JP2012194849A - Discrimination device and discrimination method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To speed up discrimination processing while maintaining precision of the discrimination processing.SOLUTION: A discrimination device includes a photography part which photographs a subject; a storage part which associatively stores a specific photographic parameter and a feature quantity of a subject as a person to be registered when the photography part photographs the subject as the person to be registered; and a discrimination part which discriminates a person to be discriminated by collating a feature quantity of the subject as the person to be discriminated that the photography part photographs with the feature quantity of the subject as the person to be registered stored in the storage part. The photography part photographs the subject as the person to be registered multiple times while changing photographic parameters when photographing the subject as the person to be registered, and photographs the subject as the person to be discriminated using the specific photographic parameter when the photographing the subject as the person to be discriminated.

Description

  The present invention relates to an identification device and an identification method.

Authentication devices that perform authentication based on finger vein images have been developed. In obtaining a vein image, a method for obtaining an appropriate image has been considered. Patent Document 1 discloses an authentication device having gain changing means for changing the amplification factor of an image pickup element in an image pickup unit.

JP 2006-155575 A

In the authentication device described in Patent Document 1, the imaging parameters (exposure, gain, etc.) are adjusted every time imaging is performed in order to obtain an image necessary for authentication. Therefore, it may take time until an image with appropriate brightness is captured. For this reason, while the vein authentication apparatus performs the authentication process, the person to be authenticated must unnaturally hold his / her finger over the apparatus, resulting in a problem that convenience is low. That is, there is a demand for speeding up the authentication device while maintaining the accuracy of the authentication process. Here, the authentication identifies whether or not the authentication (identification) target person is a registrant by collating the registered image and the like with the image obtained at the time of authentication, and based on the identification result, For example, controlling a control target such as an electronic lock. Therefore, since the accuracy and speed of authentication depend on the accuracy and speed of identification, it is desirable that the authentication can be speeded up while maintaining the accuracy of the identification process.

The present invention has been made in view of such circumstances, and an object thereof is to speed up the identification process while maintaining the accuracy of the identification process.

The main invention for achieving the above object is:
A shooting section for shooting the subject;
A storage unit for storing a specific shooting parameter and a feature amount of the subject of the registration target in association with each other when the shooting unit captures the subject of the registration target;
An identification unit for identifying the identification target person by comparing the feature quantity of the subject of the identification target person photographed by the imaging unit with the feature quantity of the subject of the registration target person stored in the storage unit; ,
Including
In the case of photographing the subject of the registration target person, the photographing unit photographs the subject of the registration target person a plurality of times by changing photographing parameters, and when photographing the subject of the identification target person. Photographing the subject of the identification subject using the specific photographing parameters;
It is an identification device.

  Other features of the present invention will become apparent from the description of the present specification and the accompanying drawings.

FIG. 1A is a block diagram of the vein authentication device 1 in the first embodiment, and FIG. 1B is a conceptual diagram of the vein authentication device 1 in the first embodiment. It is a flowchart explaining a registration process. It is a flowchart explaining a feature-value extraction process. It is a figure which shows an example of a feature point. It is a figure which shows an example of the obtained brightness | luminance gradient. It is a figure which shows an example of the histogram of a brightness | luminance gradient. FIG. 7A is a diagram illustrating an example of a luminance gradient when the coordinate axis is aligned with the reference direction, and FIG. 7B is a diagram illustrating an example of a vector obtained based on the luminance gradient of FIG. 7A. It is a 1st explanatory view of the exposure time selected. It is a 2nd explanatory view of the exposure time selected. It is a table | surface which shows the example of the database of a feature-value. It is a flowchart explaining an authentication process. It is a figure for demonstrating the concept of an authentication process. FIG. 13A is a diagram illustrating an example of information registered in the second embodiment, and FIG. 13B is a table illustrating an example of a feature amount database in the second embodiment. It is a flowchart explaining the authentication process in 2nd Embodiment.

