JP2000293689A - Personal authentication device and personal authentication method - Google Patents

Abstract

(57) [Summary] [Problem] To provide a personal authentication device and a personal authentication method capable of stably capturing a feature amount at the time of authentication. Kind Code: A1 A CR for presenting guidance information such as characters, images, or voices to a person to be authenticated who receives personal authentication.
It comprises a T display section 11, a CCD image pickup section 12 for taking an image of a tooth arrangement of a person to be authenticated and capturing the image as authentication image data, an information processing section 13, a keyboard 14 and a mouse 15. The information processing unit 13 registers an information extraction unit for extracting information on the tooth arrangement for authentication from the authentication image data of the subject, and information on the tooth arrangement of an authentication permitter who has been previously approved. A storage unit for holding
A matching unit that determines the degree of coincidence between the information on the tooth arrangement extracted by the information extracting unit and the information on the tooth arrangement of the authentication acceptor registered in the storage unit; And an output unit for designating the acceptance or rejection of the output.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a personal identification device and a personal identification method, and more particularly to a personal identification device and a personal identification method based on information relating to the shape of an individual's teeth by using a method for identifying and authenticating the individual. The present invention relates to a personal authentication device and a personal authentication method applied for a wide range of personal authentication including management, user management for a specific service device, or personal authentication for crime prevention.

[0002]

2. Description of the Related Art Conventionally, various methods have been attempted for personal authentication devices for authenticating individuals. In particular, in recent years, a personal authentication device that authenticates an individual using biometric information, which is biometric information unique to each individual including a fingerprint, a face, a voice, a retina, an iris, and the like, has been proposed. In a personal authentication device using these biometrics information as a feature amount, the user himself / herself, in contrast to a normal personal authentication device that is authenticated by personal authentication information such as a password or personal identification number that can only be known by the user,
There is a merit that the problem of not being able to respond to personal authentication due to forgetting these personal authentication information and the problem of losing the medium of the personal authentication information are obviated beforehand. Since it is not easy to simulate and imitate the biometric information of the individual, there is a merit that the problem of impersonating the individual and performing an improper act can be eliminated beforehand.

[0003]

Some of the above-described conventional personal recognition devices that use biometric information as a feature amount include a face image, a fingerprint image, and voice information as the feature amount. For example, when an individual is authenticated using a facial image as a feature amount as the biometric information, the user has a high degree of freedom to deform itself as seen in changes in facial expressions. The acquired face image information is susceptible to fluctuations each time the image information is input to the personal authentication device, and there is a problem that it is difficult to capture stable image information as a feature amount. This means that when a fingerprint image is used as a feature, there is a change due to the degree of pressing of a finger, and when voice information is used as a feature, there is a change in the vocal system. A similar problem exists in that Therefore, the conventional personal authentication apparatus using biometrics information as a feature quantity has a drawback that it is difficult to fetch a stable feature quantity due to the fluctuation of the feature quantity, which becomes a hindrance factor for personal authentication.

SUMMARY OF THE INVENTION The present invention has been made in order to solve the drawbacks caused by the conventional feature amount in the personal authentication apparatus. The purpose of the present invention is to change according to the intention of the person to be authenticated or when inputting an image. An object of the present invention is to provide a personal authentication device and a personal authentication method that can stably capture a feature amount used for personal authentication by using information about a tooth shape that is a feature amount that is difficult to perform.

[0005]

According to the present invention, there is provided a personal authentication apparatus comprising: an image pickup unit for picking up an image of a tooth shape of a person to be authenticated; an image picked up by the image pickup unit as authentication image data; Information extraction means for extracting information on the tooth shape of the person to be authenticated, storage means for registering and holding information on the tooth shape of the authentication-permitted person, and information extraction means for extracting from the authentication image data by the information extraction means The degree of coincidence of the information is detected by comparing and matching the information on the tooth shape of the person to be authenticated with the information on the tooth shape of the authenticated person registered in the storage means. It is characterized by comprising authentication determination means for determining acceptance or rejection of authentication of a person to be authenticated.

[0006] The information on the shape of the teeth may be set as information on "teeth arrangement", and the information on "teeth arrangement" may be set as information on "arrangement of tip parts of teeth". Further, the information on the shape of the tooth may be information on “unevenness on the surface of the tooth”.

