Authentication apparatus

• US 8,270,726 B2
• Filed: 03/12/2009
• Issued: 09/18/2012
• Est. Priority Date: 03/13/2008
• Status: Active Grant

• Current Assignee: Fujitsu Limited
• Original Assignee: Fujitsu Limited
• Inventors: Niinuma, Koichiro, Abiko, Yukihiro
• Primary Examiner(s): Park, Chan S
• Assistant Examiner(s): RICE, ELISA M
• Application Number:
  US12/402,803
• Publication Number:
  US 20090232368A1
• Time in Patent Office: 1,286 Days
• Field of Search: 382/199, 382115-127
• US Class Current: 382/199
• CPC Class Codes: G06V 40/1382 Detecting the live character of the finger, i.e. distinguishing from a fake or cadaver finger

Abstract

An authentication apparatus is provided. A forgery determination threshold is determined on the basis of two types of parameters, a forgery similarity and a forgery difficulty. If a calculated value for an object under test is lower than or equal to the forgery determination threshold, then, it is determined that the object is a biologic object. Thus, easiness in impersonation with a fake of a biologic object that is easy to forge may be reduced, and a false rejection due to a determination in which a biologic object is erroneously determined as a fake may be reduced.

Claims

1. An authentication apparatus for authenticating a person by determining whether input data belongs to a biologic object in accordance with a predefined condition based on biometric data of a genuine person and whether the input data matches with registered biometric data of the genuine person, said authentication apparatus comprising:

a forgery data extractor configured to extract forgery data from the input data, said forgery data being for determining whether the input data belongs to a biologic object or a fake;

a forgery similarity calculator configured to calculate forgery similarity of biometric data of the genuine person, said forgery similarity indicating similarity to a fake;

a forgery difficulty calculator configured to calculate forgery difficulty of biometric data of the genuine person, said forgery difficulty indicating difficulty of forgery;

a determination threshold setter configured to determine a forgery determination threshold on the basis of the forgery similarity and the forgery difficulty;

a forgery score calculator configured to calculate a forgery score of the input data on the basis of extracted forgery data, said forgery score indicating likelihood of forgery; and

a forgery determiner configured to determine whether the input data belongs to a biologic object by comparing calculated forgery score with the forgery determination threshold.

2. The authentication apparatus of claim 1, further comprising:

a matching data extractor configured to extract matching data from the input data, said matching data being for determining whether the input data belongs to a genuine person;

a matching score calculator configured to calculate a matching score by matching extracted matching data with registered matching data when the forgery determiner has determined that the input data belongs to a biologic object, said matching score indicating a degree of coincidence between the extracted matching data and the registered matching data; and

a matching determiner configured to determine whether the extracted matching data matches with the registered matching data by comparing calculated matching score with a matching determination threshold.

3. The authentication apparatus of claim 2, further comprising:

a genuineness determiner configured to determine whether the input data belongs to the genuine person on the basis of determination result from the forgery determiner and determination result from the matching determiner.

4. The authentication apparatus of claim 1, further comprising:

a forgery data storage configured to store registered forgery data,

wherein

said forgery score calculator calculates the forgery score on the basis of, in addition to the extracted forgery data, the registered forgery data stored in the forgery data storage.

5. The authentication apparatus of claim 1, wherein

said determination threshold setter

sets a standard forgery determination threshold as the forgery determination threshold when the forgery similarity is lower than a predefined reference threshold,

calculates, when the forgery similarity is higher than or equal to the predefined reference threshold, on the basis of the forgery difficulty, a forgery determination threshold with which difficulty in impersonation with a fake may be maintained,

sets a calculated forgery determination threshold as the forgery determination threshold when the calculated forgery determination threshold is lower than the standard forgery determination threshold, and

sets the standard forgery determination threshold as the forgery determination threshold when the calculated forgery determination threshold is higher than or equal to the standard forgery determination threshold.

6. The authentication apparatus of claim 1, wherein

said determination threshold setter

sets a standard forgery determination threshold as the forgery determination threshold when the forgery difficulty is higher than a predefined reference threshold,

calculates, when the forgery difficulty is lower than or equal to the predefined reference threshold, on the basis of the forgery similarity, a forgery determination threshold with which difficulty in erroneous determination as a fake may be maintained,

sets a calculated forgery determination threshold as the forgery determination threshold when the calculated forgery determination threshold is higher than or equal to the standard forgery determination threshold, and

sets the standard forgery determination threshold as the forgery determination threshold when the calculated forgery determination threshold is lower than the standard forgery determination threshold.

7. The authentication apparatus of claim 2, wherein

said determination threshold setter further determines the matching determination threshold on the basis of the forgery similarity and the forgery difficulty.