At least the following matters will become clear from the description of the present specification and the accompanying drawings.
A shooting section for shooting the subject;
A storage unit for storing a specific shooting parameter and a feature amount of the subject of the registration target in association with each other when the shooting unit captures the subject of the registration target;
An identification unit for identifying the identification target person by comparing the feature quantity of the subject of the identification target person photographed by the imaging unit with the feature quantity of the subject of the registration target person stored in the storage unit; ,
Including
In the case of photographing the subject of the registration target person, the photographing unit photographs the subject of the registration target person a plurality of times by changing photographing parameters, and when photographing the subject of the identification target person. Photographing the subject of the identification subject using the specific photographing parameters;
Identification device.
In this way, when shooting the subject of the registration target person, the shooting unit changes the shooting parameters to shoot a plurality of subjects, and the specific shooting parameters and characteristics of the subject of the registration target person The amount is associated and stored in the storage unit. Since a specific parameter is used when shooting the subject of the identification target person, it is possible to take a picture using an appropriate shooting parameter associated with the feature amount of the subject of the registration target person. Therefore, it is not necessary to take a plurality of images when photographing the subject of the person to be identified, so that the speed of the identification process can be increased while maintaining the accuracy of the identification process.

In this identification device, it is desirable that the feature amount is a feature amount characterized regardless of the position and orientation of the subject.
In this way, even when the subject of the registration subject is photographed and when the subject of the identification subject is photographed, even if the position and direction of the photographed subject are different, authentication can be performed easily and with high accuracy. It can be carried out.

Further, the feature amount is an amount corresponding to a feature point characterized regardless of the position and orientation of the subject, and the specific photographing parameter is obtained by changing the photographing parameter to change the subject of the registration target person. It is desirable that the shooting parameters be obtained by obtaining the largest number of feature points by shooting a plurality of times.
In this way, it is possible to perform imaging with imaging parameters that can obtain the largest number of feature points when imaging the subject of the person to be identified. That is, since the subject of the subject of identification can be photographed using the photographing parameters that most easily extract the feature amount of the subject of the subject of registration, the authentication processing can be speeded up while maintaining the accuracy of the authentication processing.

The shooting parameters include an exposure time for shooting the subject, a gain of the shooting unit,
And it is desirable that it is at least one of the light quantity of a light source.
In this way, appropriate shooting parameters can be used.

Further, the shooting parameter includes an exposure time, and the storage unit stores the specific shooting parameter and a feature amount of the subject of the registration target person in association with each other for a plurality of registration target persons, and the identification unit Is the identification target person in order from the feature quantity of the subject of the registration subject having a short exposure time included in the specific imaging parameter among the feature quantities of the subject of the plurality of registration subject persons stored in the storage unit. It is desirable to collate with the feature amount of the subject.
In this way, by collating in order from the feature amount of the subject of the registration subject with a short exposure time,
Even when a plurality of registration subjects are stored in the storage unit, while the total exposure time for collation is short, the features of the subjects of a larger number of registration subjects and the subjects of the identification subjects Since the feature quantity can be collated, the probability that the person to be identified can be identified in a short time increases. Therefore, the authentication process can be speeded up.

In addition, the storage unit stores, for a plurality of registration target persons, the specific imaging parameter and the feature amount of the subject of the registration target person in association with each other, and performs the identification for the plurality of registration target persons. The identification unit stores the feature amount of the subject of the identification target person in order from the feature amount of the subject of the registration subject who is frequently identified among the feature amounts of the subject of the registration subject. It is good also as collating with.
In this way, by performing the matching in order from the feature amount of the subject of the registration target person who is frequently identified, the probability that the authentication target person is authenticated early increases, so that the authentication process can be speeded up.

In addition, the storage unit stores, for a plurality of registration target persons, the specific imaging parameter and the feature amount of the subject of the registration target person in association with each other, and performs the authentication for the plurality of registration target persons. The frequency is stored for each time zone, and the identification unit is configured to calculate, from among the feature quantities of the subjects of the plurality of registration target persons, the feature quantities of the subject of the registration target person having a high identification frequency in the time zone in which the identification is performed. It is good also as collating with the feature-value of the subject of the said identification subject in order.
In this way, since the probability of the authentication target person being authenticated is increased by collating in order from the feature amount of the subject of the registration target person with high authentication frequency according to the time zone, the authentication process can be speeded up. Can do.

