Facial recognition apparatus, method and computer-readable medium

• US 8,406,484 B2
• Filed: 12/29/2010
• Issued: 03/26/2013
• Est. Priority Date: 01/05/2010
• Status: Active Grant

• Current Assignee: Samsung Electronics Co. Ltd.
• Original Assignee: Samsung Electronics Co. Ltd.
• Inventors: Yoon, Suk June, Roh, Kyung Shik, Hyung, Seung Yong, Ahn, Sung Hwan
• Primary Examiner(s): CHAWAN, SHEELA C
• Application Number:
  US12/980,913
• Publication Number:
  US 20110164792A1
• Time in Patent Office: 818 Days
• Field of Search: 382/118, 382/154, 382/115, 382/100, 382/117, 382/162, 382/168, 382/172, 382/173, 382/181, 382/190, 382/203, 382/219, 382/224, 382/225, 382/305, 345/592, 345/419, 345/441, 345/589, 345/427, 345/421, 358/497, 358/474
• US Class Current: 382/118
• CPC Class Codes: G06V 20/653 by matching three-dimensional models, e.g. conformal mapping of Riemann surfaces
• CPC Class Codes: G06V 40/172 Classification, e.g. identification
• CPC Class Codes: G06V 40/40 Spoof detection, e.g. liveness detection

Abstract

Two-dimensional image information and three-dimensional image information of a subject are acquired, facial recognition is performed using the two-dimensional image information to determine whether a recognized face is a registered user'"'"'s face, an elliptical model of the user is matched to the three-dimensional image information to calculate an error if it is determined that the recognized face is the user'"'"'s face, and it is determined whether the user'"'"'s face is improperly used based on the error. The subject'"'"'s face is determined using the two-dimensional image information and the three-dimensional image information of the subject and it is determined whether the recognized face is improperly used, thereby improving facial recognition reliability. Thus, information security is improved.

Claims

1. A facial recognition apparatus, comprising:

a two-dimensional information acquisition unit to acquire two-dimensional image information of a subject;

a three-dimensional information acquisition unit to acquire three-dimensional image information of the subject;

a user information database to store an elliptical model corresponding to three-dimensional face information of a user and two-dimensional face information of the user; and

a control unit to perform facial recognition using the two-dimensional image information of the subject, to determine whether a recognized face is the user'"'"'s face, to match the elliptical model of the user to the three-dimensional image information to calculate an error upon determining that the recognized face is the user'"'"'s face, and to determine whether the user'"'"'s face is improperly used based on the error.

2. The facial recognition apparatus according to claim 1, wherein the three-dimensional information acquisition unit acquires distance information corresponding to a Time-Of-Flight (TOF) of an infrared light irradiated to the subject and incident thereto.

3. The facial recognition apparatus according to claim 1, wherein the control unit includes a determination unit to determine that the user'"'"'s face is improperly used if the error exceeds a reference value and to determine that the subject is a same person as the user if the error is equal to or less than the reference value.

4. The facial recognition apparatus according to claim 3, further comprising a display unit to display whether the user'"'"'s face is improperly used according to an instruction of the control unit.

5. The facial recognition apparatus according to claim 1, wherein the control unit includes a face extraction unit to extract a face area based on the two-dimensional image information and the three-dimensional image information.

6. The facial recognition apparatus according to claim 5, wherein the face extraction unit extracts the face area from the three-dimensional image information using edge information.

7. The facial recognition apparatus according to claim 5, wherein the control unit matches the elliptical model to the three-dimensional image information of the face area using Iterative Closest Point (ICP).

8. The facial recognition apparatus according to claim 7, wherein the control unit calculates an error between a plane of the elliptical model and a plane of the three-dimensional image information.

9. The facial recognition apparatus according to claim 5, further comprising:

an input unit to receive user registration information and personal information of the user; and

an ellipsoid generation unit to generate the elliptical model using a size of an ellipsoid corresponding to the face area of the user,

wherein the control unit stores the personal information of the user if the user registration information is received through the input unit and stores the two-dimensional face information and the elliptical model of the user in the user information database.

