JPH06111006A - Individual identifying device - Google Patents

Abstract

PURPOSE:To dispense with storage of an identification number, etc., by limiting a fingerprint to be compared and collated by using a voice of a person to be identified, at the time of identifying an individual by using a fingerprint. CONSTITUTION:The device is provided with a means 10 for fetching a voice of a person to be identified, a means 1 for collating an input voice and a registered voice by using a registered voice file in which a voice of a specific person conforming to a certain condition is registered, a means 2 for fetching a registered fingerprint of the person selected by the means 1 from the file concerned in a registered fingerprint file group 102 in which the fingerprint of the specific person conforming to a certain condition is registered, and collating it with an input fingerprint inputted by a fingerprint input means 11, and a means 3 for using jointly the means 1 and the means 2, and executing a final discrimination of an individual. In such a way, there is an effect for dispensing with storage of discrimination, etc., for decreasing the number of registered fingerprints of an object to be compared, by using a voice of a person to be identified.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an individual identification device for identifying an individual.

[0002]

2. Description of the Related Art Fingerprints have long been used for identifying individuals in criminal investigations, and recently they have begun to be used for entrance control of places requiring confidentiality.

When used in criminal investigations, the individual's identification is made by comparing the suspect's fingerprint with all previously registered fingerprints. In addition, when it is used for entrance control in a sensitive place, comparing and collating the input fingerprint with all registered fingerprints is the same as in criminal investigation, so there is a time problem. By taking out the corresponding fingerprints from the registered fingerprint group by inputting, for example, from the keyboard and limiting the comparison target fingerprints, it is possible to avoid time-consuming collation of fingerprints.

In the method of comparing and collating the input fingerprint with all the registered fingerprints, the collation time increases when the number of registered fingerprints is enormous, and the registered fingerprints to be compared are limited by using the individual identification number. Then, it is necessary to remember the identification number, and in the case of forgetting it, the only situation is to compare and collate with all registered fingerprints.

[0005]

In such a conventional example, the input fingerprint and all the registered fingerprints are compared and collated with the knowledge that the collation time is increased, or the identification number is always used to shorten the collation time. There is no choice but to limit the registered fingerprints that are input and compared. There is a problem that it takes time to identify individuals,
Also, having to remember the identification number has a problem in terms of simplification.

The present invention solves such a problem, and avoids comparison / collation with all registered fingerprints, and uses the voice of the identification target person as a method of limiting the registered fingerprints to be compared. The purpose is to simplify the identification work by eliminating the need for memory.

[0007]

SUMMARY OF THE INVENTION The present invention provides a personal identification device provided with fingerprint input means for converting a given fingerprint of a person to be identified into a first feature amount corresponding to this fingerprint. Voice input means for converting a person's voice into a second feature amount corresponding to this voice, and a third feature amount corresponding to the voice of a specific person satisfying a predetermined condition and capable of being compared and collated with the second feature amount. Is registered in advance, a voice collating means for collating the second characteristic amount with the third characteristic amount, and the first characteristic amount corresponding to a fingerprint of a specific person who satisfies a predetermined condition. A plurality of registered fingerprint files in which the fourth feature amount capable of comparison and matching is registered in advance using the third feature amount as an index, and the second feature amount and the third feature amount are matched by the voice matching means. Is detected A fingerprint collation unit that selects a registered fingerprint file that uses the matched third feature amount as an index and extracts the fourth feature amount that matches the first feature amount from the registered fingerprint file. And

Here, in the registered fingerprint file, the fourth feature amount may be registered in the order of the similarity of the third feature amount which is an index with respect to the second feature amount.

Further, it is preferable to provide personal identification means for re-matching the fourth feature quantity extracted by the fingerprint matching means with the first feature quantity converted by the fingerprint inputting means.

[0010]

[Operation] The voice input means captures the voice of the person to be identified,
The voice collating means collates the input voice and the registered voice using a registered voice file in which a voice of a specific person who meets a certain condition is registered. The fingerprint matching means is the voice matching means 1
The registered fingerprint of the person selected in step 1 is taken out from the corresponding file in the registered fingerprint file group in which the fingerprint of the specific person meeting certain conditions is registered, and collated with the input fingerprint input by the fingerprint input means. The personal identification means finally identifies the individual by using the results of the voice verification means and the fingerprint verification means.

[0011]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described below with reference to the drawings. FIG. 1 shows the configuration of this embodiment. As shown in FIG. 1, in this embodiment, a voice input means 10 for inputting a voice of a person to be identified and a registered voice file 10 in which a voice of a specific person who meets a certain condition is registered in advance.
1, a voice collating means 1 for collating an input voice with a registered voice, and a fingerprint input means 11 for inputting a fingerprint of a person to be identified.
And a registered fingerprint file group 102 in which a fingerprint of a specific person meeting certain conditions is registered in advance, and a fingerprint matching means 2 for matching the registered fingerprint of the person selected by the voice matching means 1 with the input fingerprint. The personal identification means 3 for finally identifying an individual based on the results of the voice verification means 1 and the fingerprint verification means 2 described above. As shown in FIG. 1, this embodiment includes fingerprint input means 11 for converting a given fingerprint of an identification target person into a first feature amount corresponding to this fingerprint, and further as a feature of the present invention. , A voice input means 10 for converting the given voice of the person to be identified into a second feature amount corresponding to this voice, and a comparison with the second feature amount corresponding to the voice of a specific person who satisfies a predetermined condition. A registered voice file 101 in which a third feature amount capable of performing the above is registered in advance, a voice collating means 1 for collating the second feature amount with the third feature amount, and a fingerprint of a specific person who satisfies a predetermined condition. The registered fingerprint file group 102, which is a plurality of registered fingerprint files registered in advance with the third feature amount as an index, which corresponds to the first feature amount and can be compared and compared with the first feature amount, and a voice collating means. In 1 When the serial match the second feature amount and the third feature amount is detected,
The fingerprint collation unit 2 that selects the registered fingerprint file that uses the matched third feature amount as an index and extracts the fourth feature amount that matches the first feature amount from the registered fingerprint file, and the fingerprint collation unit 2. The personal identification means 3 is provided for re-matching the extracted fourth characteristic amount with the first characteristic amount converted by the fingerprint input means 11.