Further, the information extracting means receives input of authentication image data relating to the shape of the tooth of the person to be authenticated, which is imaged by the imaging means, and identifies a specific pixel accompanying the tooth shape in the authentication image data. Concatenating to generate a tip silhouette, extracting a point having a large curvature value on the tip silhouette as a curvature feature point, wherein the authentication determination means includes a collation means and an output means therein, and in the collation means, In response to the input of the curvature feature points, the correspondence between the curvature feature points and the curvature feature points corresponding to the information on the tooth shapes registered in the storage unit is determined. It is performed in the held state, and by comparing and matching the two curvature feature points, the presence or absence of coincidence between the two is detected, and the acceptance or rejection of the authentication of the authenticated person is determined and output through the output means. It may be.

Further, in the personal authentication method of the present invention, a first step of photographing the teeth arrangement of the person to be authenticated by a predetermined imaging means and inputting information regarding the teeth arrangement as authentication image data, and processing the authentication image data By doing so, as the authentication data to be used for personal authentication, the second step of extracting the features of the teeth arrangement of the person to be authenticated and the form of the features of the teeth arrangement included in the authentication data are registered in advance. A third step of associating with the form of the feature amount of the tooth arrangement included in the registration data of the subject, and receiving the third step and receiving the third step. And a fourth step of comparing and matching the feature amount of the tooth arrangement to be performed with the feature amount of the tooth arrangement included in the registration data of the authorized person to determine the degree of coincidence between the two feature amounts. It is characterized by a door.

In the second step, the tip silhouette of the tooth arrangement obtained by connecting pixels having a large difference in the luminance value of the image is obtained from the authentication image data obtained in the first step. A tip silhouette extraction processing step of extracting as a quantity, a curvature feature point search processing step of finding a curvature at each point on the tip silhouette, and extracting a point whose absolute value of the curvature exceeds a predetermined threshold value as a curvature feature point The third step may include the step of acquiring a curvature feature point of a tip silhouette of a row of teeth of the subject included in the authentication data, the curvature feature point being obtained in the curvature feature point search processing step. The correspondence between the curvature feature points of the tip silhouette of the tooth arrangement of the authorized person included in the data and the order feature of each curvature feature point is determined. A curvature feature point associating process step performed in a state in which is held, and a value of the degree of coincidence between the curvature values of the two curvature feature points associated in the curvature feature point associating process step is set to a predetermined threshold value Is determined, and if so, it is determined that the authentication data of the subject corresponds to the registration data of the authorized person, and the process proceeds to the next step. In this case, it is determined that there is no correspondence between the tip silhouettes in both data, and the curvature feature point determination processing step of rejecting the authentication of the user is performed. Is determined to correspond to the registration data of the authentication permitter, the tip silhouette of the authentication data so that the curvature characteristic points corresponding to the registration overlap each other. And linear transformation, it may have a distal silhouette associating process step be substantially matched with each other at the tip silhouette.

[0010]

According to the above configuration or method, at the time of personal authentication, information on the shape of the teeth, to which the image of the person to be authenticated is unlikely to fluctuate, is used as the feature quantity, so that it is possible to stably capture the information. Performed accurately with high precision.

[0011]

Next, the present invention will be described.

The embodiment of the present invention basically includes an image pickup means for picking up the shape of the teeth of a person to be authenticated, an image picked up by the image pickup means as authentication image data, and an individual image based on the authentication image data. Information extraction means for extracting information on the shape of the tooth of the person to be authenticated as the feature amount for authentication; storage means for registering and holding information on the tooth shape of the person to be authenticated in advance as the feature amount; The information on the tooth shape of the person to be authenticated extracted from the authentication image data is compared with the information on the tooth shape of the authenticated person registered in the storage means. Detecting the presence or absence, based on the detection result is configured to include authentication determination means to determine whether to accept or reject authentication of the person to be authenticated, information on the shape of the teeth of the opening,
It has a remarkable feature in that it is defined as a feature amount for personal authentication. In other words, the feature amount of the tooth arrangement is extracted from the authentication image data of the person to be authenticated, and the feature amount is compared with the feature amount of the tooth arrangement of the authentication acceptor registered in advance, so that the object to be authenticated is authenticated. To determine whether or not
It is determined whether to accept or reject the authentication of the person to be authenticated.