In addition, at least the following matters will become clear from the description of the present specification and the accompanying drawings.
That is,
Changing the shooting parameters and shooting the subject of the registration subject multiple times;
A specific shooting parameter when the subject of the registration target person is shot and a feature amount of the subject of the registration target person are stored in association with each other, and the subject of the identification target person is stored using the specific shooting parameter. Shooting,
Identifying the identification target person by comparing the feature quantity of the subject of the identification target person with the feature quantity of the subject of the registration target person stored in the storage unit;
It is an identification method including.
In this way, when shooting a subject of a registration target person, a plurality of subjects are shot by changing the shooting parameters, and a specific shooting parameter among them is associated with a feature amount of the subject of the registration target person. Add and remember. Since a specific parameter is used when shooting the subject of the identification target person, it is possible to take a picture using an appropriate shooting parameter associated with the feature amount of the subject of the registration target person. Therefore, it is not necessary to take a plurality of images when photographing the subject of the person to be identified, so that the speed of the identification process can be increased while maintaining the accuracy of the identification process.

=== Embodiment ===
FIG. 1A is a block diagram of the vein authentication device 1 according to the first embodiment, and FIG.
It is a key map of vein authentication device 1 in an embodiment. The vein authentication device 1 authenticates an identification (authentication) target person. That is, the vein authentication device 1 identifies the person to be identified and controls the control object 50 described later. The vein authentication device 1 includes a calculation unit 10, a sensor unit 20, and a light source unit 30. The sensor unit 20 is connected to the calculation unit 10 via an interface 28, and the light source unit 30 is connected to the calculation unit 10 via an interface 38. Further, the vein authentication device 1 includes the trigger sensor 4 via the interface 48 and the interface 58.
0 and the control target 50 are connected.

The calculation unit 10 includes a CPU (Central Processing Unit) 12 that performs calculation and an R as a storage unit.
AM (Random Access Memory) 14 and EEPROM (Electronically Erasable and Progr
ammable Read Only Memory) 16. The CPU 12 performs vein authentication by executing a program stored in the EEPROM 16. The RAM 14 stores calculation results and the like necessary for performing vein authentication. And the calculating part 10 performs the registration process and authentication process which are mentioned later. That is, the calculation unit 10 functions as an authentication unit and a storage unit, and the authentication unit includes a function as an identification unit that performs identification and a control unit that controls a control target.

The sensor unit 20 is a sensor for photographing a finger vein. That is, the subject in this embodiment is a finger, and a vein image (blood vessel image) is captured. The sensor unit 20 photographs the finger of the registration subject or the finger of the identification subject, and can adjust the exposure time at that time.

The light source unit 30 is a device for illuminating a finger to be photographed with light having a predetermined wavelength. Here, a near-infrared LED (Light Emitting Diode) light source is included, and light including a large wavelength band of 700 nm to 900 nm is irradiated. As described above, the light that includes a large wavelength band of 700 nm to 900 nm is irradiated with light that has a low absorption of blood hemoglobin and water and a high transmittance. Because. In the present embodiment, the sensor unit 2
0 and the light source unit 30 correspond to an imaging unit.

The trigger sensor 40 is a device that senses the approach of a finger and sends a trigger for starting a photographing process to the calculation unit 10. As the trigger sensor 40, for example, a capacitance sensor is used. With the trigger sensor 40, when the vein authentication device 1 is provided on a door, which will be described later, the sensor unit 20 can start the finger photographing process only by bringing the finger close to the door.

The control object 50 is an object controlled by the arithmetic unit 10 according to the identification result by the vein authentication device 1. For example, when the control object 50 is a computer, the vein authentication device 1
In accordance with the identification result of, the computer access right is given to the identification target person. When the control object 50 is an electronic door lock, the electronic door lock is unlocked according to the identification result of the vein authentication device 1. In the present embodiment, description will be given below assuming that the control target 50 is an electronic lock of a door.

In the vein authentication device 1 according to the present embodiment, two processes of a registration process and an authentication process described later are performed.

