FACE RECOGNITION METHOD AND DEVICE AND APPARATUS

• US 20180089499A1
• Filed: 04/18/2017
• Published: 03/29/2018
• Est. Priority Date: 09/24/2016
• Status: Active Grant

• Current Assignee: Nanchang O-film Bio-Identification Technology Co., Ltd. (OFILM Group Co., Ltd.)
• Original Assignee: Nanchang O-film Bio-Identification Technology Co., Ltd. (OFILM Group Co., Ltd.)
• Inventors: SUN, Wensi, BAI, Anpeng
• Granted Patent: US 10,311,289 B2
• Time in Patent Office: Days
• 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 40/193 Preprocessing; Feature extraction
• CPC Class Codes: G06V 40/45 Detection of the body part being alive

Abstract

The present disclosure relates to the field of biological recognition and provides a face recognition method and device and apparatus. The face recognition method includes: recording a face video image and dynamically tracking an eye in the face video image; detecting a blink frequency of the eye during dynamic tracking of the eye; judging whether the blinking frequency is within a preset frequency range; if the blink frequency is within the preset frequency range, determining that the face is a living face; if the blink frequency is not within the preset frequency range, determining that the face is a fake face. In the embodiment of the present disclosure, the real face and the fake face can be accurately distinguished by the blink frequency of the eye. Therefore, the safety of face recognition is improved.

Claims

1. A face recognition method, comprising:

recording a face video image and dynamically tracking an eye in the face video image;

detecting a blink frequency of the eye during dynamic tracking of the eye;

judging whether the blinking frequency is within a preset frequency range;

if the blink frequency is within the preset frequency range, determining that the face is a living face;

if the blink frequency is not within the preset frequency range, determining that the face is a fake face.

2. The method of claim 1, wherein, before detecting the blink frequency of the eye, the method further comprises:

capturing a preset number of face images from the face video image at a preset time interval;

detecting the blinking frequency of the eye, comprising;

detecting the blink frequency of the eye according to the preset number of face images.

3. The method of claim 1, wherein, detecting the blinking frequency of the eye comprises:

detecting the eyeball shape and eyeball area of the eye;

determining a state of the eye according to the eyeball shape and eyeball area of the eye, the state comprising an open-eye state and a closed-eye state;

recording the number of times the state of the eye changes from the open-eye state to the closed-eye state during a preset time period;

determining the number of times as the blink frequency of the eye.

4. The method of claim 1, wherein, the method further comprises:

extracting facial feature information of the face from the face video image;

judging whether there is a face image matching the facial feature information in a database;

if there is a face image matching the facial feature information in the database, obtaining a pre-stored blink frequency range corresponding to the face image in the database;

determining the blink frequency range as the preset frequency range.

5. The method of claim 4, wherein, extracting the facial feature information of the face from the face video image comprises:

determining a face region of the face video image,

covering a pre-stored face-region image onto the face video image and searching the position of each feature in the face region of the face video image corresponding to vicinity of the position of each feature in the covered pre-stored face-region image, so as to determine the facial feature information of the face in the face video image.

6. The method of claim 4, wherein, judging whether there is a face image matching the facial feature information in a database comprises:

when the similarity between the facial feature information and the facial feature information of the face image stored in the database exceeds a preset ratio, determining that there is a face image matching the facial feature information in the database.

7. The method of claim 1, wherein, the method further comprises:

if the blink frequency is not within the preset frequency range, outputting a warning indicative of the face being a fake face.

8. A face recognition device, comprising:

a recording unit, configured to record a face video image;

a processor, the processor comprising;

a tracking unit, configured to dynamically track an eye in the face video image;

a detecting unit, configured to detect a blink frequency of the eye during the tracking unit dynamically tracking of the eye;

a judging unit, configured to judge whether the blinking frequency detected by the detecting unit is within a preset frequency range;

a determining unit, configured to, when the judging unit judges that the blink frequency is within the preset frequency range, determine that the face is a living face;

and when the judging unit judges that the blink frequency is not within the preset frequency range, determine that the face is a fake face.