FIG. 1 is a block diagram showing the configuration of an embodiment of the personal authentication device of the present invention. When a personal computer is operated as a terminal device, the use of the personal computer is specified in terms of line protection. This is an embodiment in which the present invention is applied in order to limit the number of users (permitted users). As shown in FIG.
CRT display unit 11 for presenting guidance information such as characters, images, or voices to a person to be authenticated (hereinafter, referred to as a user who wishes to use personal authentication) in order to use a personal computer as a terminal device. A CCD imaging unit 12 which captures the tooth arrangement image of the user and captures the image as authentication image data; and an information extraction unit which extracts, from the captured authentication image data, authentication data of the user regarding the tooth arrangement used for authentication. A storage unit that stores registration data relating to the tooth arrangement of the use permitter in advance, a matching unit that compares and determines the degree of coincidence between the authentication data extracted by the information extraction unit and the registration data of the use permitter stored in the storage unit, And information including an output unit for designating acceptance or rejection of authentication to a user who wishes to use the information in response to the determination result by the matching unit. A processing section 13 controls the operation of the start-up and communication operations of the personal computer system as a terminal device configured to include a keyboard 14 and a mouse 15 functioning as control means for the personal authentication. The CRT display unit 11, the information processing unit 13, the keyboard 14, and the mouse 15 form the personal computer.

FIG. 2 is a diagram showing a flowchart of a personal authentication method by the personal authentication device of the present invention.

The operation mode of the personal authentication apparatus of the present invention includes a registration mode in which information on the tooth arrangement of a person who is permitted to use the terminal device is previously registered as a feature amount in the storage unit.
There are two operation modes, an authentication mode in which a user who wishes to use is authenticated to use a personal computer as a terminal device. Hereinafter, the registration mode and the authentication mode will be described with reference to FIGS.

First, the registration mode will be described. The flowchart of FIG. 2 is a flowchart showing a processing procedure of an authentication mode until a user who wishes to use a terminal device receives personal authentication and is accepted or rejected when using a personal computer. In Although,
The two processing procedures of the tooth arrangement image input processing (S1) and the feature quantity extraction processing (S2) are similarly applied as the same processing contents in the registration mode. In this registration mode, the personal authentication device shown in FIG. 1 performs registration processing including the above-mentioned tooth alignment image input processing (S1) and feature quantity extraction processing (S2), and the user who is permitted to use is registered. Information on the arrangement of the teeth is registered as a feature amount in a storage unit included in the information processing unit 13. As shown in FIG. 2, the content of the feature amount extraction processing (S2) is based on three processings including a tooth alignment area search processing (S21), a tip silhouette cutout processing (S22), and a curvature feature point search processing (S23). Made up of

[0017] In order to prevent unauthorized registration in the registration mode, for example, a registration applicant (hereinafter, referred to as a registration permitted person) who is previously approved as a person who intends to use a personal computer as a terminal device is referred to. In particular, a specific identification code or the like is assigned in advance, and at the start of the registration operation, whether or not the personal computer system can be started is determined based on the identification code and the like, and the unauthorized registration of the exclusion user is excluded. Authentication treatment or the like is used. With this pre-authentication process, even if a use exclusion person other than the permitted registrant attempts to register illegally, the personal computer cannot be started and registration is impossible. In the following, a description will be given of a processing procedure of the registration mode by the permitted registrant after the start-up of the personal computer system.

In step S1, the personal authentication device is shifted to the registration mode by operating the keyboard 14 and the mouse 15. Next, based on the operation guidance information provided by the CRT display unit 11, the CCD imaging unit 12 photographs the tooth arrangement of the accepted registrant who wishes to newly register as a user. At this time, image data of tooth arrangement suitable for authentication is presented as a template, and a moving image input from the CCD imaging unit 12 is displayed in real time on the CRT display unit 11.
By displaying the image as a mirror image on the monitor screen, the position and size of the tooth can be adjusted to an appropriate position and size. Displaying a mirror image on the monitor screen matches the human sense, so that positioning is easier. Also, for example, it is also possible to present instructions on the monitor screen by characters, images, sounds, or the like so as to be displayed in a size of 5 mm at the center of the monitor screen. Also, as shown in FIG. 3, on the monitor screen,
By presenting a plurality of parallel horizontal lines as the auxiliary line 21, the user can make the opening horizontal and the region in the back of the mouth that can be seen from between the upper teeth and the lower teeth at a predetermined interval. Thus, the opening posture can be easily adjusted.
In this way, as a preliminary operation for capturing a tooth alignment image with little change in the bitch angle and the roll angle, the registered person permits the registration of the tooth alignment image on the monitor screen. As described above, when the alignment of the tooth alignment image on the monitor screen of the CRT display unit 11 is completed, the registered person accepts a key input of the keyboard 14 or a mouse button of the mouse 15 to instruct the image capture. Click operation. With this operation, an image capturing instruction is performed, and a still image (hereinafter, referred to as authentication image data) including the tooth arrangement presented at the same time is captured and input to the information extracting unit 13.

