Fingerprint authentication method and system for rejecting spoof attempts

Patent metadata

Field	Value
Publication number	US-11144625-B2
Application number	US-201816616856-A
Country	US
Kind code	B2
Filing date	May 31, 2018
Priority date	Jun 7, 2017
Publication date	Oct 12, 2021
Grant date	Oct 12, 2021

Abstract

A method of authenticating a user by means of a fingerprint authentication system comprising a finger sensing arrangement, comprising the following steps for each authentication attempt in a sequence of authentication attempts: receiving a touch by a candidate finger probe on the finger sensing arrangement; acquiring a candidate fingerprint image of the candidate finger probe; determining an authentication representation based on the candidate fingerprint image; retrieving a stored enrollment representation of an enrolled fingerprint of the user; determining a match score based on a comparison between the authentication representation and the enrolment representation; determining a liveness score for the authentication attempt; determining a qualification metric for the authentication attempt based on a relation between the liveness score for the authentication attempt and a liveness score for at least one previous authentication attempt; and determining an authentication result for the authentication attempt based on the match score, and the qualification metric.

First claim

What is claimed is:

1. A method of authenticating a user by means of a fingerprint authentication system comprising a finger sensing arrangement, said method comprising the steps of, for each authentication attempt in a sequence of authentication attempts:
receiving a touch by a candidate finger probe on said finger sensing arrangement;
acquiring a candidate fingerprint image of said candidate finger probe;
determining an authentication representation based on said candidate fingerprint image;
retrieving a stored enrollment representation of an enrolled fingerprint of said user;
determining a match score based on a comparison between said authentication representation and said enrolment representation;
determining a liveness score for said authentication attempt based on said candidate fingerprint image;
determining a qualification metric for said authentication attempt based on a relation between the liveness score for said authentication attempt and a liveness score for at least one previous authentication attempt; and
determining an authentication result for said authentication attempt based on said match score, and said qualification metric.

2. The method according to claim 1 , wherein said qualification metric is further based on a relation between the match score for said authentication attempt and a match score for at least one previous authentication attempt.

3. The method according to claim 1 , wherein said qualification metric is further based on an estimated movement of said candidate finger probe since a previous authentication attempt.

4. The method according to claim 1 , wherein said qualification metric is further based on a time period since said previous authentication attempt.

5. The method according to claim 1 , wherein said qualification metric is determined based on the liveness score for said authentication attempt and liveness scores for a plurality of previous authentication attempts.

6. The method according to claim 5 , wherein said qualification metric is determined further based on the match score for said authentication attempt and match scores for said plurality of previous authentication attempts.

7. The method according to claim 1 , wherein said qualification metric is determined using an empirical model.

8. The method according to claim 7 , wherein said empirical model is determined using machine learning.

9. The method according to claim 1 , wherein the step of determining said authentication result comprises the step of:
providing, when said match score indicates that said authentication representation matches with said enrollment representation and said qualification metric indicates that a likelihood of an ongoing spoofing attempt is greater than a predefined threshold likelihood, a signal indicating a failed authentication.

10. The method according to claim 1 , further comprising the steps of, for each authentication attempt in said sequence of authentication attempts:
when the liveness score for said authentication attempt indicates a likely spoof:
providing a signal indicating a failed authentication;
determining an anti-spoofing representation based on said candidate fingerprint image; and
storing said anti-spoofing representation.

11. The method according to claim 10 , further comprising the steps of, for each authentication attempt in said sequence of authentication attempts:
retrieving a stored anti-spoofing representation associated with a previous authentication attempt;
comparing said authentication representation with said anti-spoofing representation associated with the previous authentication attempt; and
when said authentication representation matches with said anti-spoofing representation, providing a signal indicating a failed authentication.

12. A fingerprint authentication system for authenticating a user, comprising:
a fingerprint sensing arrangement;
a memory; and
processing circuitry coupled to said fingerprint sensing arrangement, said processing circuitry being configured to, for each authentication attempt in a sequence of authentication attempts:
determining a qualification metric for the authentication attempt based on a relation between the liveness score for the authentication attempt and a liveness score for at least one previous authentication attempt; and
determining an authentication result for the authentication attempt based on the match score, and the qualification metric.

13. An electronic device comprising;
a fingerprint authentication system according to claim 12 ; and
a processing unit configured to control said fingerprint authentication system to carry out a fingerprint authentication of a user, and to perform at least one action only upon successful authentication of said user.

Assignees

• Fingerprint Cards Ab

Inventors

• Setterberg Eric

Classifications

• G06V40/1306

  non-optical, e.g. ultrasonic or capacitive sensing · CPC title

• G06V40/1365

  Matching; Classification · CPC title

• G06V40/13 Primary

  Sensors therefor · CPC title

• G06F21/32 Primary

  using biometric data, e.g. fingerprints, iris scans or voiceprints · CPC title

• G06V40/1382

  Detecting the live character of the finger, i.e. distinguishing from a fake or cadaver finger · CPC title