JP2003016433A - Method, system, and program for living-body identification by fingerprint collation - Google Patents

Abstract

PROBLEM TO BE SOLVED: To decide whether a finger whose fingerprint is to be taken is a living body when a fingerprint collation is carried out. SOLUTION: A system is equipped with a finger information input part 1 which inputs the surface temperature and fine vibration of the finger due to a bloodstream as biological information together with a fingerprint image, a temperature decision part 3 which shifts to a frequency component calculating part when the surface temperature of the finger due to the bloodstream among the biological information is within the range of the temperature of the living body, and stops processing thereafter, when it is out of the range, the frequency component calculating part 4 which calculates a frequency spectrum by analyzing the frequency component of the fine vibration of the finger due to the bloodstream in the biological information, and a biological signal identification part 5 which starts fingerprint collation processing when the pattern of the frequency spectrum is a frequency spectrum pattern unique to the living body and quits subsequent processing when not.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a biometric identification method, method, and program in fingerprint collation, and more particularly, a biometric identification method, method, and program in fingerprint collation for determining whether or not a finger collecting a fingerprint is a living body. Regarding

[0002]

2. Description of the Related Art Recently, fingerprint authentication has been attracting attention from the viewpoint of the importance of security by networking and the fact that simple and accurate personal authentication can be performed.
Each fingerprint has a completely different pattern, and since it is immutable throughout life, there is a possibility that it can be a security system that can be used as a key.

By the way, the method of collecting a fingerprint pattern of a fingerprint authentication device that has been commercialized in the past is an optical method for detecting unevenness of a fingerprint by an optical sensor as a difference in light reflectance.
Capacitance methods that detect unevenness of fingerprints as differences in charge accumulation, methods that use ultrasonic waves, electric fields, etc. are generally used. However, these methods all use pattern recognition on the fingertip surface. ing.

However, these are types of image recognition, and there is no guarantee that they cannot be copied.
For example, “Technical Report of IEICE, Vol.100,
No.213, ISEC2000-45, pp.159-166, July 25, 2000, Institute of Electronics, Information and Communication Engineers, ISSN0913-5685. " We artificially created a finger, and experimentally verified whether or not the artificial finger was accepted using the 9 different fingerprint matching devices we obtained, and found that all the devices accepted the gummy artificial finger. It turned out. " That is,
In the current fingerprint collation device, it cannot be said that sufficient measures are taken against a stolen fingerprint, and further improvement in security is required.

[0005]

As described above, in the conventional fingerprint collation device, the fingerprint of the artificial finger is recognized in the same manner as the biometric fingerprint, so it is difficult to distinguish the replicas and eliminate them. is there. In particular, when personal authentication is performed using fingerprint matching, there is a drawback in that the risk of misuse by a replica or the like cannot be avoided.

That is, the personal identification technology using fingerprint collation is drawing attention as a personal authentication means instead of a password and is being commercialized. However, although there is no need to memorize the password, the convenience is great, but there is also a risk that it will be stolen by a replica or the like.

In order to avoid the above-mentioned problems, an object of the present invention is to provide a means for determining whether or not a finger for collecting a fingerprint is a living body when performing fingerprint matching. To obtain the identification method, method, and program.

[0008]

A biometric identification method in fingerprint collation according to the present invention obtains biometric information including weak vibration due to blood flow of a finger prior to fingerprint collation processing, analyzes the biometric information, and The fingerprint collation process is started when it is determined that the finger is a living body.

Further, in the biometric identification method for fingerprint collation according to the present invention, the biometric information is the surface temperature and weak vibration of the finger due to blood flow, and when the surface temperature is within the range of the body temperature of the living body, the weak vibration is a frequency. The analysis is performed to calculate a frequency spectrum, and the fingerprint matching process is started when it is determined that the peak of the frequency spectrum is a peak peculiar to a living body.

Further, the biometric identification method in fingerprint collation of the present invention includes a finger information input section having a temperature sensor and a vibration sensor for taking in finger surface temperature and weak vibration due to blood flow as biometric information together with the fingerprint image, and the fingerprint image. A feature extraction unit that extracts fingerprint data indicating a fingerprint feature point from the fingerprint data and sends it to the fingerprint information database; and a frequency component calculation unit when the surface temperature of the finger due to blood flow is within the range of the body temperature of the living body among the biological information. When the surface temperature is out of the range of the temperature of the living body, the temperature determination unit that stops the subsequent processing, and the frequency component of the weak vibration of the finger due to the blood flow in the biological information is analyzed to calculate the frequency spectrum. A frequency component calculator, and when the frequency spectrum pattern is a frequency spectrum pattern peculiar to a living body, starts fingerprint matching processing, Pattern of the serial frequency spectrum is characterized by comprising a biological signal identification unit to stop subsequent processing when not a biological characteristic of the frequency spectrum pattern.