9. The device of claim 8, wherein, the face recognition device further comprises a capturing unit configured to capture a preset number of face images from the face video image recorded by the recording unit at a preset time interval;

the detecting unit detecting the blinking frequency of the eye, comprises;

detecting the blink frequency of the eye according to the preset number of face images.

10. The device of claim 8, wherein, the detecting unit comprises:

a detecting sub unit, configured to detect the eyeball shape and eyeball area of the eye;

a determining sub unit, configured to determine a state of the eye according to the eyeball shape and eyeball area of the eye, the state comprising an open-eye state and a closed-eye state;

a recording sub unit, configured to record the number of times the state of the eye changes from the open-eye state to the closed-eye state during a preset time period;

the determining sub unit, further configured to determine the number of times as the blink frequency of the eye.

11. The device of claim 8, wherein, the processor further comprises:

an extracting unit, configured to extract facial feature information of the face from the face video image recorded by the recording unit;

the judging unit, configured to judge whether there is a face image matching the facial feature information extracted by the extracting unit in a database;

the processor further comprises a obtaining unit, configured to, when the judging unit judges that there is a face image matching the facial feature information in the database, obtain a pre-stored blink frequency range corresponding to the face image in the database;

the determining unit, configured to determine the blink frequency range as the preset frequency range.

12. The device of claim 11, wherein, the extracting unit is further configured to determine a face region of the face video image, cover a pre-stored face-region image onto the face video image and search the position of each feature in the face region of the face video image corresponding to vicinity of the position of each feature in the covered pre-stored face-region image, so as to determine the facial feature information of the face in the face video image.

13. The device of claim 11, wherein, the judging unit is further configured to, when the similarity between the facial feature information and the facial feature information of the face image stored in the database exceeds a preset ratio, determine that there is a face image matching the facial feature information in the database.

14. The device of claim 8, wherein, the face recognition device further comprises:

an outputting unit, configured to, if the blink frequency is not within the preset frequency range, output a warning indicative of the face being a fake face.

15. A face recognition apparatus, comprising:

one or more processors, one or more inputting apparatus, and a memory, the memory being configured to store instructions, the processor being configured to execute the instructions stored in the memory,

the inputting apparatus being configured to record a face video image;

when executing the instructions stored in the memory, the processor being configured to dynamically track an eye in the face video image, detect a blink frequency of the eye during dynamic tracking of the eye;

judge whether the blinking frequency is within a preset frequency range;

if the blink frequency is within the preset frequency range, determine that the face is a living face;

if the blink frequency is not within the preset frequency range, determine that the face is a fake face.

16. The apparatus of claim 15, wherein, the inputting apparatus is further configured to capture a preset number of face images from the face video image at a preset time interval;

the processor detecting the blink frequency of the eye during dynamic tracking of the eye, comprising;

detecting the blink frequency of the eye according to the preset number of face images.

17. The apparatus of claim 15, wherein the processor detecting the blink frequency of the eye comprises:

detecting the eyeball shape and eyeball area of the eye;

determining a state of the eye according to the eyeball shape and eyeball area of the eye, the state comprising an open-eye state and a closed-eye state;

recording the number of times the state of the eye changes from the open-eye state to the closed-eye state during a preset time period;

determining the number of times as the blink frequency of the eye.

18. The apparatus of claim 15, wherein the processor is further configured to:

extract facial feature information of the face from the face video image recorded by the recording unit;

judge whether there is a face image matching the facial feature information in a database;

if there is a face image matching the facial feature information in the database, obtain a pre-stored blink frequency range corresponding to the face image in the database, and determine the blink frequency range as the preset frequency range.

19. The apparatus of claim 15, wherein the face recognition apparatus further comprises an outputting apparatus configured to, when the processor judges that the blink frequency is not within the preset frequency range, output a warning indicative of the face being a fake face.

20. The apparatus of claim 19, wherein the inputting apparatus comprises a camera, and the outputting apparatus comprises a display and/or a speaker.