FIG. 2 is a flowchart of the registration process. The registration process is a process of registering the feature quantity of the registrant's finger vein. As the feature amount in the present embodiment, a position invariant and a rotation invariant are employed. This is because, for the same person's vein pattern, the same position is specified for the pattern regardless of whether the shooting range changes (when the position changes) or the shooting direction rotates (rotation changes) ( (Position invariant), and the same value as the feature quantity obtained by quantifying the characteristics of the local region around the position can be obtained (rotation invariant).

First, a vein image of a registration target person is taken (S102). The vein image can be captured using the sensor unit 20 of the vein authentication device 1 or using another device capable of capturing images. In photographing a finger vein pattern image, the light source unit 30 to 700 nm
Irradiation with light including a large wavelength band of ˜900 nm is applied. In this wavelength band, the absorption of light by hemoglobin is larger than the absorption of light by water, so an image in which the finger vein is shaded is taken. Then, a plurality of images are photographed while slightly varying the photographing parameters. Here, the imaging parameter is the exposure time. Therefore, a plurality of images are taken while changing the exposure time little by little. The amount by which the exposure time is changed can be several milliseconds to several hundred milliseconds.

Next, feature amount extraction is performed from the photographed vein image (S104). These feature quantity extraction processes are performed for each of the plurality of images taken in S102.

FIG. 3 is a flowchart for explaining the feature amount extraction processing.
In the feature amount extraction process (S104) of the present embodiment, first, image correction of the captured vein image is performed (S202). The image correction is performed mainly for the following three reasons.

(1) There is an individual difference in the light transmittance with respect to the finger, and the entire luminance of the acquired image may vary.
(2) A light-dark distribution may occur due to individual differences in finger transmittance. For example, a bright image is acquired at the joint portion of the finger, and a dark image is acquired between the joints.
(3) The living tissue between the vein and the epidermis may diffuse light and blur the photographed vein pattern.

In order to solve these problems, filter processing is performed. In order to solve the above problem (1), normalization is required, and therefore, it is necessary to remove the average value (DC component). Moreover, in order to solve the above problem (2), it is necessary to make uniform, and for this reason, it is necessary to remove moderate fluctuations. Therefore, in order to solve the problems (1) and (2), a high-pass filter is applied to the vein image.

Further, in order to solve the above problem (3), sharpness processing is necessary. Therefore, an unsharp mask is applied to the vein image to emphasize high frequency components. That is, a filter integrating these high-pass filter and unsharp mask is created and applied. Specifically, the frequency response (MTF: Modulation Transfer Function) of the two filters is integrated in the frequency space, and a filter obtained by performing inverse Fourier transform on this is applied.

Next, feature point extraction processing (S204) is performed. SIFT (Scale Invariant Feature Transform) is employed for feature point extraction and feature quantity extraction in the present embodiment. When the positional relationship between the sensor unit 20 and the finger is not fixed, the position and angle of the finger with respect to the sensor unit 20 may not be constant. In order to perform highly accurate identification in such an authentication device,
There is a demand for extracting feature points that give a feature quantity independent of position and rotation. SIFT is adopted as an example of a method that satisfies this requirement.

In the process of S204, first, in order to remove noise and obtain a stable feature point,
An averaging process is performed by applying a Gaussian filter to the vein image. And the process which cuts the component more than a certain frequency is performed. Further, the second derivative of the image to which the Gaussian filter is applied is calculated, and the extreme value is set as a feature point candidate. Further, in order to remove feature points derived from noise, a point having an absolute value of an extreme value equal to or greater than a predetermined threshold is adopted as a feature point. In the above description, the second derivative is calculated in order to obtain the feature point candidate because the edge portion having a change is extracted from the image instead of the uniform region. In addition, when light source irradiation is performed obliquely during photographing, a region that changes with a certain inclination may appear in the image, but such a region is not used as a feature point candidate. Specifically, the calculation of the second derivative is performed by performing convolution integration of a Gaussian derivative.

FIG. 4 is a diagram illustrating an example of feature points. In the figure, the photographed vein and the feature point specified at the branch point of the vein are shown. As the feature point, the location of the extreme value of the second derivative of the luminance gradient is selected, so that a portion having a large luminance change amount is selected. That is, a place that separates a vein from a place that is not so is selected as a feature point. Further, since the place of the extreme value of the second derivative is a place where the amount of change of the first derivative is maximized, a point having a larger curvature (ie, how to bend) than the surroundings is selected. Therefore, the branch point of the vein and the inside of the blood vessel are also selected as the feature points.