Further, in the biometric identification method for fingerprint collation according to the present invention, the frequency component calculation unit calculates a frequency spectrum of weak vibration of a finger due to blood flow by fast Fourier transform.

Further, in the biometric identification system for fingerprint collation of the present invention, the biometric signal identification section starts the fingerprint collation processing when the peak of the frequency spectrum is the peak of the frequency spectrum peculiar to the living body, and the peak of the frequency spectrum. When is not the peak of the frequency spectrum peculiar to the living body, the subsequent processing is stopped.

Further, the biometric identification program for fingerprint collation according to the present invention shows a finger information input step of capturing the surface temperature and weak vibration of the finger due to blood flow as biometric information together with the fingerprint image, and the fingerprint feature points from the fingerprint image. A feature extraction step of extracting fingerprint data and sending it to a fingerprint information database; and moving to a frequency component calculation step when the surface temperature of the finger due to blood flow in the biological information is within the range of the body temperature of the living body, the surface temperature is calculated. When the temperature is outside the range of the temperature of the living body, a temperature determination step of stopping the subsequent processing, a frequency component calculation step of calculating a frequency spectrum by analyzing the frequency component of the weak vibration of the finger due to the blood flow in the biological information, When the frequency spectrum pattern is a frequency spectrum pattern peculiar to a living body, the fingerprint matching process is started. The frequency spectrum of the pattern is characterized in that it comprises a biological signal identification step to stop subsequent processing when not a biological characteristic of the frequency spectrum pattern.

Further, in the biometric identification program for fingerprint collation according to the present invention, the frequency component calculating step calculates the frequency spectrum of the weak vibration of the finger due to the blood flow by the fast Fourier transform.

Further, in the biometric identification program for fingerprint collation according to the present invention, the biometric signal identifying step starts the fingerprint collation processing when the peak of the frequency spectrum is the peak of the frequency spectrum peculiar to a living body, and the peak of the frequency spectrum. When is not the peak of the frequency spectrum peculiar to the living body, the subsequent processing is stopped.

That is, according to the present invention, in the fingerprint collation device, biometric information is obtained from the surface temperature of the finger surface due to blood flow and the vibration frequency component due to blood flow, and fingerprint authentication is performed after identifying the biometric finger or replica. . Therefore, it becomes possible to prevent the replica from being stolen in the fingerprint matching device, and the security of the fingerprint matching device can be improved.

The fingertips of the living body have subtle vibrations due to muscular movements, weak vibrations due to blood flow, and fingertip movements (vibrations) caused by living human beings, and these are detected to calculate the frequency spectrum. Then, by comparing the characteristic "whether it is within a certain range (whether it is peculiar to the living body)", it is possible to identify whether the finger that picked up the fingerprint is a living body or a replica. Improved personal authentication can be performed.

[0018]

DETAILED DESCRIPTION OF THE INVENTION The present invention will be described below with reference to the drawings.

FIG. 1 is an explanatory view showing an embodiment of the present invention. In the figure, a biometric identification method in fingerprint collation according to the present invention includes a finger information input unit 1 including a temperature sensor and a vibration sensor that capture the surface temperature and weak vibration of a finger due to blood flow as biometric information together with a fingerprint image, and the fingerprint image. A feature extraction unit 2 that extracts fingerprint data indicating fingerprint feature points from the fingerprint data and sends it to the fingerprint information database 8;
Of the biological information, when the surface temperature of the finger due to the blood flow is within the range of the body temperature of the living body, the temperature component shifts to the frequency component calculation unit, and when the surface temperature is outside the range of the temperature of the living body, a temperature determination unit that suspends the subsequent processing. 3, a frequency component calculation unit 4 that analyzes a frequency component of weak vibration of a finger due to blood flow in the biometric information and calculates a frequency spectrum, and fingerprint collation when the frequency spectrum pattern is a frequency spectrum pattern peculiar to a living body. The biological signal identifying section 5 which starts the processing and stops the subsequent processing when the frequency spectrum pattern is not the frequency spectrum pattern peculiar to the living body.
And.

When the biometric signal identifying section 5 determines that the finger is a living body, the fingerprint matching section 6 refers to the fingerprint information database 8 to perform fingerprint matching processing. Furthermore, the collation determination unit 7 confirms the result of the above fingerprint collation processing and makes a determination.