As the teeth arrangement used for authentication, it is generally preferable to use only a small amount of opening the mouth and use several upper and lower front teeth which have a low psychological resistance as the characteristic amount. Alternatively, the characteristic amount can be appropriately set according to the ease of presenting the tooth arrangement of the registered permitted person including only the lower teeth, or according to a request such as authentication accuracy.
In order to absorb a change in the angle of view at the time of imaging in the CCD imaging unit 12, the instruction is presented plural times by presenting instructions so that the yaw angle with respect to the CCD imaging unit 12 is imaged at different angles. It is preferable to use each of these images for creating registration data described later. Examples of the instruction include an instruction indicating a direction in which a face should be turned on a personal computer screen,
The installation, such as changing the position of the currently input image for checking the input image and the visual interface such as the click button that triggers the image capture to the left, center, and right of the monitor screen, is available. Conceivable. It should be noted that these instructions can be performed by voice in addition to characters or images. The registered acceptor changes the direction of the face according to the above-described instructions, and can check the image including the currently input tooth arrangement on the monitor screen, whereby the CCD imaging unit 12 has a different yaw angle. A plurality of tooth alignment images are easily captured.

Next, a description will be given of a feature amount extraction process (S2) mainly performed in the information extraction unit included in the information processing unit 13 in the registration mode. As shown in FIG. 2, the content of this process is a tooth alignment region search process (S
21) and three processes including a tip silhouette cutout process (S22) and a curvature feature point search process (S23). The tooth alignment area search processing (S21) is intended to extract an area where the tooth alignment exists from the authentication image data input from the CCD imaging unit 12.

In step S21, first, in order to speed up the tooth alignment area search process (S21), the information extraction unit uses the original image based on the authentication image data (FIG. 4A).
Pyramid image (see Fig. 4)
(See (b)). At this time, the degree to which the resolution is reduced is determined by a balance between the authentication performance and the processing capability of the authentication device. In FIG. 4, on the pyramid image, the image area is divided into a plurality of areas in the Y-axis direction, and an edge detection operator typified by a Sobel Operator or the like is applied to each of the divided areas, whereby Y
The amount of change in the luminance value in the axial direction is detected. The number of divisions of the image area is preferably three or more in order to separate the lips and both cheeks.However, if the number of divisions is increased, the processing amount increases. Will decide. FIGS. 5A and 5B show the tooth arrangement areas of the upper teeth and the lower teeth extracted by the tooth arrangement area search processing (S21).

Now, taking the upper teeth as an example, FIG.
As can be seen from (a), there is generally a large difference in the brightness value between the teeth and the region deep in the mouth. For this reason, a large output value is obtained at the boundary between the tooth and the mouth area as an output value accompanying the change amount of the luminance value of the authentication image data, whereas the output value at the boundary between the tooth and the mouth is obtained. Since the change amount of the luminance value is small in the area, the output value is a small value. As a result, when the output values of the operator are accumulated on the Y axis, the distribution has a large peak value near the tip of the tooth. Therefore, threshold value processing is performed on the cumulative distribution, and the peak value having the cumulative value equal to or larger than the threshold value is extracted to extract the maximum position of the tooth arrangement in each region. Can be. Then, by cutting out a region including these tip positions, a tooth alignment region can be extracted. Note that the order of the area division and the application of the edge detection operator may be reversed. The same applies to the case of taking the lower teeth of FIG. 5B as an example. When a color image is used, a tooth alignment region can be extracted by focusing on the tooth color.

Next, a tip silhouette cutting process (S2
2) will be described. The purpose of this process is to obtain a tooth alignment feature amount to be used for authentication from a tooth alignment region cut out from the original image.