Next, feature amount extraction is performed (S206). The feature amount extraction is performed by performing the following processing on each feature point obtained in the above processing. First, the brightness gradient around the feature point is calculated.

FIG. 5 is a diagram illustrating an example of the obtained luminance gradient. FIG. 5 shows a plurality of grids (in this embodiment, 8 × 8 grids) centering on the feature points. And the brightness | luminance gradient in each square is shown as a vector amount.

Next, a histogram is created for the luminance gradient obtained as shown in FIG. The direction with the highest frequency is set as the reference direction of the feature amount.

FIG. 6 is a diagram illustrating an example of a histogram of luminance gradient. In this figure, a histogram in 36 directions is shown. In this figure, the value of the direction described as “peak” is the highest. Therefore, this direction becomes the reference direction of the feature amount.

Next, centering on the feature point, the 8 × 8 is again set in accordance with the reference direction selected in the above-described processing.
Create the squares. Then, this 8 × 8 cell is made to correspond to the 4 × 4 cell,
The luminance gradient for each square of × 4 is decomposed into eight-direction vectors for each square.

FIG. 7 is a diagram illustrating an example of the luminance gradient when the coordinate axis is aligned with the reference direction. In FIG. 7A, the direction of the thick arrow is the above-described reference direction, and an 8 × 8 cell whose direction is aligned with the reference direction is re-created, and the luminance gradient is obtained again using the reference direction as a reference. . In addition, FIG.
B corresponds to the 8 × 8 cells corresponding to the 4 × 4 cells, and the luminance gradient is decomposed into vectors in eight directions for each cell.

Here, since it is decomposed into vectors in 8 directions, 0 °, 45 °, 90 °, 135 °,
A vector scalar quantity is obtained for each direction of 180 °, 225 °, 270 °, and 315 °. In addition, since these scalar quantities are obtained for each 4 × 4 cell, 4 × 4 × 8 = 128-dimensional scalar quantities can be obtained. In the present embodiment, the feature quantity at the feature point is a multi-dimensional scalar quantity.

In this way, a plurality of feature points and a feature amount corresponding to each feature point can be obtained for one image. Such processing is performed for each of a plurality of images taken in S102, and feature amounts corresponding to the respective images are obtained. Next, a shooting parameter to be registered is selected based on the number of feature points of the obtained plurality of images (S106).

FIG. 8 is a first explanatory diagram of the selected exposure time. In the figure, 11 image samples are shown, from an image with a short exposure time to an image with a long exposure time. In the figure, a graph showing the number of feature points obtained in an image taken by changing the exposure time is shown. The horizontal axis of the graph is the exposure time, and the vertical axis is the number of feature points in each image. The point with the shortest exposure time corresponds to the image of I-01, and the longest point corresponds to the image of I-11.

In I-01, since the exposure time is the shortest, the vein image is darkest overall. On the other hand, since the exposure time is the longest in I-11, the whole vein image is brightest. Referring to the graph, if the exposure time is too short, the number of feature points is small because the contrast is low. As the exposure time is increased, the number of feature points increases. However, if the exposure time is excessively long, a phenomenon of whiteout occurs in the vein image, and the number of feature points decreases rapidly.

In the present embodiment, among the 11 kinds of shooting parameters (here, exposure time), a shooting parameter that maximizes the number of feature points obtained from the shot image is selected in S106. In FIG. 8, since the number of obtained feature points is the maximum in the image of I-07, the exposure time used for the shooting of I-07 is selected here.

FIG. 9 is a second explanatory diagram of the selected exposure time. This figure shows the number of feature points with respect to the vein image and exposure time when a vein image of a person with high finger transmittance is taken. FIG.
In FIG. 8, the graph of the person is represented by a dotted line, and the graph of the person having a higher light transmittance to the finger than the person in FIG. 8 is represented by a solid line. Thus, referring to the graph of a person with high transmittance, the exposure time with the maximum number of obtained feature points is shorter than that of the person in FIG. As described above, since appropriate photographing parameters differ depending on a person, it is desirable to select appropriate photographing parameters based on a plurality of vein images having different photographing parameters.

