System and method for face capture and matching

• US 9,773,157 B2
• Filed: 12/07/2015
• Issued: 09/26/2017
• Est. Priority Date: 01/30/2012
• Status: Active Grant

• Current Assignee: Accenture Global Services Limited (Accenture PLC)
• Original Assignee: Accenture Global Services Limited (Accenture PLC)
• Inventors: Bataller, Cyrille, Astrom, Anders
• Primary Examiner(s): Gilliard, Delomia L
• Application Number:
  US14/961,294
• Publication Number:
  US 20160104033A1
• Time in Patent Office: 659 Days
• Field of Search: None
• CPC Class Codes: G06F 18/22 Matching criteria, e.g. proximity measures
• CPC Class Codes: G06F 18/28 Determining representative reference patterns, e.g. by averaging or distorting; Generating dictionaries
• CPC Class Codes: G06T 11/80 Creating or modifying a manually drawn or painted image using a manual input device, e.g. mouse, light pen, direction keys on keyboard
• CPC Class Codes: G06T 17/00 Three dimensional [3D] modelling, e.g. data description of 3D objects
• CPC Class Codes: G06T 7/246 using feature-based methods, e.g. the tracking of corners or segments
• CPC Class Codes: G06V 10/772 Determining representative reference patterns, e.g. averaging or distorting patterns; Generating dictionaries
• CPC Class Codes: G06V 20/52 Surveillance or monitoring of activities, e.g. for recognising suspicious objects recognising microscopic objects G06V20/69
• CPC Class Codes: G06V 40/166 using acquisition arrangements
• CPC Class Codes: G06V 40/172 Classification, e.g. identification
• CPC Class Codes: G06V 40/173 face re-identification, e.g. recognising unknown faces across different face tracks

Abstract

According to an example, a face capture and matching system may include a memory storing machine readable instructions to receive captured images of an area monitored by an image capture device, and detect one or more faces in the captured images. The memory may further store machine readable instructions to track movement of the one or more detected faces in the area monitored by the image capture device, and based on the one or more tracked detected faces, select one or more images from the captured images to be used for identifying the one or more tracked detected faces. The memory may further store machine readable instructions to select one or more fusion techniques to identify the one or more tracked detected faces using the one or more selected images. The face capture and matching system may further include a processor to implement the machine readable instructions.

Claims

1. A face capture and matching system comprising:

a non-transitory machine readable storage medium storing machine readable instructions to;

receive captured images of an area monitored by an image capture device;

detect at least one face in the captured images;

track movement of the at least one detected face in the area monitored by the image capture device;

based on the at least one tracked detected face, select at least one image from the captured images to be used for identifying the at least one tracked detected face; and

select at least one fusion technique to identify the at least one tracked detected face using the at least one selected image;

selectively queue selected ones of the captured images to process the selected ones of the captured images based on a determination of whether a predetermined number of the at least one detected face for movement tracking exceeds a specified threshold; and

a processor to implement the machine readable instructions.

2. The face capture and matching system of claim 1, further comprising machine readable instructions to:

provide feedback to the image capture device to enhance image characteristics.

3. The face capture and matching system of claim 1, wherein the machine readable instructions to track the movement of the at least one detected face in the area monitored by the image capture device further comprise machine readable instructions to:

assign a tracking identification (ID) to the at least one tracked detected face.

4. The face capture and matching system of claim 1, wherein the machine readable instructions to track the movement of the at least one detected face in the area monitored by the image capture device further comprise machine readable instructions to:

determine that the at least one tracked detected face is of a same person if the at least one tracked detected face is located within a predetermined distance in consecutive captured images.

5. The face capture and matching system of claim 4, further comprising machine readable instructions to:

increase or decrease the predetermined distance based on an image capture rate of the image capture device.

6. The face capture and matching system of claim 1, wherein the machine readable instructions to track the movement of the at least one detected face in the area monitored by the image capture device further comprise machine readable instructions to:

determine that the at least one tracked detected face is of a same person if the at least one tracked detected face is located along a calculated direction of movement in consecutive captured images.

7. The face capture and matching system of claim 1, wherein the machine readable instructions to track the movement of the at least one detected face in the area monitored by the image capture device further comprise machine readable instructions to:

determine that the at least one tracked detected face is of a same person based on a size of the at least one tracked detected face in consecutive captured images.

8. The face capture and matching system of claim 1, wherein the machine readable instructions to select the at least one fusion technique to identify the at least one tracked detected face using the at least one selected image further comprise machine readable instructions to:

use a single best face image fusion technique based on detection of a frontal face of a predetermined quality related to a quality of the at least one detected face in the captured images to identify the at least one tracked detected face.