In step S22, a line figure appearing at the boundary between the region deep in the mouth of the registered person and the tooth arrangement is defined as a tip silhouette of the tooth arrangement, and the tip silhouette is extracted as a feature amount of the tooth arrangement. To this end, the above-described edge detection operator is applied in all directions to the tooth arrangement area obtained in the above-described tooth arrangement area search processing (S21), thereby extracting pixels having a large change in luminance value. Then, as shown in FIGS. 6A and 6B, a combination of these pixels is extracted as a “tip silhouette”. In FIGS. 6A and 6B, the tip silhouettes 22 and 23 are indicated by thick black lines, respectively. In this case, the brightness difference between the lips or gums and the back of the mouth may be detected, but since these are small compared to the brightness difference between the teeth and the back of the mouth, Can be easily removed,
This is not a practical problem. Further, in the case of a color image, the color difference may be focused on and then separated.

Next, as a final processing procedure included in the feature amount extraction processing (S2), a curvature characteristic point search processing (S2
3) will be described. This process is a process for obtaining a curvature characteristic point serving as a reference in an alignment process (S3) described later.

In step S23, a curvature detection operator is applied to the tip silhouette obtained in the tip silhouette cutout processing (S22), and the curvature at each point on the tip silhouette is determined. Here, the curvature detection operator means a process of calculating the curvature at each point on the tip silhouette, including the sign.

The curvature will be described. The curvature is a value indicating the degree of rotation of a tangent at an arbitrary point on the curve. Focusing on the point a on the curve, the slope s− of the straight line composed of the point a (−k) at the k-th left from the point a and the point a
And the angle between the k-th point a (+ k) to the right of the point a and the slope s + of the straight line composed of the point a can be defined as k-curvature at the point a. Therefore, the curvature varies depending on the value of k. It is necessary to set the value of k to an optimum value depending on the target application. In the personal authentication device according to the present embodiment, the curvature value of the boundary between adjacent teeth or the tip of a sharp tooth becomes a large value.
Determine the value of k. Then, as shown in FIG. 7, the absolute value of the curvature is calculated from the tip silhouette 24 of the tooth arrangement.
A point having a value equal to or larger than a predetermined curvature value is extracted as a curvature feature point 25. In FIG. 7, what is indicated by a circle is
These are the curvature feature points, and generally, there are a plurality of curvature feature points. As a result, the boundary positions of adjacent teeth or the tip positions of sharp teeth are extracted as the curvature feature points 25, respectively. In the registration mode, these curvature feature points are not sent to the matching unit included in the information processing unit 13 shown in FIG. The information is input and stored in the storage unit.

That is, in the registration mode, the information extraction unit of the information processing unit 13 performs the above-described feature amount extraction processing, so that the tip silhouette of the tooth arrangement and the tip silhouette of the tip silhouette are registered for each registered person. The position and curvature value of each curvature feature point used for the alignment process are extracted, registered and held in the storage unit included in the information processing unit 13 as registration data used at the time of personal authentication. Then, with the keyboard 14 or the mouse 15,
The registration mode is ended by inputting a registration mode end command to the information processing section 13.

Next, the authentication mode will be described with reference to the block diagram of FIG. 1 and the flowchart of the personal authentication method of FIG. For the activation of the authentication mode, for example, the personal computer may be automatically shifted to the authentication mode when the power to the personal computer is turned on.

As shown in FIG. 2, the processing contents in the authentication mode mainly include the tooth arrangement image input processing (S1) performed in the CRT display section 11 and the CCD image pickup section 12, and mainly in the information processing section 13. Feature quantity extraction processing (S2) performed in the information extraction unit;
The alignment processing (S3) and the authentication determination processing (S4) mainly performed in the collating unit and the output unit included in the information processing unit 13 are included. Of these,
Teeth alignment image input processing (S1) and feature extraction processing (S1)
The processing contents of 2) are almost the same as the processing contents of the tooth arrangement image input processing (S1) and the feature quantity extraction processing (S2) in the above-mentioned registration mode. The difference is that in this authentication mode, the curvature feature points and the like extracted by the feature amount extraction processing (S2) performed in the information extraction unit of the information processing unit 13 are not used for registration but for a personal computer as a terminal device. The verification unit included in the information processing unit 13 as information for determining whether to accept or reject the use of the personal computer with respect to the authenticated person who wishes to use the personal computer, that is, the use applicant. It is transmitted to. Accordingly, the processing contents of the tooth arrangement image input processing (S1) and the feature quantity extraction processing (S2) are omitted to avoid duplication of description,
Alignment processing (S3) and authentication determination processing (S4)
The processing contents related to the above will be described.