Next, the imaging parameter selected in S106 and a plurality of feature amounts obtained based on the vein image captured with the imaging parameter are stored in the calculation unit 10 (RAM 14) (
S108).

FIG. 10 is a table showing an example of a feature amount database. In the table, an exposure time as a photographing parameter and a 128-dimensional scalar amount of 0 to 127 as a feature amount are shown. In this manner, a plurality of specific imaging parameters selected in S106 for a specific registration target person and a plurality of feature amounts for the registration target person are registered. In the present embodiment, since the shooting parameter having the largest number of extracted feature points is selected in S106, a plurality of feature amounts corresponding to the feature points are registered. Note that a feature amount related to one feature point may be registered for one registrant. When the registration process is completed, the authentication process can be performed using the registrant's feature amount stored in the calculation unit 10 of the vein authentication device 1. Hereinafter, the authentication process will be described.

FIG. 11 is a flowchart of the authentication process. FIG. 12 is a diagram for explaining the concept of the authentication process. In the authentication process, first, a vein image photographing process (S302) is performed. The vein image capturing process (S302) in the authentication process is a process for capturing a vein image of the finger of the person to be identified, similar to the vein image capturing process in the registration process described above. One vein image is taken.

Next, feature amount extraction processing (S304) is performed. Since the feature amount extraction processing is the same as the processing in steps S202 to S206 described above, description thereof is omitted. Here, all feature quantities of feature points in one vein image photographed in S302 are extracted.

The feature quantity obtained in this way can be obtained as a 128-dimensional scalar quantity as described above. In FIG. 12, the registered registrant feature amounts are shown as “registered feature amounts” as a plurality of bar graphs. Further, the feature quantity of the person to be identified obtained in step S304 is shown as a plurality of bar graphs as “authentication feature quantities”. Conceptually, in the following S306 and S308, a registered feature amount similar to the authentication feature amount is searched, and the identification target person is identified based on whether or not the index indicated as the similarity is larger than a predetermined amount.

Next, collation processing (S306) is performed. In this process, in order to compare with the registered feature value, the similarity between the registered feature value and the feature value acquired at the time of authentication is obtained. Here, the similarity can be obtained based on the Euclidean distance for the obtained 128-dimensional scalar quantity. And when these distance values are smaller than a predetermined threshold value, it can be determined that the similarity is high. Here, the similarity is obtained based on the Euclidean distance, but the similarity may be obtained based on the city distance or the Mahalanobis distance. By doing in this way, it is possible to obtain the degree of similarity between the feature quantity obtained based on the vein image photographed in the authentication process and the registered feature quantity. And it counts about the number of the feature-values determined that the calculated | required similarity is high (the value of distance is smaller than a predetermined threshold value).

Next, it is determined whether or not the identification target person is a registrant (registration target person) based on the counted value (S308). Specifically, when the counted value is larger than a predetermined count value, it is determined that the person to be identified is a registrant (YES in S308). On the other hand, if the counted value is less than or equal to the predetermined count value, it is determined that the person to be identified is not a registered person (S308).
NO). If the determination result (result of S308) is YES, a permission signal for permitting the operation of the controlled object is transmitted to the electronic lock 50 that is the controlled object (S310). Judgment result (S30
When the result of 8) is NO, a rejection signal not permitting the operation of the control target is transmitted to the electronic lock 50 (S312). The electronic lock 50 is unlocked when a signal to permit is received, and is not unlocked when it receives a signal that is not permitted, that is, does not operate.

As described above, authentication is performed by setting the shooting parameter when shooting the subject of the authentication target to the shooting parameter having the maximum number of feature points when shooting the subject of the registration target. Sometimes it is possible to perform highly accurate authentication processing without taking a plurality of images. In addition, the authentication process can be speeded up.

