Method for selecting frames used in face processing

1. A method used in a personal equipment for selecting frames used in face processing, comprising:

a) Capturing with a camera video data featuring a face of an individual;

b) Determining with a processing unit at least one image quality indicator (q ij ) for at least some frames in the video data;

c) Using the quality indicator (q ij ) for selecting a subset of the frames (Sel 1 , Sel 2 );

d) Detecting a sequence of frames (fperiod) corresponding to a movement of a body portion in the video data and/or corresponding to a response window (rw) during which a response to a challenge should be given;

e) Adding to the subset at least one second frame (Sel 3 , Sel 4 ) within a predefined interval (Int 1 , Int 2 ) before or after said the sequence; and

f) Storing the subset of frames (Sel 1 , Sel 2 , SeI 3 , Sel 4 ) in a memory.

2. The method of claim 1 , comprising:

prompting a user for a response to a challenge;

detecting the movement of a body portion as the response;

and wherein second frame (Sel 3 ) is selected within the predefined interval (Int 1 ) before or after the sequence (fperiod) during which the movement of a body portion is detected.

3. The method of claim 2 , wherein the response corresponds to one of the following voluntary movements:

a blink;

a head movement;

a movement of the eyebrow, cheek, and/or mouth; and/or

a displacement of the camera relative to the face.

4. The method of claim 1 , comprising:

prompting a user for a response to a challenge;

waiting during the response window (rw) for the movement of a body portion as the response to the challenge;

and wherein the second frame (Sel 4 ) is selected within the predefined interval (Int 2 ) before said prompting the user or after the end of the response window (rw).

5. The method of claim 1 the movement being an involuntary movement, the second frame (Sel 3 ) being selected within the predefined interval (Int 1 ) before or after the sequence (fperiod) during which the movement of a body portion is detected.

6. The method of claim 1 wherein a frame (Sel 3 ) immediately adjacent to the sequence (fperiod) is selected as the second frame.

7. The method of claim 1 , wherein consecutive frames are excluded in said subset.

8. The method of claim 1 , wherein the quality indicator (q ij ) depends on at least one of the following parameters for the whole frame:

quality of illumination;

sharpness;

blur;

noise level.

9. The method of claim 1 , wherein the quality indicator (q ij ) depends on at least one of the following parameters separately determined for the portion of the image corresponding to the face:

quality of illumination;

quality of illumination;

sharpness;

blur;

noise level;

face orientation;

size of the face;

occlusion or reflects;

user expression;

likelihood that the image corresponds to a face; and

liveness.

10. The method of claim 1 , performed during an enrolment phase of a biometric face recognition process.

11. The method of claim 1 , performed during a test phase of a biometric face recognition process.

12. The method of claim 11 , the selection depending on a similarity of at least one frame quality indicator of at least one test frame with at least one reference frame.

13. The method of claim 1 , further comprising:

pre-processing the frames (Sel 1 , Sel 2 , . . . ) in the subset;

sending the pre-processed frames remotely.

14. The method of claim 13 , the pre-processing comprising rectifying the pose of the individual.

15. The method of claim 13 , further comprising:

receiving the remotely sent frames in a remote equipment;

extracting biometric features from the frames.

16. User equipment for selecting frames used in face processing, comprising:

a) a video camera for capturing video data with the face of an individual;

b) a processing unit arranged for:

determining at least one image quality indicator (q ij ) for at least some frames in the video data;

using the quality indicator for selecting a subset of the frames (Sel 1 , Sel 2 . . . );

determining a sequence of frames (fperiod) corresponding to a movement of a body portion in the video data and/or corresponding to a response window (rw) during which a response to a challenge should be given;

adding to the subset at least one second frame (Sel 3 , Sel 4 ) within a predefined interval (Int 1 , Int 1 ) before or after the sequence;

c) a memory for storing the frames in the subset.

17. A non-transitory computer-program product tangibly embodied in a machine-readable non-transitory storage medium that includes instructions configured to cause one or more processors to:

a) capture with a camera video data featuring a face of an individual;

b) determine with a processing unit at least one image quality indicator (q ij ) for at least some frames in the video data;

c) using the quality indicator (q ij ) for selecting a subset of the frames (Sel 1 , Sel 2 );

d) detect a sequence of frames (fperiod) corresponding to a movement of a body portion in the video data and/or corresponding to a response window (rw) during which a response to a challenge should be given;

e) add to the subset at least one second frame (Sel 3 , Sel 4 ) within a predefined interval (Int 1 , Int 1 ) before or after the sequence; and

f) store the subset of frames (Sel 1 , Sel 2 , SeI 3 , Sel 4 ) in a memory.