Analysis of reflections of projected light in varying colors, brightness, patterns, and sequences for liveness detection in biometric systems

• US 10,599,912 B2
• Filed: 07/18/2017
• Issued: 03/24/2020
• Est. Priority Date: 12/23/2016
• Status: Active Grant

• Current Assignee: Aware Incorporated (Crowley Holdings Preferred LLC)
• Original Assignee: Aware Incorporated (Crowley Holdings Preferred LLC)
• Inventors: Benini, David, Hills, Louis Scott, Riopka, Taras
• Primary Examiner(s): Perlman, David
• Application Number:
  US15/652,485
• Publication Number:
  US 20180181794A1
• Time in Patent Office: 980 Days
• Field of Search: None
• CPC Class Codes: G06V 40/161 Detection; Localisation; Normalisation
• CPC Class Codes: G06V 40/168 Feature extraction; Face representation
• CPC Class Codes: G06V 40/18 Eye characteristics, e.g. of the iris
• CPC Class Codes: G06V 40/45 Detection of the body part being alive

Abstract

Spoofed faces are unlikely to reflect a likeness detection image in the same manner as live faces. To determine likeness of a subject face, a likeness detection image is displayed on a screen of a face recognition device in the direction of the subject face, and a reflection image of the subject face is captured while the likeness detection image is displayed. The reflection image is analyzed to determine whether the reflection image contains any reflections of the likeness detection image, and a likeness determination is made based on the location and/or character of any such reflections.

Claims

1. A method of determining face liveness, comprising:

during both an enrollment phase and a face recognition phase;

automatically detecting, by a mobile device, a level of light;

displaying a captured image of a subject'"'"'s face on a screen;

conducting a preliminary analysis of an eye of the subject'"'"'s face to determine whether there are any preexisting reflections in the eye;

selecting a liveness detection image based upon whether a reflection of the image will be more or less readily detectable given one or more current environmental condition;

causing, with a processor of the mobile device, the liveness detection image to be displayed on the screen of the mobile device during a first period of time, the displayed liveness detection image brightness adjusted based on the detected level of light;

receiving, at the processor, a reflection image from a camera utilized by the mobile device, the reflection image captured during the first period of time and showing the subject'"'"'s face;

displaying the captured image of a subject'"'"'s face or another captured image of the subject'"'"'s face on the screen;

analyzing, with the processor, the reflection image to identify any reflections of the liveness detection image in the reflection image;

making, with the processor, a liveness determination based on the analysis;

selecting one captured image determined to be live; and

storing the one captured image determined to be live; and

normalizing the one captured image.

2. The method of determining face liveness of claim 1, further comprising receiving, at the processor, an initial image of the subject face, the initial image captured during a period of time preceding the first period of time.

3. The method of determining face liveness of claim 2, wherein the analyzing comprises:

comparing the reflection image to the initial image; and

identifying portions of the reflection image that are different than the initial image.

4. The method of determining face liveness of claim 1, wherein the analyzing comprises:

dividing the reflection image into a plurality of zones; and

determining whether a reflection of the liveness detection image appears in each of the plurality of zones.

5. The method of claim 1, wherein the liveness detection image is a solid white image.

6. The method of claim 1, wherein the liveness detection image is a patterned image.

7. The method of claim 1, wherein the liveness detection image is a colored image.

8. The method of claim 1, further comprising:

receiving information about a light level from a light meter utilized by the mobile device; and

causing, with the processor, a brightness of the screen to be adjusted based on the received information.

9. The method of claim 1, further comprising:

causing, with the processor of the mobile device, a plurality of liveness detection images to be displayed on the screen of the mobile device during the first period of time;

receiving, at the processor, a plurality of reflection images from the camera utilized by the mobile device;

the plurality of reflection images captured during the first period of time, each of the plurality of reflection images captured while a different one of the plurality of liveness detection images was caused to be displayed on the screen of the mobile device;

analyzing, with the processor, each of the plurality of reflection images to identify any reflections of the liveness detection image in each of the plurality of reflection images; and

making, with the processor, a liveness determination based on the analysis.

10. The method of claim 9, wherein at least one of the plurality of liveness detection images is different than at least another one of the plurality of liveness detection images.