FIG. 13A is a diagram illustrating an example of information registered in the second embodiment. In the figure, the ID of the person to be registered, the shooting parameters, and the frequency are shown. In the second embodiment, feature quantities of a plurality of registration subjects are stored. The registration process for registering each registrant is the same as the registration process (S102 to S108) of the first embodiment. For this reason, the imaging parameters associated with the feature amount are stored for each registration target person. In the present embodiment, as described below, the frequency with which the person to be registered performs authentication is recorded in association with the ID, and the shooting parameters with the highest frequency are used as shooting parameters at the time of authentication. Here, first, a shooting parameter whose registration target person ID is “2” is employed. Here, “Frequency”
Can be the authentication frequency for the past few days. For the same configuration as the first embodiment,
The same reference numerals as those in the first embodiment are attached and detailed description is omitted.

FIG. 13B is a table illustrating an example of a feature amount database according to the second embodiment. The table shows the ID of the registrant and 128-dimensional scalar quantities from 0 to 127 as feature quantities.
The reason why a plurality of the same numbers are arranged as the registrant ID is that feature quantities relating to a plurality of feature points are registered for one registrant. Note that a feature amount related to one feature point may be registered for one registrant.

The number of feature quantities to be stored is arbitrary. However, when multiple registrants are registered, at the time of identification described later, if the number of registered feature quantities of a specific registrant is extremely greater than the number of registered feature quantities of other registrants, If the target person matches a registrant who has a large number of registrations, the possibility of erroneous identification increases probabilistically. In order to reduce such an authentication bias, it is also possible to store a number of feature quantities each having a certain number of restrictions for each registrant. In this case, the feature amount to be stored may be selected at random, or a predetermined number may be registered from the top by providing a selection criterion. One registrant may register the feature quantities of a plurality of fingers. In this way, the number of fingers that can be authenticated can be increased.

FIG. 14 is a flowchart of authentication processing in the second embodiment. First, one of the registration subjects is selected (S402). At this time, the table of FIG. 13A is referred to, and the registration target person with the highest “frequency” is selected. Here, the registration target person whose ID is “2” is selected.

Next, a shooting process is performed using the shooting parameters of the selected registration target person (S404).
. Here, since the shooting parameter with ID “2” is selected in S402, 100 is selected.
(Msec) is selected. Then, the finger of the person to be identified is photographed with this photographing parameter, and one vein image is acquired.

Next, a feature amount extraction process is performed on the vein image obtained in S404 (S406).
Since the feature amount extraction processing has already been described, the description is omitted.

Next, collation processing (S408) is performed. The collating process is also substantially the same as the above-described step S306. However, in the second embodiment, since the feature amounts of a plurality of registration target persons are registered, a count value is obtained for each registration target person ID.

Next, a determination is made as to whether the selected registration target person has the largest count value (
S410). That is, here, a determination is made as to whether or not the count value of the registration target person whose ID is “2” is the largest. When the count value is the largest, that is, when the count value is larger than the count value of another registrant, it is determined that the registration target person is the identification target person (YES in S410). On the other hand, when the count value of the selected registration target person is not the largest, that is, when the count value of other registrants is less than or equal to, whether there is a registration target person not selected among the registration target persons It is determined (S414). When there is a registration target person who has not yet been selected (YES in S414), the next registration target person with the highest "frequency" is selected. Here, the registration target person whose ID is “4” is selected. And it returns to step S404 and the same process is performed. On the other hand, when there is no registration target person who has not yet been selected (S414 is NO), it is determined that there is no identification target person among the registration target persons (S414 is NO).

If the determination result (result of S410) is YES, a permission signal for permitting the operation of the controlled object is transmitted to the electronic lock 50 that is the controlled object (S412). Further, when there is no registration target person who is not selected in S414 (S414 is NO), a rejection signal not permitting the operation of the control target is transmitted to the electronic lock 50 (S418). The electronic lock 50 is unlocked when a signal to permit is received, and is not unlocked when it receives a signal that is not permitted, that is, does not operate.

In this way, even when the feature quantities of a plurality of registration target persons are registered, the identification target person can be identified early by collating in order from the feature quantity of the subject of the registration target person with a high authentication frequency. Since the probability of authentication increases, authentication processing can be performed at high speed while maintaining authentication accuracy.

In the above-described embodiment, the photographing parameter is the exposure time of the photographing unit. However, when amplification processing is performed on the obtained luminance, the gain may be used. Alternatively, the light amount of the light source unit 30 may be used.