8. The authentication apparatus of claim 7, wherein

said determination threshold setter

sets a standard forgery determination threshold and a standard matching determination threshold as the forgery determination threshold and the matching determination threshold, respectively, when the forgery similarity is lower than a predefined first reference threshold and the forgery difficulty is higher than a predefined second reference threshold,

calculates, on the basis of the forgery similarity, a forgery determination threshold with which difficulty in erroneous determination as a fake may be maintained when the forgery similarity is higher than or equal to the predefined first reference threshold or the forgery difficulty is lower than or equal to the predefined second reference threshold,

sets a calculated forgery determination threshold and the standard matching determination threshold as the forgery determination threshold and the matching determination threshold, respectively, when difficulty in impersonation with a fake may be maintained with the calculated forgery determination threshold,

calculates a matching determination threshold that may maintain difficulty in impersonation with a fake when difficulty in impersonation with a fake may not be maintained with the calculated forgery determination threshold,

sets the calculated forgery determination threshold and the standard matching determination threshold as the forgery determination threshold and the matching determination threshold, respectively, when a calculated matching determination threshold is lower than the standard matching determination threshold, and

sets the calculated forgery determination threshold and the calculated matching determination threshold as the forgery determination threshold and the matching determination threshold, respectively, when the calculated matching determination threshold is higher than or equal to the standard matching determination threshold.

9. The authentication apparatus of claim 1, wherein

said forgery similarity calculator calculates, at registration of biometric data of the genuine person, forgery similarity of biometric data of the genuine person on the basis of a forgery score of the registered biometric data.

10. The authentication apparatus of claim 1, wherein

said forgery similarity calculator calculates, at registration of biometric data of the genuine person, forgery similarity of biometric data of the genuine person on the basis of an average value of forgery scores of a plurality of registered biometric data.

11. The authentication apparatus of claim 2, wherein

said forgery similarity calculator calculates forgery similarity of biometric data of the genuine person on the basis of a forgery score of previous input data of the genuine person, a matching score of said previous input data being higher than or equal to a predefined matching threshold.

12. The authentication apparatus of claim 2, wherein

said forgery similarity calculator calculates forgery similarity of biometric data of the genuine person on the basis of an average value of forgery scores of a plurality of previous input data of the genuine person, matching scores of said plurality of previous input data being higher than or equal to a predefined matching threshold.

13. The authentication apparatus of claim 2, wherein

said forgery difficulty calculator calculates the forgery difficulty of biometric data of the genuine person on the basis of the matching determination threshold.

14. The authentication apparatus of claim 2, wherein

said forgery difficulty calculator calculates the forgery difficulty of biometric data of the genuine person on the basis of the calculated matching score.

15. The authentication apparatus of claim 1, wherein

said biometric data is fingerprint data, and

said forgery difficulty calculator calculates the forgery difficulty of biometric data of the genuine person on the basis of unsharpness of fingerprint ridges.

16. The authentication apparatus of claim 1, wherein

said biometric data is fingerprint data, and

said forgery difficulty calculator calculates the forgery difficulty of biometric data of the genuine person on the basis of state of finger skin.

17. An authentication method executed by an authentication apparatus for authenticating a person by determining whether input data belongs to a biologic object in accordance with a predefined condition based on biometric data of a genuine person and whether the input data matches with registered biometric data of the genuine person, said authentication method comprising:

extracting forgery data from the input data, said forgery data being for determining whether the input data belongs to a biologic object or a fake;

calculating forgery similarity of biometric data of the genuine person, said forgery similarity indicating similarity to a fake;

calculating forgery difficulty of biometric data of the genuine person, said forgery difficulty indicating difficulty of forgery;

determining a forgery determination threshold on the basis of the forgery similarity and the forgery difficulty;

calculating a forgery score of the input data on the basis of extracted forgery data, said forgery score indicating likelihood of forgery; and

determining whether the input data belongs to a biologic object by comparing calculated forgery score with the forgery determination threshold.

18. A non-transitory computer-readable medium storing a program that causes a computer to execute an authentication method for authenticating a person by determining whether input data belongs to a biologic object in accordance with a predefined condition based on biometric data of a genuine person and whether the input data matches with registered biometric data of the genuine person, said authentication method comprising:

extracting forgery data from the input data, said forgery data being for determining whether the input data belongs to a biologic object or a fake;

calculating forgery similarity of biometric data of the genuine person, said forgery similarity indicating similarity to a fake;

calculating forgery difficulty of biometric data of the genuine person, said forgery difficulty indicating difficulty of forgery;

determining a forgery determination threshold on the basis of the forgery similarity and the forgery difficulty;

calculating a forgery score of the input data on the basis of extracted forgery data, said forgery score indicating likelihood of forgery; and

determining whether the input data belongs to a biologic object by comparing calculated forgery score with the forgery determination threshold.