-142255 A

  The present invention has been made in view of the above problems in the prior art, and an object of the present invention is to provide a novel biometric personal authentication method and biometric personal authentication system having impersonation resistance.

  As a result of intensive studies on a novel biometric personal authentication configuration having impersonation resistance, the present inventor has paid attention to the bending and stretching movements of fingers among a number of human movements as information unique to the user. The present inventor, in addition to the human finger joint being a one-degree-of-freedom hinge joint, there is a tendon between the first joint and the second joint, and the movements of both are interlocked. The reproducibility of finger flexion and extension was considered extremely high. Under this concept, the time-series change of the silhouette shape in the flexion and extension movement of the finger was defined as a feature amount using the concept of curvature, and authentication information was successfully generated from this feature amount. It has come.

  That is, according to the present invention, there is provided a personal authentication method using finger flexion / extension movement measurement, which is a continuous image obtained by imaging a person's finger flexion / extension movement, and a plurality of digital images obtained by imaging the side surfaces of the finger. Obtaining an image; for each of a plurality of digital images of the side surface of the finger, extracting a contour line of the finger; obtaining a curvature profile for the extracted contour line; and calculating a bending angle of the finger from the curvature profile. And storing the curvature profile in correspondence with the bending angle of the finger, and performing authentication by checking the curvature profile of the person to be authenticated for each predetermined bending angle of the finger. A personal authentication method is provided. In the present invention, the collation is performed by obtaining a correlation coefficient for each predetermined finger bending angle for the curvature profile of the person to be authenticated, and comparing the average value of the correlation coefficient with a predetermined threshold value. It can be carried out. In the present invention, the curvature profile can be obtained by scanning the contour line. Furthermore, in the present invention, the bending angle of the finger can be obtained based on position information corresponding to the extreme value of the curvature of the contour line.

  Further, according to the present invention, there is provided a personal authentication system using finger flexion / extension movement measurement, which is a continuous image obtained by imaging a person's finger flexion / extension movement, and a plurality of digital images obtained by imaging the side surfaces of the finger. A sensor unit that acquires an image; and a personal authentication unit to which information of a digital image acquired by the sensor unit is input. The personal authentication unit includes a feature amount calculation unit, a feature amount storage unit, a collation unit, The feature amount calculation unit extracts a contour line of the finger for each of a plurality of digital images of the side surface of the finger, derives a curvature profile of the extracted contour line, and calculates the curvature of the curvature Functional means for calculating the bending angle of the finger from the profile and then defining the curvature profile as a feature quantity in association with the bending angle of the finger; and functional means for storing the feature quantity in the feature quantity storage unit , The verification unit And an individual including a functional means for performing authentication by comparing the feature quantity of the person to be authenticated defined by the feature quantity calculation unit with the feature quantity stored in the feature quantity storage unit for each predetermined finger bending angle. An authentication system is provided. In the present invention, the verification unit calculates a correlation coefficient for each predetermined finger bending angle for the curvature profile of the person to be authenticated, and compares the average value of the correlation coefficient with a predetermined threshold value. Functional means can be included. In the present invention, the feature amount calculation unit may include functional means for deriving the curvature profile by scanning a contour line of the finger. Furthermore, in the present invention, the feature amount calculation unit can include a functional unit that calculates a bending angle of the finger based on position information corresponding to an extreme value of a curvature of the contour line of the finger.

  As described above, according to the present invention, a novel biometric personal authentication method and biometric personal authentication system having impersonation resistance are provided.

  Hereinafter, the present invention will be described with reference to embodiments shown in the drawings, but the present invention is not limited to the embodiments shown in the drawings.