The finger information input section 1 is provided with a temperature sensor and a vibration sensor, and simultaneously acquires a fingerprint image, observes the surface temperature of the finger with the temperature sensor, and observes the frequency of weak vibration of the finger with the vibration sensor.

The feature extraction unit 2 extracts fingerprint data from the above fingerprint image and registers it in the fingerprint information database 8. At the same time, the above-mentioned surface temperature and weak vibration are detected.

The temperature determination unit 3 determines whether or not the finger placed on the finger information input unit 1 is a living body based on the temperature. In order not to complicate the process, the temperature determination unit 3 targets the temperature (30 to 40 ° C.) that can be the body temperature of the human body, and evaluates only whether or not the detected temperature is within this range.

If the detected temperature is out of this range, the process is forcibly terminated and it is determined that the placed finger is not a living finger. If the detected temperature is 30-40
If it is within the range of ° C, the process shifts to the process of biometric identification using weak vibration due to blood flow.

When it is determined that the surface temperature of the finger placed on the finger information input unit 1 is within the range that can be the temperature of the human body, the frequency component calculation unit 4 calculates the frequency component of weak vibration due to blood flow, The frequency spectrum is calculated by the Fourier function.

The biometric signal identifying section 5 determines whether or not the placed finger is a living body by comparing whether or not the calculated peak of the frequency spectrum is a peak peculiar to the living body.

When the biometric signal identifying unit 5 does not obtain a peak peculiar to a living body, that is, when the living body is not a living body, the processing is forcibly terminated and the fingerprint collation is not performed. When the biological signal identifying unit 5 obtains a peak peculiar to a living body and determines that the living body is a living body, the fingerprint matching unit 6 performs matching with the fingerprint information registered in the fingerprint information database 8 in advance.

As described above, the security of the fingerprint collation apparatus can be improved by performing the double check of the surface temperature of the finger and the biometric identification based on the weak vibration before performing the fingerprint collation.

Next, the operation of the above biometric identification method will be described.

First, the surface temperature of the finger captured by the finger information input unit 1 is evaluated by the temperature determination unit 3. In the temperature determination unit 3, the surface temperature can be the temperature of the human body.
Evaluate whether it falls within the range of 40 ° C.

If the temperature taken in by the finger information input section 1 deviates from this range, it is determined that the finger placed in the finger information input section 1 is not a living body, and the processing is forcibly terminated. To do. If the sensed temperature is within this range,
The finger is determined to be a living body, and the process of the frequency component calculation unit 4 is performed.

In the frequency component calculation unit 4, the finger information input unit 1
Fourier transform the frequency of the weak vibration of the fingertip obtained in
A spectrum of frequency components as shown in FIG. 2 is calculated.
The fingertips of the living body should have weak vibrations due to blood flow, etc., and the biological signal identifying unit 5 determines whether or not the living body is a living body by looking at the spectrum pattern calculated here as the biological information.

Here, when the finger of the replica is placed on the finger information input section 1, a weak vibration due to the blood flow does not occur on the surface of the replica, and even if artificially trying to cause it, a spectrum pattern peculiar to the living body is obtained. Hard to get.

The biological signal identifying section 5 determines whether or not the spectrum pattern calculated by the frequency component calculating section 4 is peculiar to a living body. If it is not peculiar to the living body, the process is forcibly terminated. If it is peculiar to the living body, the process proceeds to the fingerprint matching unit 6.

For the first time, the finger placed on the finger information input section 1 is judged to be a finger having a temperature close to the body temperature (30 to 40 ° C.) and having vibration peculiar to the living body, and the finger of the living body is determined. Only the fingerprint collation unit 6 will be processed.

The fingerprint collation unit 6 collates the fingerprint information registered in advance in the fingerprint information database 8 with the fingerprint data fetched from the finger information input unit 1 via the feature extraction unit 2 by the method conventionally proposed. Then, it is possible to perform authentication with higher security than before.

By adding frequency information to the registered fingerprint data registered in the fingerprint information database 8 at the same time (minutia + frequency spectrum pattern),
The accuracy of personal authentication can be further improved.

Further, by incorporating the frequency biometric identification function in the fingerprint authentication unit, it is possible to realize a security system / fingerprint authentication system having high safety.

The biometric identification method in the fingerprint collation described above operates by executing a program stored in the main memory (not shown) of the central processing unit.

[0040]

As described above in detail, according to the present invention, the following effects can be obtained.