21. The method of determining face liveness of claim 1, further comprising excluding the preexisting reflections from consideration when analyzing the eye to identify the reflection image.

11. A face enrollment or face recognition device, comprising:

a camera;

a display screen;

a processor;

and a memory storing instructions for execution by the processor that, when executed by the processor, cause the processor, for both an enrollment phase and a face recognition phase, to;

display a captured image of a subject'"'"'s face on the display screen;

automatically detect a level of light;

conduct a preliminary analysis of an eye of the subject'"'"'s face to determine whether there are any preexisting reflections in the eye;

selecting a liveness detection image based upon whether a reflection of the image will be more or less readily detectable given one or more current environmental condition;

cause the liveness detection image to be displayed on the display screen the displayed liveness detection image brightness adjusted based on the detected level of light;

cause the camera to capture a reflection image of a subject face while the liveness detection image is displayed on the display screen;

display the captured image of a subject'"'"'s face or another captured image of the subject'"'"'s face on the display screen;

analyze the reflection image to identify one or more reflections of the liveness detection image in the reflection image;

make a liveness determination based on the analysis;

select one captured image determined to be live; and

normalize the one captured image.

12. The face enrollment or face recognition device of claim 11, further comprising:

causing a plurality of liveness detection images to be displayed on the display screen in sequence;

causing the camera to capture at least one reflection image of the subject face while each of the plurality of liveness detection images is displayed on the display screen; and

analyzing each reflection image to identify one or more reflections of one of the plurality of liveness detection images therein.

13. The face enrollment or face recognition device of claim 11, wherein the memory stores additional instructions that, when executed, further cause the processor to:

determine that the subject face is live if the only reflection of the liveness detection image is in the iris of each eye of the subject face; and

determine that the subject face is spoofed if a background of the subject face in the reflection image contains a reflection of the liveness detection image.

14. The face enrollment or face recognition device of claim 11, wherein the face recognition device is a mobile device.

15. The face enrollment or face recognition device of claim 11, wherein the liveness detection image is one of a solid white image, a patterned image, and a colored image.

16. The face enrollment or face recognition device of claim 11, wherein the memory stores additional instructions for execution by the processor that, when executed by the processor, cause a brightness of the display screen to be adjusted based on a detected light level.

17. The face enrollment or face recognition device of claim 11, wherein analyzing the reflection image to identify one or more reflections of the liveness detection image in the reflection image comprises:

determining whether the subject face in the reflection image is wearing eyeglasses, and, if so, determining whether there is a reflection of the liveness detection image on the eyeglasses in the reflection image.

22. The face enrollment or face recognition device of claim 11, further comprising excluding the preexisting reflections from consideration when analyzing the eye to identify the reflection image.

18. A method for determining whether a subject face is live or spoofed, for both an enrollment process and face recognition process, comprising:

conducting, for both the enrollment process and the face recognition process, a preliminary analysis of an eye of the subject face to determine whether there are any preexisting reflections in an eye;

selecting, for both the enrollment process and the face recognition process, a liveness detection image based upon whether a reflection of the image will be more or less readily detectable given one or more current environmental condition;

projecting, for both the enrollment process and the face recognition process, the liveness detection image onto a subject face during a first time period;

capturing, for both the enrollment process and the face recognition process, a reflection image of the subject face during the first time period;

identifying, for both the enrollment process and the face recognition process, at least one reflection of the liveness detection image in the reflection image;

determining, for both the enrollment process and the face recognition process, a location in the reflection image, relative to the subject face, of the at least one reflection;

based on the determined location, making a liveness determination, for both the enrollment process and the face recognition process;

selecting, for both the enrollment process and the face recognition process, one captured image determined to be live;

normalizing, for both the enrollment process and the face recognition process, the one captured image; and

storing, for both the enrollment process and the face recognition process, the one captured image.

19. The method of claim 18, wherein the making a liveness determination further comprises:

determining that the subject face is spoofed if the determined location comprises a background of the subject face in the reflection image.

20. The method of claim 18, wherein the liveness detection image is one of a solid white image, a black and white patterned image, a solid colored image, and a colored patterned image.

23. The method of claim 18, further comprising excluding the preexisting reflections from consideration when analyzing the eye to identify the reflection image.