  FIG. 1 shows a biometric authentication system 10 (hereinafter referred to as an authentication system 10) according to an embodiment of the present invention. The authentication system 10 according to the present embodiment includes a sensor unit 12 and a personal authentication unit 14. The sensor unit 12 is configured such that a finger 18 (preferably an index finger) of the user 16 who is the person to be authenticated is inserted, and the bending / extending motion of the finger 18 can be imaged. The personal authentication unit 14 includes a CPU that functions as a system controller, a RAM that provides an execution space for application software, a ROM that stores data for processing, a job log, or a program, and a storage such as a hard disk. Comprising a personal computer with a device, using MacOS ™, Windows ™, UNIX ™, LINUX ™ or any other suitable operating system, C, By storing and executing an application program written in C ++, Visual C ++, Visual Basic, Java (registered trademark), etc., personal authentication of the user 16 is performed based on the image of the finger 18 captured by the sensor unit 12. To provide a functional means of Utame. Moreover, in this embodiment, the authentication result derived | led-out by the personal authentication part 14 can be displayed on the authentication result display part 20, and the user 16 can know an authentication result by visually recognizing this. Although not shown in FIG. 1, the authentication result derived by the personal authentication unit 14 may be output to a desired control means (for example, door lock opening / closing control means).

  FIG. 2 is a diagram showing the sensor unit 12 in the present embodiment, FIG. 2 (a) shows a side view thereof, and FIG. 2 (b) shows a top view thereof. As shown in FIG. 2A, the sensor unit 12 includes a housing 30 as an imaging space. An opening 32 for inserting the finger 18 is formed in the housing 30. When performing personal authentication, the user 16 moves the finger 18 from the opening 32 with the belly of the finger 18 facing down. Insert into the housing 30. The inside of the housing 30 has sufficient space for inserting the finger 18 to at least the second joint (proximal interphalangeal joint) and bending and extending and bending the first and second joints of the finger 18. Yes. In FIG. 2, for convenience of explanation, the finger 18 inserted into the housing 30 is shown through.

  Further, as shown in FIG. 2B, the sensor unit 12 in the present embodiment is configured to include an imaging unit 34 that is positioned so as to be capable of imaging the side surface of the housing 30. The imaging means 34 in the present invention may be any means that can continuously take images, convert them into digital signals and output them, and a CCD video camera can be used. 2B shows a mode in which the image pickup unit 34 is positioned and arranged on the side of the housing 30. However, the present invention is not limited to this, and it is more compact by using an optical system including a reflection mirror. It can also be configured.

  In addition, the side surface 30a of the housing 30 is made of a light-transmitting material in view of imaging by the imaging means 20, and the other side surface 30b takes into account the contrast in relation to the color of the finger 18. And functions as a background in the imaging of the finger 18. In the present embodiment, it may be configured to detect that the user 16 has inserted the finger 18 into the housing 30 with a sensor or the like and automatically start imaging of the side surface side of the housing 30.

  In use, the user 16 bends and stretches the finger 18 with the finger 18 inserted into the housing 30. As a result, the imaging unit 20 continuously images the bending / extending motion of the finger 18 and is acquired as a plurality of two-dimensional image information. FIG. 3 shows a continuous image of the finger 18 of the user 16 imaged in time series by the imaging means 20. As shown in FIG. 3, it will be understood that the silhouette shape of the finger 18 as seen from its side changes with its bending and stretching movement. Since this time-series change has high reproducibility in terms of the structure of the finger joint, in the present invention, the change is extracted as a feature amount and used as authentication information unique to the user 16.

  Continuous two-dimensional image information obtained by imaging from the side the bending and stretching motion of the finger 18 acquired by the sensor unit 12 in the above-described procedure is converted into a digital signal and input to the personal authentication unit 14. In the personal authentication unit 14, the feature amount is calculated from the two-dimensional image information of the finger 18, and based on this, unique authentication information for the user 16 is generated. Hereinafter, the process of calculating the feature amount in the present embodiment will be described with reference to FIGS.

  In the calculation of the feature amount in the present embodiment, first, predetermined image processing is performed on the digital image acquired from the sensor unit 12. This point will be described below with reference to FIG. FIG. 4 shows in a time series manner how image processing is performed on one of the continuous two-dimensional images of the finger 18 acquired from the sensor unit 12. In the present embodiment, first, the image obtained by the sensor unit 12 shown in FIG. 4A is binary with the background pixel being black (luminance value 0) and the finger portion being white (luminance value 255). Binarization processing is performed. FIG. 4B shows an image after binarization processing. Next, in the image after the binarization process shown in FIG. 4B, an edge pixel extraction process is performed to extract an edge pixel that becomes the outline of the finger 18. Edge pixel extraction can be performed using a Laplacian filter or the like as appropriate. FIG. 4C shows an image after the edge pixel extraction process. In the image after the edge pixel extraction processing shown in FIG. 4C, the contour line on the side surface of the finger 18 is configured as a thin line having a width of one pixel.