10. A facial recognition method, comprising:

acquiring, by a processor, two-dimensional image information and three-dimensional image information of a subject;

performing, by the processor, facial recognition using the two-dimensional image information;

determining, by the processor, whether a recognized face is a registered user'"'"'s face;

matching, by the processor, an elliptical model of the user to the three-dimensional image information to calculate an error upon determining that the recognized face is the user'"'"'s face; and

determining, by the processor, whether the user'"'"'s face is improperly used based on the error.

11. The facial recognition method according to claim 10, wherein the acquiring of the three-dimensional image information includes:

irradiating the subject with infrared light;

acquiring distance information corresponding to a Time-Of-Flight (TOF) of the infrared light; and

generating three-dimensional image information corresponding to the distance information.

12. The facial recognition method according to claim 10, wherein the matching of the elliptical model of the user to the three-dimensional image information is performed using Iterative Closest Point (ICP).

13. The facial recognition method according to claim 12, wherein the calculating of the error includes calculating an error between a plane of the elliptical model and a plane of the three-dimensional image information.

14. The facial recognition method according to claim 10, wherein the determining of whether the user'"'"'s face is improperly used based on the error includes:

determining whether the subject is the user'"'"'s face used improperly if the error exceeds a reference value; and

determining whether the subject is a same person as the user if the error is equal to or less than the reference value.

15. The facial recognition method according to claim 10, wherein the matching of the elliptical model of the user to the three-dimensional image information further includes:

extracting a face area based on the two-dimensional image information and the three-dimensional image information; and

matching the elliptical model to the face area.

16. The facial recognition method according to claim 10, further comprising:

receiving personal information of the user if user registration information is input;

receiving two-dimensional image information and three-dimensional image information of the user;

performing facial recognition using the two-dimensional image information to acquire two-dimensional face information;

generating the elliptical model using the two-dimensional image information and the three-dimensional image information; and

storing the personal information of the user, the two-dimensional face information and the elliptical model to perform user registration.

17. At least one non-transitory computer readable medium comprising computer readable instructions that control at least one processor to implement the method of claim 10.

18. A facial recognition method, comprising:

acquiring, by a processor, three-dimensional image information of a subject;

extracting, by the processor, a face area from the three-dimensional image information;

generating, by the processor, an elliptical model using an ellipsoid corresponding to the face area;

performing, by the processor, facial recognition using the three-dimensional image information of the subject;

determining, by the processor, whether a recognized face is a user'"'"'s face;

matching, by the processor, the elliptical model to the three-dimensional image information to calculate an error upon determining that the recognized face is the user'"'"'s face; and

determining, by the processor, whether the user'"'"'s face is improperly used based on the error.

19. The facial recognition method according to claim 18, wherein the matching of the elliptical model to the three-dimensional image information of the face area is performed using Iterative Closest Point (ICP).

20. The facial recognition method according to claim 18, wherein the generating of the ellipsoid includes generating an ellipsoid formed by two diameters of the ellipsoid corresponding to the face area and a normal line to an intersection point between the two diameters, and the normal line has a predetermined diameter.

21. The facial recognition method according to claim 18, wherein the determining of whether the user'"'"'s face is improperly used includes:

determining whether the user'"'"'s face is improperly used if the error exceeds a reference value; and

determining whether the subject is a same person as the user if the error is equal to or less than the reference value.

22. At least one non-transitory computer readable medium comprising computer readable instructions that control at least one processor to implement the method of claim 18.

23. A method, comprising:

acquiring, by a processor, two-dimensional and three-dimensional information of a first face;

performing, by the processor, facial recognition on the two-dimensional information and determining whether the first face matches a second face stored in a user information database;

comparing, by the processor, an elliptical model of the first face based on the three-dimensional information with an elliptical model of the second face stored in the user information database; and

determining, by the processor, if the second face differs from the first face by more than a predetermined amount based on the performing and the comparing.