JP2009003511A - Collation device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To correctly achieve collation even when an angle by which a user places his or her finger is deviated. <P>SOLUTION: This collation device is provided with a face reading part 1 for acquiring the face image of a user; a face position detection means 5 for detecting a face position from the face image; a database 8 for storing the face image registration data, fingerprint registration data and face position registration data of the user; an angle deviation calculation means 6 for calculating a difference between the data of the face position detected by the face position detection means 5 and the face position registration data of the database 8, and for calculating the deviation of an angle by which the user places his or her finger from a prescribed reference position based on the difference; a finger reading part 2 fixed so as to always have the same position relation to the face reading part 1 for reading the fingerprint data of the user; and a finger authentication means 7 for generating compensation fingerprint data by compensating the fingerprint registration data of the database 8 based on the calculated deviation of the angle, and for collating the read fingerprint data with the compensation fingerprint data for authentication. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to a collation device, and more particularly to a collation device for performing identity verification using a user's fingerprint.

  In recent years, a collation device that collates read fingerprint data with registered fingerprint data for user authentication has been proposed (see, for example, Patent Document 1).

  In the collation device described in Patent Document 1, the three-dimensional shape data of the user's finger is read, collated with the three-dimensional data stored in the data storage unit, and collation is performed when it is determined that they match. Since it did in this way, since fingerprint collation by a finger and fingerprint collation by a non-finger can be distinguished, impersonation is prevented and security is improved.

JP 2006-215975 A

  However, in such a conventional collation device, the structure is such that the finger is placed on the sensor part or the finger is held over the sensor for collating the finger, but the angle at which the user places the finger on the sensor part Since there is a deviation in each case, there is a problem that even if the deviation of the angle at which the finger is placed exceeds a preset allowable range, even the person himself / herself cannot perform the matching process correctly.

  The present invention has been made to solve such a problem, and an object of the present invention is to obtain a collation device that can perform collation correctly even if the angle at which the user places his finger on the sensor is deviated.

  The present invention relates to a face image reading unit that captures a face of a user and obtains image data of the face, a face position detection unit that detects a face position in the read face image data, and a user image captured at the time of registration. A database in which face image registration data, fingerprint registration data, and face position registration data of the face position in the face image registration data are stored in advance, the face position data detected by the face position detection means, and the database Difference calculating means for obtaining a difference from the stored face position registration data; and an angle deviation calculating means for calculating a deviation of an angle at which the user's finger is placed from a predetermined reference position based on the obtained difference. The fingerprint registration data stored in the database is compensated based on the angular deviation calculated by the angular deviation calculating means. Compensation fingerprint data generating means for generating compensated fingerprint data, finger reading means fixed to be always in the same positional relationship with respect to the face image reading means, and reading the fingerprint data of the user, and the read fingerprint It is a collation device comprising finger authentication means for collating data and the generated compensation fingerprint data to authenticate a user.

  The present invention relates to a face image reading unit that captures a face of a user and obtains image data of the face, a face position detection unit that detects a face position in the read face image data, and a user image captured at the time of registration. A database in which face image registration data, fingerprint registration data, and face position registration data of the face position in the face image registration data are stored in advance, the face position data detected by the face position detection means, and the database Difference calculating means for obtaining a difference from the stored face position registration data; and an angle deviation calculating means for calculating a deviation of an angle at which the user's finger is placed from a predetermined reference position based on the obtained difference. The fingerprint registration data stored in the database is compensated based on the angular deviation calculated by the angular deviation calculating means. Compensation fingerprint data generating means for generating compensated fingerprint data, finger reading means fixed to be always in the same positional relationship with respect to the face image reading means, and reading the fingerprint data of the user, and the read fingerprint Since the verification device includes a finger authentication means for verifying the user by verifying the data and the generated compensation fingerprint data, the correct verification is performed even if the angle at which the user places the finger with respect to the sensor is shifted. The effect that it can be performed is acquired.

Embodiment 1 FIG.
A configuration of a collation device according to Embodiment 1 of the present invention will be described with reference to FIG. In FIG. 1, 1 is a face reading unit that captures a user's face and obtains face image data, 2 is a finger reading unit that reads fingerprint data of the user's finger, and 3 is registration of face image data and fingerprint data. A processing unit 8 that performs collation is a database that stores user ID numbers, face image registration data, face position registration data, and fingerprint registration data. Although omitted in FIG. 1, an input means such as a numeric keypad connected to the processing unit 3 for inputting the user ID number may be further provided as necessary.