  In the generation of the authentication information in the present embodiment, the curvature is sequentially calculated along the contour line of the side surface of the finger 18 shown in FIG. Hereinafter, this curvature calculation process will be described with reference to FIGS. 5 and 6.

  FIG. 5 is a diagram in which the diagram shown in FIG. 4C is reversed in black and white for convenience of explanation. In the present embodiment, the contour line E shown in FIG. 5 starts from the point “START” shown in the drawing and reaches the point “END” shown in the drawing, as indicated by the broken line arrow in the drawing, The curvature is sequentially calculated by scanning the contour line E. Here, the curvature means an amount representing the degree of bending of the curve. In the case of two dimensions, the curvature is given by the reciprocal (1 / R) of the circumscribed circle radius R of the triangle formed by any three points on the curve, The greater the degree of bending, the larger the curve.

  6A is an enlarged view of a portion surrounded by a broken-line circle in FIG. 5, and FIG. 6B is an enlarged view of a portion surrounded by a square solid line in FIG. 6A. FIG. In the present embodiment, as shown in FIG. 6B, for the circumscribed circle S of the triangle T formed by three points (I, II, III) existing on the contour E, the reciprocal of the radius R (1 / R) is defined as the curvature. Each of the three points (I, II, III) on the contour line E shown in FIG. 6B is defined as one pixel (one pixel) constituting the contour line E. While the distance d1 between the pixels (II) and the distance d2 between the pixels (II) and (III) are fixed, the point (I) on the contour E is moved one pixel at a time. Each time, the coordinates of the above-mentioned three points (I, II, III) are obtained, and the curvature (1 / R) of the position is calculated from these values. In the present embodiment, the integrated value of the pixels that have moved with the calculation of the curvature is defined as the number of tracking pixels.

  FIG. 7 is a diagram for explaining the correlation between the curvature and the position on the contour line E. In FIG. FIG. 7A shows the relationship between the curvature of the contour E calculated by the above-described procedure and the number of tracking pixels. This relationship defines the curvature profile of the contour line E on the side of the finger 18. As indicated by circles 1 to 6 in the figure, the relationship between the curvature of the contour line E and the number of tracking pixels indicates six extreme values (maximum value / minimum value). Here, it should be understood that the extreme values indicated by reference numerals 1 to 6 correspond to the curvatures of the positions indicated by reference numerals 1 to 6 surrounded by circles on the contour line E in FIG. 7B, reference numerals 1 and 6 related to the second joint of the finger, reference numerals 2 and 5 related to the first joint of the finger, and reference numeral 3 corresponding to the base of the nail. And the curvature of the part 4 corresponding to the tip of the finger always show extreme values (maximum value / minimum value). Here, since the curvature is always the minimum value in the portion 4 corresponding to the tip of the finger, the position where the calculated curvature shows the minimum value is specified as the tip of the finger, and from the position of the tip of the finger The positions of the reference numerals 1 to 6 can be specified a priori.

  What should be noted here is that the above-described “curvature profile” is unique to the user and changes with the bending and stretching motion of the finger 18. That is, even for the finger of the same person, the “curvature profile” at that time varies depending on how the fingers are bent. This is because the authentication system 10 of the present embodiment acquires “curvature profile” for the number of frames of continuous images for one user at the time of authentication, that is, for one user, It means that a large amount of unique authentication information is acquired in a short time. The present invention pays attention to the change of the “curvature profile” in the flexion and extension movement of the finger, and defines this change, that is, the total of a plurality of “curvature profiles” as the feature amount.