Further, in the above-described embodiment, matching with the feature amount is performed in descending order of “frequency”. However, matching may be performed in ascending order of exposure time. In this case, since the total exposure time at the time of authentication can be shortened, the authentication process can be speeded up.

In addition, the authentication frequency of the registration target person may be stored for each time period, and the feature amounts may be collated in the order of the registration target person with the highest authentication frequency during the authentication time. In this case, it is possible to speed up the authentication process because the probability that the identification target person is authenticated at an early stage increases by performing the matching in order from the feature amount of the subject of the registration target person who has a high authentication frequency according to the time zone. it can.

In the feature point search method and the feature amount calculation, SIFT is used in the above description, but the method of obtaining the feature point and feature amount is not limited to this.

For example, when MSER (Maximally Stable External Regions) is used, and the region is divided into grayscale images and the threshold is gradually increased, the binarization result is the most unchanged region. It is good also as adopting as a point.

Also, when obtaining the feature amount, SURF (Speeded Up Robust Features), GLOH (Gradient
ent Location and Orientation Histogram) and Shape Context may be used.

1 vein authentication device,
10 arithmetic units, 12 CPU, 14 RAM, 16 EEPROM,
20 Sensor part,
30 light source section,
40 trigger sensor,
50 Control target

Claims (8)

A shooting section for shooting the subject;
A storage unit for storing a specific shooting parameter and a feature amount of the subject of the registration target in association with each other when the shooting unit captures the subject of the registration target;
An identification unit for identifying the identification target person by comparing the feature quantity of the subject of the identification target person photographed by the imaging unit with the feature quantity of the subject of the registration target person stored in the storage unit; ,
Including
In the case of photographing the subject of the registration target person, the photographing unit photographs the subject of the registration target person a plurality of times by changing photographing parameters, and when photographing the subject of the identification target person. Photographing the subject of the identification subject using the specific photographing parameters;
Identification device. The identification device according to claim 1, wherein the feature amount is a feature amount characterized regardless of a position and a posture of the subject. The feature amount is an amount corresponding to a feature point characterized regardless of the position and posture of the subject,
The identification device according to claim 1, wherein the specific shooting parameter is a shooting parameter that obtains the largest number of feature points by shooting the subject of the registration target person a plurality of times by changing the shooting parameter. . The shooting parameters include an exposure time for shooting the subject, a gain of the shooting unit, and
The identification device according to claim 1, wherein the identification device is at least one of the light amounts of the light sources. The shooting parameter includes an exposure time,
The storage unit stores the specific shooting parameter and the feature amount of the subject of the registration target person in association with each other for a plurality of registration target persons,
The identification unit is configured in order from the feature amount of the subject of the registration subject having a short exposure time included in the specific imaging parameter among the feature amounts of the subject of the plurality of registration subjects stored in the storage unit. The identification apparatus according to claim 4, wherein collation is performed with a feature amount of a subject of an identification target person. The storage unit stores, for a plurality of registration target persons, the specific imaging parameter and the feature amount of the subject of the registration target person in association with each other, and the frequency of performing the identification for the plurality of registration target persons. Remember,
The identification unit collates with the feature amount of the subject of the identification target in order from the feature amount of the subject of the registration subject who frequently performs the identification among the feature amounts of the subject of the plurality of registration subjects. The identification device according to claim 1. The storage unit stores, for a plurality of registration target persons, the specific imaging parameter and the feature amount of the subject of the registration target person in association with each other, and the frequency of performing the identification for the plurality of registration target persons. Remember every time zone,
The identification unit includes, in order from the feature amount of the subject of the plurality of registration subjects, the feature amount of the subject of the identification subject in order from the feature amount of the subject of the registration subject who is frequently identified in the time zone in which the identification is performed. The identification device according to any one of claims 1 to 4, wherein collation with an amount is performed. Changing the shooting parameters and shooting the subject of the registration subject multiple times;
A specific shooting parameter when the subject of the registration target person is shot and a feature amount of the subject of the registration target person are stored in association with each other, and the subject of the identification target person is stored using the specific shooting parameter. Shooting,
Identifying the identification target person by comparing the feature quantity of the subject of the identification target person with the feature quantity of the subject of the registration target person stored in the storage unit;
Including the identification method.