  Here, in the collation device in the present embodiment, a face collation device that constitutes a face authentication unit that performs collation from face image data and a finger authentication unit that performs collation from a finger fingerprint are configured. The finger collating devices are fixed so as to always have the same positional relationship with each other, and the collation is performed by combining both of these two collating devices. However, the face authentication may be used only for detecting the user's face position. The face authentication is not necessarily performed by comparing the photographed face image data with the face image registration data registered in the database. It is not necessary to determine whether or not there is a high possibility.

  Further, as shown in FIG. 1, the processing unit 3 includes a face authentication unit 4 for collating the read face image data with the registered face image registration data, and the position of the face from the entire face image data. A face position detecting means 5 for detecting the angle, an angle deviation calculating means 6 for calculating the angle deviation of the finger based on the difference between the position of the read face and the registered face position registration data, and the read fingerprint data. Finger authentication means 7 for collating with existing fingerprint registration data is provided.

  Next, the operation of the verification device according to the present embodiment will be described. First, the flow during registration will be described with reference to FIG. First, the user inputs his / her ID number (step S1). The input ID number is registered in the database 8. Thereby, the registration process is started. Note that the ID number does not necessarily have to be entered. By pressing a registration button (not shown) provided in the processing unit 3, a unique ID number unique to the user is automatically assigned. Also good.

  Next, the user places his / her finger on a reading device such as a sensor constituting the finger reading unit 2 in a state in which the face is directed toward an imaging device such as a camera constituting the face reading unit 1. At this time, a predetermined reference position indicating a position where the finger is placed is displayed on the reading device so that the user's fingerprint is registered at the correct position, and if the user does not place the finger at that position, an error occurs. Either end the registration process or read the user's attention by voice or the like and prompt the user to correct the position where the finger is placed. Thereby, the face of the user to be registered is photographed and the fingerprint data of the finger is read. The face image obtained by the face reading unit 1 is converted into data by the face authentication means 4 and registered in the database 8 as face image registration data. Similarly, face position data P (x, y) indicating the position coordinates of the user's face in the entire face image is calculated from the face image by the face position detecting means 5 and registered in the database 8 as face position registration data. . As the face position, the position of one or more predetermined locations of the user's face (for example, the center point of one of the left and right eyes, nose, mouth, etc.) is detected and the face position data P (x, y) is also used. Well, or based on the top point, bottom point, rightmost point, and leftmost point of the detected face part, draw a center line vertically and horizontally with respect to the user's face image, and The intersection may be the face position data P (x, y). The fingerprint data read by the finger reading unit 2 is converted into data by the finger authentication means 7 and registered in the database 8 together with the face image registration data and the face position registration data as fingerprint registration data (step S2). As a result, the registration process ends and the user is registered in the database 8.

  Next, the flow at the time of collation will be described with reference to FIG. First, the user places his / her finger on a reading device such as a sensor constituting the finger reading unit 2 in a state where his / her face is directed to a photographing device such as a camera constituting the face reading unit 1. Thereby, an image of the user's face is photographed by the face reading unit 1, and face position data is detected from the face image read by the face reading unit 1 by the face position detecting unit 5 (step S10). Next, the face authentication means 4 compares the face image data read by the face reading unit 1 with the face image registration data registered in the database 8 and performs authentication processing (step S11). As a result of authentication, if it is determined that the user is not similar and the user is not himself, the process is terminated as it is. On the other hand, if they are coincident or similar and the user is highly likely to be the user himself / herself, the angle deviation calculation means 6 uses the face position registration data P (x, y) registered in the database 8 and the face position detection. A difference from the face position data detected by the means 5 is obtained, and an angle at which the user's finger placed on the finger reading unit 2 is displaced from a predetermined reference position is calculated as an angle deviation θ based on the difference ( Step S12). At this time, it is also detected whether the shifted direction is the right direction or the left direction.