  This is a significant difference from conventional authentication systems that use fingerprints or vein static two-dimensional images. For example, consider a case where a static two-dimensional image of the side of a user's finger is stolen. Even if the authentication system 10 according to the present embodiment is attempted to impersonate with such a static two-dimensional image, the authentication system 10 is not limited to information on time-series changes accompanying the flexion and extension movements of the finger, that is, a plurality of information. Such a spoofing is not successful because it requires the user to have a “curvature profile”. Here, if the reproducibility of the finger flexion / extension motion is low, it is not established as an authentication method before the problem of impersonation strength, but the human finger joint is a hinge joint with one degree of freedom. Since the movements of the joint and the second joint are linked, the reproducibility of the bending / extending movement of the finger is extremely high, so that the authentication method of the present invention functions effectively.

  Here, among the “curvature profiles” corresponding to the number of frames of the above-described continuous image, which one is adopted to generate the authentication information is determined by comparing the processing cost of the information processing apparatus and the authentication accuracy. Can do. In this embodiment, a “curvature profile” at a predetermined “finger bending angle” is extracted, and authentication information is generated from the plurality of “curvature profiles”. In the present embodiment, the “finger bending angle” described above is defined as θ shown in FIG. That is, in the present embodiment, the vertical bisector L1 passing through the midpoint M of the line segment connecting the pixel O at the position of reference numeral 1 and the pixel P at the position of reference numeral 6 related to the second joint of the finger. Is the reference line, and the angle θ formed by the line segments L2 and L1 connecting the midpoint M and the pixel Q at the position of the reference numeral 4 corresponding to the tip of the finger is defined as “finger bending angle”. The method for obtaining the positions of the reference numerals 1, 4, and 6 has already been described above.

  The feature amount calculation procedure in the present embodiment has been described above with reference to FIGS. 4 to 8. Subsequently, the algorithm for personal authentication in the authentication system 10 of the present embodiment will be described below with reference to FIG. To do.

  FIG. 9 shows a functional block diagram of the authentication system 10 of the present embodiment. The personal authentication unit 14 described above with reference to FIG. 1 includes a feature amount calculation unit 40, a feature amount storage unit 42, and a collation unit 46. When using the authentication system 10 and registering as a user, the feature quantity calculation unit 40 is configured to obtain a series of digital images acquired from the sensor unit 12 according to the procedures described above with reference to FIGS. For all, a “curvature profile” and a “finger bending angle θ” corresponding to the “curvature profile” are obtained, and data associating the two are defined as feature quantities. The feature quantity is associated with the user's personal information (name, sex, age, etc.) and stored in the feature quantity storage unit 42 constituted by a storage device such as a hard disk.

  Next, the procedure of personal authentication performed by the authentication system 10 of the present embodiment will be described below with reference to FIG. First, when a user who is a person to be authenticated inserts a finger into the sensor unit 12 of the authentication system 10 and performs a bending / extending exercise, the feature amount calculation unit 40 performs an In accordance with the procedure described above with respect to 4 to 8, a “curvature profile” is derived, and the result obtained by associating this with the finger bending angle θ is output to the matching unit 46 as “inquiry data”.

  The collation unit 46 extracts a “curvature profile” corresponding to a predetermined finger bending angle θ from the “inquiry data” input from the feature amount calculation unit 40. Here, for convenience of explanation, “curvature profiles” corresponding to seven angles of finger bending angles θ = 15 °, 20 °, 25 °, 30 °, 35 °, 40 °, and 45 ° are extracted. A case will be described as an example.

  Next, the collation unit 46 reads the “registration data” stored in the feature amount storage unit 42 for each registered user, and among them, the finger bending angle θ = 15 °, 20 °, 25 °, 30 °, Extract “curvature profile” corresponding to 35 °, 40 ° and 45 °.

  Thereafter, the matching unit 46 extracts the “curvature profile” extracted from the “query data” input from the feature amount calculation unit 40 and the “curvature” extracted from each “registered data” stored in the feature amount storage unit 42. The correlation coefficient with the “profile of” is calculated for each finger bending angle θ. At this time, in order to match the number of data used for calculating the correlation coefficient, it is preferable to set a predetermined position indicating the extreme value of the curvature on the contour line E as a reference point and to align the number of data to be compared. For example, with reference numeral 5 (position inside the first joint) shown in FIG. 7B as a reference point, X [Pixel] and “END” are traced back to the “START” side shown in FIG. "Curvature profile" in the range of the number of tracking pixels of Y [Pixel] and (X + Y) [Pixel] down to the "" side is set as a correlation coefficient calculation target.