Next, the operation of this embodiment will be described. Figure 2
6 is a flow chart showing the processing of the individual identification method of the present embodiment.
First, as shown in FIG. 3, the voice collating means (step S21) compares the voices of the identification target person (speaking the surname of the identification target person) captured by the voice inputting means 10 with each other.
What is converted into a feature amount that can be collated (hereinafter referred to as input voice) is obtained (step S31). Then, in order to check whether the input voice is registered in the registered voice file 101, the input voice and the registered voice in the registered voice file are compared (step S33), and if they are not equal, the next registered voice is obtained (step S33). S34), and the comparison is performed again. This comparison process is continued until an equal registered voice is found or compared with all the registered voices in the registered voice file 101 (step S32). If the same registered voice is not found after all comparisons, the identification is rejected. In the registered voice file 101 used here, a voice of a specific person who satisfies a certain condition (a voice of a surname of a specific person who satisfies a certain condition) is used as a feature amount that can be compared and collated as in the input voice. The converted one (hereinafter referred to as a registered voice) is stored. Then, in the fingerprint collating means (step S22), when the input voice is found in the registered voice, as shown in FIG. 4, the registered fingerprint file group 1
One of the registered fingerprint files in which the registered fingerprint corresponding to the input voice is stored is selected from 02 (step S4
1) The corresponding registered fingerprint is taken out from the file (step S42). The registered fingerprint file used here is
Similar to the input voice, the fingerprints converted into feature quantities that can be compared and collated (hereinafter referred to as registered fingerprints) are stored in the order of high similarity of voices. Next, in order to check whether the input fingerprint is registered in the registered fingerprint file, the input fingerprint and the registered fingerprint in the registered fingerprint file are compared (step S44). If they are not equal, the next registered fingerprint in the same registered fingerprint file is compared. Is obtained (step S45), and the comparison is performed again. This comparison process continues until either an equal enrollment fingerprint is found or compared to all enrollment fingerprints in the selected enrollment fingerprint file (step S43).
If no identical registered fingerprint is found after all comparisons, the identification is rejected.

Finally, the personal identification means 23 is performed only when the voice collation means 21 and the fingerprint collation means 22 do not reject the identification, and finally identifies the individual and ends the processing.

[0014]

The present invention compares all registered fingerprints.
Collation can be avoided, and by using the voice of the person to be identified, there is an effect that storage of an identification number or the like for limiting the registered fingerprint to be compared becomes unnecessary.

[Brief description of drawings]

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

FIG. 2 is a flowchart showing the operation of the embodiment of the present invention.

FIG. 3 is a flowchart showing the operation of the embodiment of the present invention.

FIG. 4 is a flowchart showing the operation of the embodiment of the present invention.

[Explanation of symbols]

 1 Voice Matching Means 2 Fingerprint Matching Means 3 Personal Identification Means 10 Voice Input Means 11 Fingerprint Input Means 101 Registered Voice Files 102 Registered Fingerprint Files

─────────────────────────────────────────────────── ─── Continuation of front page (51) Int.Cl. ⁵ Identification code Office reference number FI technical display location G10L 3/00 531 L 7627-5H

Claims (3)

[Claims] 1. A personal identification device having a fingerprint input means for converting a given fingerprint of an identification person into a first feature amount corresponding to this fingerprint, wherein a given identification person's voice corresponds to this voice. And a registered voice in which a third feature amount corresponding to a voice of a specific person who satisfies a predetermined condition and capable of being compared and collated with the second feature amount is registered in advance. A file, a voice collating means for collating the second characteristic amount with the third characteristic amount, and a fourth characteristic capable of comparing and collating with the first characteristic amount corresponding to a fingerprint of a specific person who satisfies a predetermined condition. A plurality of registered fingerprint files whose amount is registered in advance using the third feature amount as an index, and when the voice matching means detects a match between the second feature amount and the third feature amount, Matched third special Select registered fingerprint file to the amount as the index, the personal identification device characterized by comprising a fingerprint verification means for extracting the fourth feature value matching the first feature amount from the registered fingerprint file. 2. The personal identification device according to claim 1, wherein in the registered fingerprint file, the fourth characteristic amount is registered in order of similarity of the third characteristic amount that is an index with respect to the second characteristic amount. 3. The personal identification device according to claim 1, further comprising personal identification means for re-collating the fourth feature quantity extracted by the fingerprint collating means with the first feature quantity converted by the fingerprint input means.