  Next, as fingerprint authentication processing, first, the finger reading unit 2 captures or scans the fingerprint of the user's finger and reads the fingerprint data (step S13). Next, the finger authentication means 7 generates compensated fingerprint data obtained by rotating the fingerprint registration data registered in the database 8 by the angle deviation θ obtained in step S12 (step S14), and then this angle. Compensation fingerprint data rotated by the shift θ is compared with the read fingerprint data for authentication (step S15). As a result of the authentication, if the compensated fingerprint data and the read fingerprint data match (or are similar), it is determined that the user is the user himself / herself. , Output the judgment result. At this time, for a slight deviation, a conventional allowable range is provided, and if it is within the allowable range, it is automatically corrected and collated.

  As described above, in the present embodiment, in the series of processes described above, the angle deviation θ of the angle at which the finger is placed is determined from the difference between the registered face position registration data and the read face position data. By compensating the fingerprint data at the time of finger authentication, authentication is performed without being affected by the angle even in the case where collation was not possible due to the finger angle deviation in the past. The effect that processing can be performed is obtained.

  As a method for calculating the angular deviation θ of the finger from the difference in face position, for example, the following method can be given. Recognize the position of the center point of the black eye of the user's right eye and the center point of the center point of the black eye of the left eye as the face position, and based on those position data, the user's face is directed to the right or left with respect to the front position. Calculate how much angle is rotating in the direction. At this time, it is determined that the angle of the user's finger is also rotated by the same angle in the same direction as the face from the predetermined position, and the angle is calculated as the angle deviation θ. Alternatively, instead of directly using the angle as the angle deviation θ, the angle deviation θ may be obtained by a predetermined function using the value of the angle as an input value.

  As another method, as the face position, the position of one or more predetermined locations (for example, one of the left and right eyes, nose, mouth) of the user's face is detected and registered in the database 8. The data of the face position P (x, y) corresponding to the location is compared, and the angle deviation θ placed by the user's finger is obtained based on the difference. As a method of obtaining, for example, since there is a body wrinkle for each user, at the time of registration, an actual measurement value of the angle deviation θ is detected for each difference of the face position, and the data of the difference and the actual measurement value of the angular deviation θ It is also possible to generate a data table in which these are stored in correspondence with each other and obtain the angle deviation θ from the table.

  In the above description, an example has been described in which the angle deviation θ of the angle of the finger is obtained from the difference in the face position, and the fingerprint verification is performed by rotating the registered fingerprint data by the amount of θ. Not only in that case, but when the facial position difference exceeds a predetermined threshold, the registered fingerprint registration data is rotated in the left and right directions by a predetermined angle proportional to the difference, and if it still cannot be collated, In addition, the fingerprint registration data may be shifted in the back direction and the front direction, and sequentially moved to a position where the fingerprint can be verified.

  In the above description, an example in which finger authentication is performed after performing face authentication has been described, but the order may be reversed.

It is a block diagram which shows the structure of the collation apparatus which concerns on Embodiment 1 of this invention. It is the flowchart which showed the operation | movement at the time of registration of the collation apparatus which concerns on Embodiment 1 of this invention. It is the flowchart which showed the operation | movement at the time of collation of the collation apparatus which concerns on Embodiment 1 of this invention.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 1 face reading part, 2 finger reading part, 3 process part, 4 face authentication means, 5 face position detection means, 6 angle deviation calculation means, 7 finger authentication means, 8 database.

Claims (2)

A face image reading means for photographing a user's face and obtaining image data of the face;
Face position detecting means for detecting a face position in the read image data of the face;
A database in which face image registration data and fingerprint registration data of a user photographed at the time of registration and face position registration data of a face position in the face image registration data are stored in advance;
Difference calculation means for obtaining a difference between the face position data detected by the face position detection means and the face position registration data stored in the database;
Based on the obtained difference, an angle deviation calculating means for calculating a deviation of the angle at which the user's finger is placed from a predetermined reference position;
Compensation fingerprint data generation means for generating compensation fingerprint data that compensates for the fingerprint registration data stored in the database based on the angle deviation calculated by the angle deviation calculation means;
A finger reading unit that is fixed so as to always have the same positional relationship with the face image reading unit, and that reads the fingerprint data of the user;
A collation apparatus comprising: finger authentication means for authenticating a user by collating the read fingerprint data with the generated compensation fingerprint data. Further comprising face authentication means for authenticating a user by comparing the face image data read by the face image reading means with the face image registration data stored in the database;
The collation apparatus according to claim 1, wherein the face authentication unit performs user authentication by the finger authentication unit when it is determined that there is a high possibility that the user is the user himself / herself.

Images