  Finally, the correlation coefficients of the “curvature profile” at the finger bending angles θ = 15 °, 20 °, 25 °, 30 °, 35 °, 40 °, 45 ° calculated in this way are as follows. Find the average value of.

  The collation unit 46 determines whether or not the average value of the seven correlation coefficients obtained in the above-described procedure is equal to or greater than a predetermined threshold, reads “registered data” from which the average value equal to or greater than the threshold is derived, It is determined that the person to be authenticated is a user of the “registration data”. On the other hand, when the average value of the seven correlation coefficients does not exceed the predetermined threshold, the collation unit 46 determines that the authentication of the person to be authenticated has failed. The above-described determination result is output from the verification unit 46 to the authentication result display unit 20 as shown in FIG. Note that the above-described threshold value can be appropriately determined in view of the required accuracy of authentication and the like.

  Hereinafter, the biometric personal authentication method of the present invention will be described more specifically using examples, but the present invention is not limited to the examples described later.

  The authentication system 10 described above with reference to FIG. 1 was constructed, and an operation experiment of this system was performed with the cooperation of four subjects (A, B, C, D). First, for each of four subjects (A, B, C, D), a “curvature profile” is calculated from a continuous image of bending and stretching motions of the index finger side surface imaged using the authentication system 10, Recorded as "registration data" of people. “Registration data” includes seven “finger bending angles θ (15 °, 20 °, 25 °, 30 °, 35 °, 40 °, 45 °)” and “curvature profile” for each angle. Was calculated and recorded as “registration data”.

  Next, using the authentication system 10, one person and three persons other than the person collated one “registration data”. In addition, it collated several times (8-10 times) about the combination of one "registration data" and a test subject. In this example, for “curvature profiles” obtained for seven “finger bending angles θ (15 °, 20 °, 25 °, 30 °, 35 °, 40 °, 45 °)”, Correlation coefficients with “registered data” were calculated for each angle, and an average value of these seven correlation coefficients was used as a verification target. In this embodiment, the person is determined to be the person when the average value of the correlation coefficient is equal to or greater than a predetermined value. In this embodiment, the person collation criterion is “0.967 or more”.

  The experimental results are shown in Tables 1 to 4 below. Table 1 below shows the results of verification by three persons other than the principal and the person against the “registration data” of the subject A, and Table 2 below shows the results of comparison between the subject and the person other than the principal and the person against the “registration data” of the subject B Table 3 below shows the results of verification by three persons, and Table 3 below shows the results of verification by three persons other than the principal and the person against the “registration data” of subject C. The results obtained by comparing the “data” with three persons other than the principal and the principal are shown. In the following Tables 1 to 4, the calculated correlation coefficients are shown for each “bending angle θ of the finger” (15 ° to 45 °), and the average value (Avr.) Is shown at the bottom.

As shown in Tables 1 to 4 above, the authentication system of the present embodiment is a result of determining the person 31 times in 35 trials when the person collates the subject's own “registration data”. It became. Specifically, for subject A (see Table 1 (I)) and subject B (see Table 2 (I)), all of the average values of the correlation coefficients exceeded the personal verification criterion “0.967”. Subject A and subject B could be determined as the principals. On the other hand, for subject C, the second and third times (Table 3 (I)
For the subject D, the average value of the correlation coefficient was lower than the personal verification criterion “0.967” at the 8th and 9th times (see Table 4 (I)), and the verification by the principal was rejected (the principal) Rejection rate = 4/35 = 11.4%).

On the other hand, in the authentication system of the present embodiment, when another person other than the registrant collated, the other person was not accepted once in a total of 117 trials. Specifically, when subjects B, C, and D collate against “registration data” of subject A (see Tables 1 (II) to (IV)), for subject B “registration data” When subjects A, C, and D collate (Table 2 (II)
(See (IV)), when subjects A, B and D collate against “registration data” of subject C (see Tables 3 (II) to (IV)), and “registration data” of subject D When subjects A, B, and C collate (Table 4 (II)
In all cases (see (IV)), all of the average values of the correlation coefficients were below the personal verification standard of “0.967”.