First, the alignment process (S3) mainly performed in the collating unit included in the information processing unit 13 in the authentication mode will be described. This alignment processing (S3) includes a curvature feature point in the authentication data of the applicant who is extracted by the information extraction unit;
By correlating with the curvature feature point of the user who is registered in the storage unit, the direction and size of the tip silhouette of the user who wants to use are corrected, and the user who is registered in the storage unit is used. The purpose is to match the shape close to the tip silhouette. As shown in the flowchart of FIG. 2, the processing content includes three processes including a curvature feature point association process (S31), a coincidence determination process (S32), and a tip silhouette association process (S33). .

First, in step S31, the curvature value of the curvature feature point in the authentication data of the use applicant inputted from the information extraction unit of the information processing unit 13 and the curvature feature point of the use-permitted person registered in the storage unit. And calculates the degree of coincidence by using the reciprocal of the distance between them or the correlation value. If the numbers of the curvature feature points are different, matching is performed by shifting the position of the curvature feature point so that the degree of coincidence becomes highest. In this shifted matching, m consecutive elements are extracted from a vector having a small number of dimensions (the number of dimensions m) and a vector having a large number of dimensions (the number of dimensions n, n> m), and the degree of coincidence between the two vectors is obtained. . This processing is performed as many times as possible (n−m + 1) while sequentially shifting the positions of the curvature feature points, and the correspondence that gives the maximum matching degree among the obtained matching degrees is obtained. If the numbers of curvature feature points match, only one matching is needed.

The specific processing contents of the curvature feature point association processing (S31) will be described with reference to the examples of FIGS. 8 (a) and 8 (b). FIG. 8A shows curvature characteristic points (1 to 9: 1 to 9) of registration data registered in the storage unit.
9 are extracted and shown. On the other hand, in FIG. 8B, seven curvature feature points (indicated by 1 to 7: enclosed in squares) are used as authentication data extracted from the authentication image data of the applicant by the information extraction unit. It is shown. In this example, the authentication data extracted by the information extraction unit has only seven curvature feature points.
Among these curvature feature points, seven consecutive curvature feature points are selected from the curvature feature points of the registered data in accordance with the smaller number of curvature feature points. The degree of coincidence between the curvature values of the selected seven curvature characteristic points and the curvature values of the curvature characteristic points of the authentication data is calculated. In this case, there are three combinations for selecting seven consecutive curvature feature points from the nine curvature feature points in the registered data. Excluding the condition that the curvature feature points are continuous, any combination of seven curvature feature points can be used without changing the arrangement order. Are available in 36 combinations. Which of these is used is determined in consideration of the processing capability of the personal authentication device.

Next, in step S32, it is determined whether or not the degree of coincidence obtained in step S31 is greater than a preset threshold value.
If the obtained degree of coincidence is equal to or smaller than the set threshold value, it is determined that the tip silhouette cannot be associated in the next step.
Move to rejection processing of and reject the authentication. On the other hand, if the obtained degree of coincidence exceeds the set threshold value, the flow shifts to the next step S33 for associating the tip silhouette. In step S33, a process of linearly transforming the tip silhouette of the authentication data based on the correspondence between the curvature feature points obtained in step S31 and superimposing the tip silhouette on the registration data is performed. In this superimposition process, as shown in FIGS. 9A and 9B, the positions of the curvature feature points 1 and 7 at both end points of the curvature feature point group of the authentication data correspond to the curvature feature points of the corresponding registration data. A linear conversion process is performed on each curvature feature point so as to match the positions of the curvature feature points 2 and 8 at both end points of the point group. By this linear conversion processing, the size and the direction of the tip silhouette of both the authentication data and the registration data are substantially in the same state. Information on the tip silhouette of both data in this state is transmitted to the matching unit.

Next, a description will be given of an authentication determination process (S4) mainly performed in the collating unit and the output unit in the authentication mode. This authentication judgment processing (S4)
In the above-mentioned alignment processing (S3), the front end silhouettes of the associated authentication data and the registered data are compared and collated, and it is determined whether or not both front end silhouettes match. The purpose is to determine whether or not. As shown in the flowchart of FIG. 2, the content of the processing is the coincidence calculation processing (S4
1) and two processes mainly including a coincidence determination process (S42).