The first effect is that by combining the temperature of the fingertip and the frequency information of the fingertip, the security of the authentication device using fingerprint collation can be significantly improved.

The second effect is that the frequency component of the weak frequency due to the blood flow of the fingertip is used as the biological information, making it difficult to make a replica and significantly reducing the risk of plagiarism.

[Brief description of drawings]

FIG. 1 is an explanatory diagram showing an embodiment of the present invention.

FIG. 2 is an explanatory diagram showing an example of a spectrum pattern of weak vibration.

[Explanation of symbols]

1 Finger information input section 2 Feature extraction unit 3 Temperature determination unit 4 Frequency component calculator 5 Biometric signal identification section 6 Fingerprint matching unit 7 Collation judging section 8 Fingerprint information database

Claims (8)

[Claims] 1. Prior to the fingerprint matching process, biometric information including weak vibration due to blood flow of a finger is acquired, the biometric information is analyzed, and the fingerprint matching process is performed when the finger is determined to be a living body based on the result. A biometric identification method in fingerprint matching, characterized by starting. 2. The biometric identification method in fingerprint matching according to claim 1, wherein the biometric information is a finger surface temperature and weak vibration due to blood flow, and when the surface temperature is within a range of a body temperature of the living body, the weak vibration is obtained. A frequency analysis is performed to calculate a frequency spectrum, and when it is determined that the peak of the frequency spectrum is a peak peculiar to a living body, a fingerprint matching process is started, and a biometric identification method in fingerprint matching. 3. A finger information input unit including a temperature sensor and a vibration sensor that captures surface temperature and weak vibration of a finger due to blood flow as biometric information together with a fingerprint image, and fingerprint data indicating a fingerprint feature point is extracted from the fingerprint image. If the surface temperature of the finger due to blood flow in the biological information is within the range of the body temperature of the living body, the frequency component calculator moves to the frequency component calculating unit, and the surface temperature is the temperature of the living body. When the temperature is out of the range, the temperature determination unit that stops the subsequent processing, the frequency component calculation unit that calculates the frequency spectrum by analyzing the frequency component of the weak vibration of the finger due to the blood flow in the biological information, and the pattern of the frequency spectrum Is a frequency spectrum pattern peculiar to a living body, fingerprint matching processing is started, and the frequency spectrum pattern is a frequency spectrum peculiar to a living body. A biometric identification method in fingerprint collation, comprising: a biometric signal identification unit that suspends subsequent processing when the pattern is not a several-spectrum pattern. 4. The biometric identification method for fingerprint collation according to claim 3, wherein the frequency component calculation unit calculates a frequency spectrum of weak vibration of the finger due to blood flow by fast Fourier transform. Identification method. 5. The biometric identification method for fingerprint matching according to claim 3 or 4, wherein the biometric signal identification unit starts fingerprint verification processing when the peak of the frequency spectrum is a peak of a frequency spectrum unique to a living body, A biometric identification method in fingerprint matching, characterized in that when the peak of the frequency spectrum is not the peak of the frequency spectrum peculiar to a living body, the subsequent processing is stopped. 6. A finger information input step of taking in the surface temperature and weak vibration of a finger due to blood flow as biometric information together with a fingerprint image, and extracting fingerprint data indicating a characteristic point of the fingerprint from the fingerprint image and transferring it to a fingerprint information database. When the surface temperature of the finger due to the blood flow in the biological information is within the range of the body temperature of the living body, the process proceeds to the frequency component calculation step. When the surface temperature is outside the range of the temperature of the living body, A temperature determination step of stopping the process, a frequency component calculation step of calculating a frequency spectrum by analyzing frequency components of weak vibration of the finger due to blood flow in the biological information, and a frequency spectrum pattern in which the frequency spectrum pattern is peculiar to a living body. If it is, the fingerprint matching process is started, and the frequency spectrum pattern has A biometric identification program in fingerprint collation, comprising a biometric signal identification step of stopping the subsequent processing when it is not a vector pattern. 7. The biometric identification program for fingerprint matching according to claim 6, wherein the frequency component calculating step calculates a frequency spectrum of weak vibration of the finger due to blood flow by fast Fourier transform. Identification program. 8. The biometric identification program for fingerprint collation according to claim 6 or 7, wherein the biometric signal identifying step starts fingerprint collation processing when the peak of the frequency spectrum is a peak of a frequency spectrum peculiar to a living body, A biometric identification program in fingerprint matching, characterized in that when the peak of the frequency spectrum is not the peak of the frequency spectrum peculiar to a living body, the subsequent processing is stopped.