Facial recognition method and related product

• US 10,769,464 B2
• Filed: 08/21/2018
• Issued: 09/08/2020
• Est. Priority Date: 09/12/2017
• Status: Active Grant

• Current Assignee: Guangdong OPPO Mobile Telecommunications Corporation Limited
• Original Assignee: Guangdong OPPO Mobile Telecommunications Corporation Limited
• Inventors: Zhou, Haitao, Wang, Lizhong, Hui, Fangfang
• Primary Examiner(s): Patel, Jayesh A
• Application Number:
  US16/106,954
• Publication Number:
  US 20190080188A1
• Time in Patent Office: 749 Days
• Field of Search: None
• CPC Class Codes: G06V 10/94 Hardware or software architectures specially adapted for image or video understanding
• CPC Class Codes: G06V 40/171 Local features and components; Facial parts eye characteristics G06V40/18; Occluding parts, e.g. glasses; Geometrical relationships
• CPC Class Codes: G06V 40/172 Classification, e.g. identification
• CPC Class Codes: G06V 40/193 Preprocessing; Feature extraction
• CPC Class Codes: G06V 40/45 Detection of the body part being alive

Abstract

Embodiments of the present disclosure relate to mobile terminal technologies, a facial recognition method and related products are provided. The method includes: acquiring a face image by a mobile terminal; extracting face feature information from the face image and matching the face feature information with a predetermined face feature template by a central processing unit (CPU) of the mobile terminal; performing liveness detection according to the face image by a graphics processing unit of the mobile terminal when the CPU extracts the face feature information from the face image and matches the face feature information with the predetermined face feature template.

Claims

1. A facial recognition method, comprising:

acquiring a face image by a mobile terminal;

extracting face feature information from the face image and matching the face feature information with a predetermined face feature template by a central processing unit of the mobile terminal;

performing liveness detection according to the face image by a graphics processing unit of the mobile terminal, when the central processing unit extracts the face feature information from the face image and matches the face feature information with the predetermined face feature template;

wherein a length of time required for extracting and matching the face feature information is less than a length of time required for the liveness detection, and a length of time required for facial recognition is equal to the length of time required for the liveness detection.

2. The method according to claim 1, wherein the liveness detection comprises blink detection and depth of field detection, and performing the liveness detection according to the face image comprises:

performing the blink detection and the depth of field detection in parallel.

3. The method according to claim 2, wherein acquiring the face image comprises:

acquiring more than one face image at different collection angles.

4. The method according to claim 3, wherein acquiring more than one face image at different collection angles comprises:

acquiring more than one face image at different collection angles by enabling the mobile terminal to be oriented to a face of a user at the different collection angles.

5. The method according to claim 3, wherein performing the depth of field detection comprises:

determining a plurality of values of depth of field in different regions of each face image from the more than one face image; and

determining that a detection object is a non-living body when the plurality of values of depth of field are identical.

6. The method according to claim 3, wherein performing the blink detection comprises:

extracting a plurality of eye regions from the more than one face image; and

analyzing change in the plurality of eye regions to determine whether a detection object is a living body.

7. The method according to claim 6, wherein analyzing change in the plurality of eye regions to determine whether a detection object is a living body comprises:

comparing the plurality of eye regions with each other to determine whether change occurs in an area of pupil, an area of white of eye and an area of eyelid; and

when the change occurs in any of the area of pupil the area of white of eye and the area of eyelid, determining a detection object as a living body, when no change occurs in any of the area of pupil the area of white of eye and the area of eyelid, determining the detection object as a non-living body.

8. The method according to claim 1, before acquiring the face image, further comprising:

triggering facial recognition in response to unlocking the mobile terminal, accessing to information stored in the mobile terminal or a payment process.

9. A mobile terminal, comprising a central processing unit, a graphics processing unit, a front camera and a memory, wherein the central processing unit, the graphics processing unit, the front camera and the memory are coupled with each other, wherein

the memory is configured to store a face feature template;

the front camera is configured to acquire a face image;

the central processing unit is configured to extract face feature information from the face image and to match the face feature information with the face feature template; and

the graphics processing unit is configured to perform liveness detection according to the face image when the central processing unit extracts the face feature information from the face image and matches the face feature information with the face feature template;

wherein a length of time required for extracting and matching the face feature information is less than a length of time required for the liveness detection, and a length of time required for facial recognition is equal to the length of time required for the liveness detection.

10. The mobile terminal according to claim 9, wherein the liveness detection comprises blink detection and depth of field detection, when performing the liveness detection according to the face image, the graphics processing unit is configured to:

perform the blink detection and the depth of field detection in parallel.

11. The mobile terminal according to claim 10, wherein when acquiring the face image, the front camera is configured to:

acquire more than one face image at different collection angles.

12. The mobile terminal according to claim 11, wherein when performing the depth of field detection, the GPU is configured to:

determine a plurality of values of depth of field in different regions of each face image from the more than one face image; and

determine that a detection object is a non-living body when the plurality of values of depth of field are identical.

13. The mobile terminal according to claim 11, wherein when performing the blink detection, the graphics processing unit is configured to:

extract a plurality of eye regions from the more than one image; and

analyze change in the plurality of eye regions to determine whether a detection object is a living body.

14. The mobile terminal according to claim 13, wherein when analyzing change in the plurality of eye regions to determine whether a detection object is a living body, the graphics processing unit is configured to:

compare the plurality of eye regions with each other to determine whether change occurs in an area of pupil, an area of white of eye and an area of eyelid; and

when the change occurs in any of the area of pupil the area of white of eye and the area of eyelid, determine a detection object as a living body, when no change occurs in any of the area of pupil the area of white of eye and the area of eyelid, determine the detection object as a non-living body.

15. The mobile terminal according to claim 9, wherein before acquiring the face image, the CPU is configured to:

trigger facial recognition in response to unlocking the mobile terminal, accessing to information stored in the mobile terminal or a payment process.

16. A non-transitory computer-readable storage medium, having a computer program for electronic data exchange stored thereon, wherein the computer program causes a computer to perform the method comprising:

acquiring a face image by a mobile terminal;

extracting face feature information from the face image and matching the face feature information with a predetermined face feature template by a central processing unit of the mobile terminal;

performing liveness detection according to the face image by a graphics processing unit of the mobile terminal when the central processing unit extracts the face feature information from the face image and matches the face feature information with the predetermined face feature template;

wherein a length of time required for extracting and matching the face feature information is less than a length of time required for the liveness detection, and a length of time required for facial recognition is equal to the length of time required for the liveness detection.