First, in step S41, the degree of coincidence between the tip silhouette of the authentication data of the applicant and the tip silhouette of the registered data is calculated. FIG. 10 shows a calculation example in the case where the reciprocal of the distance between each other is used as the matching degree. In FIG. 10, the part shown by the solid line is the tip silhouette 26 of the registration data, and the part shown by the dotted line is the tip silhouette 27 of the authentication data. The hatched portion is a portion where the shape is different between the two tip silhouettes, and indicates a portion where the distance between the two is large. Therefore, the smaller the area of the hatched portion, the higher the degree of coincidence between the two tip silhouettes. This area can be determined, for example, as the sum of the differences between the Y coordinates of the two tip silhouettes at each point on the X axis. In this embodiment, as described above, the reciprocal of the distance between the tip silhouettes is used as a measure of the degree of coincidence of the tip silhouettes. However, the present invention is not limited to this. It is also possible to use a correlation value or the like indicating the degree of similarity between them as a measure of the degree of coincidence.

Next, in step S42, it is determined whether or not the value of the degree of coincidence of the tip silhouette calculated in step S41 is larger than a predetermined threshold value. If the value of the matching degree exceeds the set threshold value, it is determined that the tip silhouettes of both the authentication data and the registration data of the user who wish to use match, and the process proceeds to step S43 to use the data. Accept the applicant's certification. If the calculated degree of coincidence is equal to or smaller than the set threshold value, it is determined that the two tip silhouettes do not coincide with each other, and step S44 is performed.
And reject the authentication of the applicant. If the authentication of the user is accepted and the use of the personal computer is accepted, the output unit outputs, for example, a use permission signal or a screen saver cancel signal, and the like. Use of a personal computer becomes possible.

In the above embodiment, the authentication judgment is performed by using the degree of coincidence in the coordinate values of the tip silhouette of the tooth arrangement. However, the present invention is not limited to such an authentication judgment. Instead of the tip silhouette of the tooth arrangement, it is also possible to use three-dimensional information based on unevenness of the tooth surface photographed using a stereo camera or a range finder. Furthermore, it is also possible to extract not only the tip of the tooth arrangement but also the shape of the tooth including the boundary between the teeth or the entire shape of the tooth arrangement as a contour image, and perform authentication using this contour image. Of course, it is needless to say that these features can be appropriately combined and used.

Further, in order to increase the efficiency of the authentication processing procedure in the above embodiment, a threshold value used as a reference value in the matching degree judgment processing at the time of authentication is set in advance for each registrant. It is registered in the storage unit together with the ID number and the registration data, and at the time of authentication, by selecting the target registration data through the collation determination of the ID number, the authentication time is set using the registration threshold value of the registrant. Can be shortened, and the authentication accuracy can be improved.

In the above embodiment, the case where the present invention is applied to the use management of a personal computer is described as an example. It goes without saying that the present invention can be effectively applied to a wide range of personal authentication systems including those for authentication.

[0041]

As described above, according to the present invention, the information on the shape of the teeth which does not easily fluctuate at the time of capturing the authentication image is used as the feature amount for the personal authentication, thereby stabilizing the image information for authentication. In this case, it is possible to obtain the information accurately, and the authentication accuracy of the personal authentication can be improved.

Further, in the authentication image data, since the contrast between the arrangement of the teeth and the region at the back of the mouth is high, there is an effect that the authentication feature can be stably extracted without applying special illumination. .

[Brief description of the drawings]

FIG. 1 is a block diagram showing a configuration of an embodiment of a personal authentication device of the present invention.

FIG. 2 is a diagram showing a flowchart of an embodiment of the personal authentication method of the present invention.

FIG. 3 is a diagram showing image information shown as guidance used for adjusting an opening posture at the time of authentication using a tooth arrangement as a feature amount.

FIG. 4 is a diagram illustrating a pyramid image and a region division state during a region search process.

FIG. 5 is a diagram showing a tooth alignment region cut out by the region search processing.

FIG. 6 is a diagram showing a tip silhouette.

FIG. 7 is a diagram showing extraction of a curvature feature point.

FIG. 8 is a diagram showing the correspondence between curvature feature points.

FIG. 9 is a diagram showing association of a tip silhouette.