  As described above, according to the present invention, a novel biometric personal authentication method and biometric personal authentication system having impersonation resistance are provided. The present invention is expected to contribute to the improvement of impersonation resistance.

The figure which shows the biometric authentication system of this embodiment. The figure which shows the sensor part in this embodiment. The figure which shows the continuous image of the user's finger | toe imaged in time series. The figure which shows the image processing performed about the two-dimensional image of a finger in time series. FIG. 5 is a diagram showing the diagram shown in FIG. The conceptual diagram for demonstrating the calculation method of a curvature. The figure which shows the correspondence of a curvature and the position on a thin line. The figure which shows the definition of the bending angle (theta) of a finger | toe. The functional block diagram of the biometric authentication system of this embodiment.

Explanation of symbols

DESCRIPTION OF SYMBOLS 10 ... Biometric authentication system, 12 ... Sensor part, 14 ... Personal authentication part, 16 ... User, 18 ... Finger, 20 ... Authentication result display part, 30 ... Case, 32 ... Opening part, 34 ... Imaging means, 40 ... Feature amount calculation unit, 42 ... feature amount storage unit, 46 ... collation unit,

Claims (8)

A personal authentication method using finger flexion and extension measurement,
Obtaining a plurality of digital images obtained by imaging a side surface of the finger, which is a continuous image obtained by imaging the bending motion of the finger of the person to be authenticated;
For each of a plurality of digital images on the side of the finger,
Extracting the outline of the finger;
Obtain a curvature profile for the extracted contour line,
After obtaining the bending angle of the finger from the curvature profile,
Storing the curvature profile in association with the bending angle of the finger;
A personal authentication method, wherein authentication is performed by comparing the curvature profile of the person to be authenticated for each bending angle of a predetermined finger.   The verification is performed by obtaining a correlation coefficient for each predetermined finger bending angle with respect to the curvature profile of the person to be authenticated, and comparing an average value of the correlation coefficient with a predetermined threshold value. 1. The personal authentication method according to 1.   The personal authentication method according to claim 1, wherein the curvature profile is obtained by scanning the contour line.   The personal authentication method according to claim 1, wherein the bending angle of the finger is obtained based on position information corresponding to an extreme value of the curvature of the contour line. It is a personal authentication system using finger flexion and extension movement measurement,
A sensor unit that captures a plurality of digital images obtained by imaging the side surface of the finger, which is a continuous image obtained by imaging the flexion and extension motion of the finger of the person to be authenticated;
A personal authentication unit to which information of a digital image acquired by the sensor unit is input,
The personal authentication unit
A feature amount calculation unit, a feature amount storage unit, and a matching unit;
The feature amount calculation unit includes:
For each of a plurality of digital images on the side of the finger,
Extracting the outline of the finger;
Deriving the curvature profile of the extracted contour line,
After calculating the bending angle of the finger from the curvature profile,
Functional means for defining the curvature profile as a feature value in association with the bending angle of the finger;
Functional means for storing the feature quantity in a feature quantity storage unit,
The collation unit
Personal authentication including functional means for performing authentication by collating the feature quantity of the person to be authenticated defined by the feature quantity calculation unit and the feature quantity stored in the feature quantity storage unit for each bending angle of a predetermined finger system.   The verification unit includes a functional unit that calculates a correlation coefficient for each bending angle of a predetermined finger for the curvature profile of the person to be authenticated, and compares an average value of the correlation coefficient with a predetermined threshold value. The personal authentication system according to claim 5.   The personal authentication system according to claim 5, wherein the feature amount calculation unit includes a function unit that derives a profile of the curvature by scanning an outline of the finger.   The said feature-value calculation part contains the function means which calculates the bending angle of the said finger | toe based on the positional information corresponding to the extreme value of the curvature of the said finger | toe outline, The said any one of Claims 5-7. Personal authentication system.

Images