9. The face capture and matching system of claim 1, wherein the machine readable instructions to select the at least one fusion technique to identify the at least one tracked detected face using the at least one selected image further comprise machine readable instructions to:

use a matching template fusion technique based on a combination of a plurality of frontal faces to generate a fused matching template to identify the at least one tracked detected face.

10. The face capture and matching system of claim 1, wherein the machine readable instructions to select the at least one fusion technique to identify the at least one tracked detected face using the at least one selected image further comprise machine readable instructions to:

use a three-dimensional (3D) model fusion technique based on generation of a 3D model of a face from a plurality of detected faces to identify the at least one tracked detected face.

11. The face capture and matching system of claim 1, wherein the machine readable instructions to select the at least one fusion technique to identify the at least one tracked detected face using the at least one selected image further comprise machine readable instructions to:

select the at least one fusion technique from a plurality of fusion techniques including;

a single best face image fusion technique based on detection of a frontal face of a predetermined quality to identify the at least one tracked detected face,

a matching template fusion technique based on a combination of a plurality of frontal faces to generate a fused matching template to identify the at least one tracked detected face, and

a three-dimensional (3D) model fusion technique based on generation of a 3D model of a face from a plurality of detected faces to identify the at least one tracked detected face.

12. The face capture and matching system of claim 1, further comprising machine readable instructions to:

identify the at least one tracked detected face using the at least one selected image;

match the identified face to a predetermined list of captured faces; and

generate an alert based on the matched face.

13. The face capture and matching system of claim 12, wherein the machine readable instructions to generate the alert based on the matched face further comprise machine readable instructions to:

generate an e-mail to alert a user of the face capture and matching system of the matched face.

14. The face capture and matching system of claim 12, wherein the machine readable instructions to generate the alert based on the matched face further comprise machine readable instructions to:

generate at least one of a color coded signal and an audio signal to alert a user of the face capture and matching system of the matched face.

15. A method for face capture and matching, the method comprising:

receiving captured images of an area monitored by an image capture device;

detecting at least one face in the captured images;

tracking movement of the at least one detected face in the area monitored by the image capture device by

determining that the at least one tracked detected face is of a same person if the at least one tracked detected face is located within a predetermined distance in consecutive captured images, and increase or decrease the predetermined distance based on an image capture rate of the image capture device,

determining that the at least one tracked detected face is of the same person if the at least one tracked detected face is located along a calculated direction of movement in the consecutive captured images, or

determining that the at least one tracked detected face is of the same person based on a size of the at least one tracked detected face in the consecutive captured images;

based on the at least one tracked detected face, selecting at least one image from the captured images to be used for identifying the at least one tracked detected face; and

selecting, by a processor, at least one fusion technique to identify the at least one tracked detected face using the at least one selected image.

16. The method according to claim 15, further comprising:

selectively queuing selected ones of the captured images to process the selected ones of the captured images based on a determination of a predetermined number of the at least one detected face for movement tracking.

17. The method according to claim 15, further comprising:

selectively queuing the captured images to process the captured images that include a predetermined number of the at least one detected face that meet a predetermined quality metric related to a quality of the at least one detected face in the captured images for movement tracking.

18. The method according to claim 15, further comprising:

selectively queuing the captured images to process the captured images for movement tracking based on at least one of a predetermined time lag and a maximum number of images queued.

19. A non-transitory computer readable medium having stored thereon a computer executable program to provide face capture and matching, the computer executable program when executed causes a computer system to:

receive captured images of an area monitored by an image capture device;

detect at least one face in the captured images;

track movement of the at least one detected face in the area monitored by the image capture device;

based on the at least one tracked detected face, select at least one image from the captured images to be used for identifying the at least one tracked detected face; and

select, by a processor, at least one fusion technique to identify the at least one tracked detected face using the at least one selected image by using a matching template fusion technique based on a combination of a plurality of frontal faces to generate a fused matching template to identify the at least one tracked detected face.

20. The non-transitory computer readable medium according to claim 19, the computer executable program when executed further causes the computer system to:

selectively queue selected ones of the captured images to process the selected ones of the captured images based on a determination of whether a predetermined number of the at least one detected face for movement tracking exceeds a specified threshold,

selectively queue the captured images to process the captured images that include the predetermined number of the at least one detected face that meet a predetermined quality metric related to the quality of the at least one detected face in the captured images for movement tracking, or

selectively queue the captured images to process the captured images for movement tracking based on at least one of a predetermined time lag and a maximum number of images queued.