FIG. 10 is a diagram showing a degree of coincidence of a tip silhouette.

[Explanation of symbols]

 Reference Signs List 11 CRT display unit 12 CCD imaging unit 13 Information processing unit 14 Keyboard 15 Mouse 21 Auxiliary line 22, 23, 24, 26, 27 Tip silhouette 25 Curvature feature point

Claims (7)

[Claims] An image capturing means for capturing an image of a tooth of a person to be authenticated, and an image captured by the image capturing means is used as authentication image data, and information on a tooth shape of the person to be authenticated is extracted from the authentication image data. Information extraction means, storage means for registering and holding information on the tooth shape of the authorized person, information on the tooth shape of the person to be authenticated extracted from the authentication image data by the information extraction means, Detecting the degree of coincidence of both information by comparing and matching information on the tooth shape of the authorized person registered in the storage means,
An authentication determining unit configured to determine whether to accept or reject authentication of a person to be authenticated based on the detection result. 2. The personal authentication apparatus according to claim 1, wherein the information on the shape of the teeth is information on “teeth alignment”. 3. The personal authentication device according to claim 2, wherein the information on the “teeth alignment” is information on a “teeth tip alignment”. 4. The personal authentication apparatus according to claim 1, wherein the information on the shape of the tooth is information on “unevenness on the surface of the tooth”. 5. The information extracting means receives input of authentication image data relating to the shape of a tooth of a person to be authenticated, which is imaged by the imaging means, and identifies a specific pixel accompanying the tooth shape in the authentication image data. Concatenating to generate a tip silhouette, extracting a point having a large curvature value on the tip silhouette as a curvature feature point, wherein the authentication determination means includes a collation means and an output means therein, and in the collation means, In response to the input of the curvature feature points, the correspondence between the curvature feature points and the curvature feature points corresponding to the information on the tooth shapes registered in the storage unit is determined. It is performed in the held state, and the degree of coincidence between the two is detected by comparing and matching the curvature feature points with each other, and the acceptance or rejection of the authentication of the authenticated person is determined and output through the output means. Personal identification device according to claim 1, 2, 3 or 4, wherein. 6. A first step in which an image of a tooth arrangement of a person to be authenticated is imaged by a predetermined imaging means, and information relating to the tooth arrangement is input as authentication image data. A second step of extracting a feature amount of the tooth arrangement of the authenticator; and a third step of associating the feature amount of the tooth arrangement included in the authentication data with the feature amount of the tooth arrangement included in the registration data of the authentication permitter registered in advance. Receiving the third step, and comparing the characteristic amount of the tooth arrangement included in the authentication data extracted in the second step with the characteristic amount of the tooth arrangement included in the registration data of the authorized person. A fourth step of collating and judging the degree of coincidence of the feature amounts with each other. 7. The method according to claim 1, wherein the second step is a step of extracting a tip end silhouette of the tooth arrangement obtained by connecting pixels having a large difference in luminance value of the image from the authentication image data obtained in the first step. A tip silhouette cutout processing step of extracting as a quantity, a curvature feature point search processing step of finding a curvature at each point on the tip silhouette, and extracting a point whose absolute value of the curvature exceeds a predetermined threshold value as a curvature feature point And wherein the third step includes a curvature feature point of a tip silhouette of a row of teeth of a person included in the authentication data obtained in the curvature feature point search processing step and an authentication included in the registration data. The correspondence with the curvature feature point of the tip silhouette of the tooth arrangement of the acceptor is performed in a state where the order relationship between the curvature feature points is maintained. A curvature feature point associating processing step, and whether or not the value of the degree of coincidence of the curvature values of the two curvature feature points associated in the curvature feature point associating processing step exceeds a preset threshold value; Is determined, if it exceeds, it is determined that the authentication data of the authenticated person corresponds to the registration data of the authorized person, and the process proceeds to the next step. A curvature feature point determination processing step of determining that the tip silhouette is not compatible and rejecting the authentication of the subject; and in the curvature feature point determination processing, the authentication data of the subject is registered as an authentication permitter. When it is determined that the correspondence is established with the data, the tip silhouette of the authentication data is linearly converted so that the curvature feature points having the correspondence overlap each other, and the mutual 7. The personal authentication method according to claim 6, further comprising: a tip silhouette associating processing step of substantially matching silhouettes.