BIOMETRIC AUTHENTICATION USING BIOLOGIC TEMPLATES

• US 20080212846A1
• Filed: 12/26/2007
• Published: 09/04/2008
• Est. Priority Date: 01/09/2007
• Status: Abandoned Application

• Current Assignee: silex technology, Inc. (Muratec)
• Original Assignee: silex technology, Inc. (Muratec)
• Inventors: Yamamoto, Kazuya, Ichikawa, Shota, Hamaguchi, Koji
• Application Number:
  US11/964,400
• Publication Number:
  US 20080212846A1
• Time in Patent Office: Days
• US Class Current: 382/115
• CPC Class Codes: G06F 21/32 using biometric data, e.g. fingerprints, iris scans or voiceprints
• CPC Class Codes: G06F 21/83 input devices, e.g. keyboards, mice or controllers thereof
• CPC Class Codes: G06F 2221/2101 Auditing as a secondary aspect
• CPC Class Codes: G06V 10/75 Organisation of the matching processes, e.g. simultaneous or sequential comparisons of image or video features; Coarse-fine approaches, e.g. multi-scale approaches; using context analysis; Selection of dictionaries
• CPC Class Codes: G06V 10/993 Evaluation of the quality of the acquired pattern
• CPC Class Codes: G06V 40/1365 Matching; Classification
• CPC Class Codes: G06V 40/50 Maintenance of biometric data or enrolment thereof

Abstract

Tools and techniques for biometric authentication obtain a biologic information input such as a fingerprint image, which is to be accepted or rejected as being input from an authentic user. Calculated matching scores show respective degrees of similarity between the biologic information input and several templates. The templates include a fixed registration biologic information template, as well as non-fixed learning biologic information templates which are subject to replacement.

Claims

1. A biometric apparatus comprising:

a memory unit storing a registration biologic information template which is fixed and at least one learning biologic information template which is subject to replacement; and

a determining unit capable of operating in connection with the memory unit to;

receive an inputted biologic information,

calculate a plurality of matching scores showing a respective degree of similarity between an inputted biologic information and each of a plurality of biologic information templates,

determine whether the inputted biologic information is accepted based on the plurality of matching scores, and

output a signal showing whether the inputted biologic information is accepted.

2. The apparatus of claim 1, wherein the memory unit stores a plurality of learning biologic information templates, and the determining unit calculates matching scores for each of a plurality of learning biologic information templates.

3. The apparatus of claim 1, further comprising a template learning unit capable of operating in connection with the memory unit and the determining unit to replace a learning biologic information template stored in the memory unit with the inputted biologic information in response to a determination of acceptance of the inputted biologic information by the determining unit.

4. The apparatus of claim 3, wherein the template learning unit operates to replace the learning biologic information template which has a matching score indicating a least degree of similarity between the learning biologic information template and the inputted biologic information.

5. The apparatus of claim 1, wherein the determining unit is further capable of operating in connection with the memory unit to calculate an average score as an average value of the matching scores and to determine whether the inputted biologic information is accepted based on whether the average score indicates at least a predetermined average degree of similarity.

6. The apparatus of claim 1, wherein the determining unit is further capable of operating in connection with the memory unit to calculate an average score as an average value of the matching scores and to determine whether the inputted biologic information is accepted based firstly on whether the average score indicates at least a predetermined average degree of similarity, and based secondly on whether each of the matching scores indicates at least a respective predetermined individual degree of similarity.

7. The apparatus of claim 1, further comprising a template registration unit which registers the registration biologic information template and the at least one learning biologic information template in the memory unit, such that the registration biologic information template cannot be replaced during normal operation of the apparatus and the at least one learning biologic information template can be replaced during normal operation of the apparatus.

8. The apparatus of claim 1, wherein the determining unit is further capable of operating in connection with the memory unit to select a learning biologic information template whose matching score is an extreme after the inputted biologic information is accepted by the determining unit, and to accumulate data based on the number of times a given learning biologic information template is selected.

9. A method for biometric authentication, comprising the steps of:

obtaining a biologic information input which is to be accepted or rejected as being input from an authentic user;

calculating a matching score which shows a degree of similarity between the biologic information input and a fixed registration biologic information template;

for each of one or more non-fixed learning biologic information templates, calculating a matching score which shows a degree of similarity between the biologic information input and the learning biologic information template;

determining whether the biologic information input is accepted as input from an authentic user based on the matching scores; and

outputting a signal showing whether the biologic information input is accepted as input from an authentic user.

10. The method of claim 9, further comprising the step of replacing a learning biologic information template with a biologic information input which has been accepted as input from an authentic user.

11. The method of claim 10, wherein the replacing step replaces the learning biologic information template which has a matching score indicating the least degree of similarity with the biologic information input.

12. The method of claim 9, wherein the method further comprises calculating an average matching score as an average value of the matching scores, and the determining step determines whether the biologic information input is accepted as input from an authentic user based on whether the average matching score indicates at least a predetermined average degree of similarity.

13. The method of claim 9, wherein the method further comprises calculating an average matching score as an average value of the matching scores, and the determining step determines whether the biologic information input is accepted as input from an authentic user based on whether two conditions hold, namely, the average matching score indicates at least a predetermined average degree of similarity, and each of the matching scores indicates at least a respective predetermined individual degree of similarity.

14. The method of claim 9, wherein the method further comprises the steps of:

selecting a learning biologic information template based on its matching score;

incrementing a selection-times counter associated with the selected learning biologic information template;

replacing a learning biologic information template whose selection-times counter satisfies a predetermined condition; and

initializing the selection-times counter of the replaced learning biologic information template.

15. The method of claim 9, wherein the method further comprises the steps of:

selecting a learning biologic information template whose matching score, of all the current matching scores, indicates the least degree of similarity with biologic information input;

incrementing a selection-times counter associated with the selected learning biologic information template; and

replacing a learning biologic information template whose selection-times counter satisfies a predetermined condition.

16. The method of claim 15, wherein the selection-times counter satisfies at least one of the following predetermined conditions:

(a) the selection-times counter is larger than a predetermined standard value, (b) the selection-times counter has the largest value of the current selection-times counters, (c) the selection-times counter has the largest value of the current selection-times counters and also the absolute value of the difference between the largest selection-times counter and the next largest selection-times counter is larger than a predetermined value.

17. The method of claim 9, wherein the method further comprises the steps of:

selecting a learning biologic information template whose matching score, of all the current matching scores, indicates the greatest degree of similarity with biologic information input;

incrementing a selection-times counter associated with the selected learning biologic information template; and

replacing a learning biologic information template whose selection-times counter satisfies a predetermined condition.

18. The method of claim 17, wherein the selection-times counter satisfies at least one of the following predetermined conditions:

(a) the selection-times counter is smaller than a predetermined standard value, (b) the selection-times counter has the smallest value of the current selection-times counters, (c) the selection-times counter has the smallest value of the current selection-times counters and also the absolute value of the difference between the smallest selection-times counter and the next smallest selection-times counter is larger than a predetermined value.