DISTANT FACE RECOGNITION SYSTEM

• US 20110285845A1
• Filed: 05/21/2010
• Published: 11/24/2011
• Est. Priority Date: 05/21/2010
• Status: Active Grant

• Current Assignee: Honeywell International Inc.
• Original Assignee: Honeywell International Inc.
• Inventors: Bedros, Saad J., Miller, Ben, Janssen, Michael
• Granted Patent: US 9,262,668 B2
• Time in Patent Office: Days
• US Class Current: 348/143
• CPC Class Codes: G06V 40/16 Human faces, e.g. facial parts, sketches or expressions
• CPC Class Codes: G06V 40/166 using acquisition arrangements
• CPC Class Codes: G06V 40/167 using comparisons between temporally consecutive images
• CPC Class Codes: G06V 40/173 face re-identification, e.g. recognising unknown faces across different face tracks
• CPC Class Codes: G06V 40/174 Facial expression recognition

Abstract

A method and system for automatic face recognition. A primary and a plurality of secondary video cameras can be provided to monitor a detection area. The primary video camera can detect people present in the detection zone. Data can be then transmitted to a prioritizor module that produces a prioritized list of detected people. The plurality of secondary video cameras then captures a high-resolution image of the faces of the people present in the detection area according to the prioritized list provided by the prioritizor module. The high-resolution images can be then provided to a face recognition module, which is used to identify the people present in the detection area.

Claims

1. A method for automatic face recognition at a distance, said method comprising:

monitoring a scene for a presence of at least one individual and tracking said at least one individual moving through said scene using at least one primary camera;

providing tracking and prediction outputs from said primary camera to a prioritizor module;

compiling a ranked list of individuals being tracked through said scene via said prioritizor module;

controlling at least one secondary camera by aiming said cameras at each individual in said scene according to said ranked list provided by said prioritizor;

capturing a plurality of high quality facial images of each face present in said scene using said at least one secondary cameras; and

identifying said at least one individual based on said plurality of high quality facial images using a face recognition module.

2. The method of claim 1 further comprising configuring said at least one primary camera as a wide-angle camera.

3. The method of claim 1 further comprising configuring said at least one secondary camera as a high-resolution camera having digital pan, tilt and zoom features.

4. The method of claim 3 further comprising achieving said plurality of high quality facial images by pan, tilt and zooming said at least one secondary camera onto a targeted face with respect to said at least one individual.

5. The method of claim 1 further comprising:

prioritizing data with respect to said scene utilizing overall scene information regarding said scene obtained from said at least one primary camera in association with target information obtained from said at least one secondary camera;

constructing said ranked list of individuals based on said overall scene information and said target information; and

continuing or interrupting obtaining said target information from said at least one secondary camera.

6. The method of claim 5 further comprising employing a controller with respect to said at least one secondary camera for face detection, face tracking, and facial quality.

7. The method of claim 5 further comprising providing additional target location information from a tracking target with respect to said at least one primary camera.

8. The method of claim 5 further comprising utilizing a minimization of predicted error for a control of a pan, a tilt and a zoom of said at least one secondary camera and said at least one primary camera, wherein said predicted error with respect to said pan and said tilt is measured based on centering of a face of said at least one individual in a high resolution image and wherein said predicted error in said zoom is measured with a face size of said at least one individual in a high resolution image.

9. A system for automatic face recognition at a distance, comprising:

at least one primary camera for monitoring a scene for a presence of at least one individual and tracking said at least one individual moving through said scene;

a prioritizor module configured to provide tracking and prediction outputs from said at least one primary camera;

a ranked list compiled by said prioritizor module of individuals being tracked through said scene;

at least one secondary camera controlled by said prioritizor module wherein said prioritizor module aims each of said at least one secondary cameras at each individual in said scene according to said ranked list;

wherein said secondary camera captures a plurality of high quality facial images of each face present in said scene; and

a face recognition module for identifying said at least one individual based on said plurality of high quality facial images.

10. The system of claim 9 wherein said at least one primary camera is configured as a wide-angle camera.

11. The system of claim 9 wherein said at least one secondary camera is configured as a high-resolution camera having digital pan, tilt and zoom features.

12. The system of claim 11 wherein said plurality of high quality facial images are captured by pan, tilt and zooming said at least one secondary camera onto a targeted face with respect to said at least one individual.

13. The system of claim 9 wherein said prioritizor module is further configured to:

prioritize data with respect to said scene utilizing overall scene information regarding said scene obtained from said at least one primary camera in association with target information obtained from said at least one secondary camera;

construct said ranked list of individuals based on said overall scene information and said target information; and

continue or interrupt obtaining said target information from said at least one secondary camera.

14. The system of claim 13 further comprising a controller adapted to control said at least one secondary camera for face detection, face tracking, and facial quality.

15. The system of claim 13 wherein said primary camera is configured to provide additional target location information to said prioritizor module while tracking a target.

16. The system of claim 13 wherein said prioritizor module is further configured to utilize a minimization of predicted error for a control of a pan, a tilt and a zoom of said at least one secondary camera and said at least one primary camera, wherein said predicted error with respect to said pan and said tilt is measured based on centering of a face of said at least one individual in a high resolution image and wherein said predicted error in said zoom is measured with a face size of said at least one individual in a high resolution image.

17. A system for automatic face recognition at a distance, said system comprising:

at least one primary camera for monitoring a scene for a presence of at least one individual and tracking said at least one individual moving through said scene wherein said primary camera is capable of capturing an image of each face present in said scene;

a prioritizor module configured to provide tracking and prediction outputs from said at least one primary camera;

a ranked list compiled by said prioritizor module of individuals being tracked through said scene;

at least one secondary camera controlled by said prioritizor module wherein said prioritizor module aims each of said at least one secondary cameras at each individual in said scene according to said ranked list wherein said secondary camera captures a plurality of high quality facial images of each face present in said scene; and

a face recognition module for identifying said at least one individual based on said plurality of high quality facial images.

18. The system of claim 17 wherein said primary camera is configured as a wide angle camera with a digital pan, tilt, and zoom capability and said secondary camera is configured as a high resolution camera with a digital pan, tilt, and zoom capability.

19. The system of claim 18 wherein said prioritizor module is further configure to compile said ranked list using overall scene information from said primary camera and target information from said secondary camera, and wherein said prioritizor module is configured to interrupt or continue capture by said secondary cameras according to said overall scene information and said target information.

20. The system of claim 18 wherein said prioritizor module is further configured to utilize a minimization of predicted error for a control of a pan, a tilt and a zoom of said at least one secondary camera and said at least one primary camera, wherein said predicted error with respect to said pan and said tilt is measured based on centering of a face of said at least one individual in a high resolution image and wherein said predicted error in said zoom is measured with a face size of said at least one individual in a high resolution image.