Low threshold face recognition

• US 9,076,029 B2
• Filed: 12/03/2012
• Issued: 07/07/2015
• Est. Priority Date: 06/29/2010
• Status: Active Grant

• Current Assignee: Apple Inc.
• Original Assignee: Apple Inc.
• Inventors: Free, Robert Mikio
• Primary Examiner(s): Koziol, Stephen R
• Assistant Examiner(s): LE, TOTAM HA
• Application Number:
  US13/692,500
• Publication Number:
  US 20140112552A1
• Time in Patent Office: 946 Days
• Field of Search: 382/118
• US Class Current: 1/1
• CPC Class Codes: G06V 40/161 Detection; Localisation; Normalisation
• CPC Class Codes: G06V 40/165 using facial parts and geometric relationships
• CPC Class Codes: G06V 40/172 Classification, e.g. identification

Abstract

Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, are disclosed for reducing the impact of lighting conditions and biometric distortions, while providing a low-computation solution for reasonably effective (low threshold) face recognition. In one aspect, the methods include processing a captured image of a face of a user seeking to access a resource by conforming a subset of the captured face image to a reference model. The reference model corresponds to a high information portion of human faces. The methods further include comparing the processed captured image to at least one target profile corresponding to a user associated with the resource, and selectively recognizing the user seeking access to the resource based on a result of said comparing.

Claims

1. An image processing method, the method comprising:

determining a skin tone hue value;

determining if each pixel'"'"'s hue value in an image is within a range defined by a radial distance from the skin tone hue value;

for pixels that are not within the range, excluding the pixels'"'"' values from the image processing;

for pixels that are within the range, increasing a score of each such pixel;

performing facial recognition based on the remaining pixel scores, detecting parts of a face using shape detection, and

recognizing specific ranges of distance from the skin tone hue value correlated with specific parts of the face.

2. The method of claim 1, further comprising:

determining the radial distance for each pixel'"'"'s hue value from the skin tone hue value; and

increasing the scores proportionately to the determined distance.

3. The method of claim 1, wherein the facial recognition further comprises:

modifying pixels'"'"' scores based on a search characteristic related to the shape detection.

4. The method of claim 3, wherein the facial recognition further comprises:

normalizing the detected face to a specified scale and orientation.

5. The method of claim 4, wherein the facial recognition further comprises:

comparing the normalized face to a reference image.

6. The method of claim 5, wherein the comparing further comprises:

creating a matching map by subtracting the normalized face and the reference image from each other using the pixels'"'"' scores based on a set of rules;

deciding if the reference image and normalized face match;

if the reference image and normalized face match, performing a device operation;

if the reference image and normalized face are not a match, comparing the normalized face to a different reference image.

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

determining a center of mass of a face; and

tracking facial motion based on the changing center of mass.

8. The method of claim 1, wherein the skin tone hue has a value of 25 and the range is within 30 degrees of the skin tone hue.

9. A non-transitory machine readable storage device, storing program instructions that when executed cause an executing device to perform an image processing method, the method comprising:

determining a skin tone hue value;

determining if each pixel'"'"'s hue value in an image is within a range defined by a radial distance from the skin tone hue value;

for pixels that are not within the range, excluding the pixels'"'"' values form the image processing;

for pixels that are within the range, increasing a score of each such pixel;

performing facial recognition based on the remaining pixel scores, detecting parts of a face using shape detection, and

recognizing specific ranges of distance from the skin tone hue value correlated with specific parts of the face.

10. The non-transitory machine readable storage device of claim 9, where the instructions comprise:

determining the radial distance for each pixel'"'"'s hue value from the skin tone hue value; and

increasing the scores proportionately to the determined distance.

11. The non-transitory machine readable storage device of claim 9, where the instructions comprise:

modifying pixels'"'"' scores based on a search characteristic related to the shape detection.

12. The non-transitory machine readable storage device of claim 11, where the instructions comprise:

normalizing the detected face to a specified scale and orientation.

13. The non-transitory machine readable storage device of claim 12, where the instructions comprise:

comparing the normalized face to a reference image.

14. The non-transitory machine readable storage device of claim 13, where the instructions comprise:

creating a matching map by subtracting the normalized face and the reference image from each other using the pixels'"'"' scores based on a set of rules;

deciding if the reference image and normalized face match;

if the reference image and normalized face match, performing a device operation;

if the reference image and normalized face are not a match, comparing the normalized face to a different reference image.

15. The non-transitory machine readable storage device of claim 9, where the instructions comprise:

determining a center of mass of a face; and

tracking facial motion based on the changing center of mass.

16. The non-transitory machine readable storage device of claim 9, wherein the skin tone hue has a value of 25 and the range is within 30 degrees of the skin tone hue.

17. An image processing method, the method comprising:

determining a skin tone hue value;

determining if each pixel'"'"'s hue value in an image is within a range defined by a radial distance from the skin tone hue value;

for pixels that are not within the range, excluding the pixels'"'"' values from the image processing;

for pixels that are within the range, increasing a score of each such pixel;

performing facial recognition based on the remaining pixel scores,

using shape detection to detect parts of a face, and modifying pixels scores based on a search characteristic related to the shape detection, and

recognizing specific ranges of distance from the skin tone hue value correlated with specific parts of the face.

18. An image processing method, the method comprising:

converting a color image to a hue image;

determining a skin tone hue value;

determining if each pixel'"'"'s hue value in the hue image is within a range defined by a radial distance from the skin tone hue value;

for pixels that are not within the range, excluding the pixels'"'"' values from the image processing;

for pixels that are within the range, increasing a score of each such pixel;

performing facial recognition based on the remaining pixel scores, detecting parts of a face using shape detection, and

recognizing specific ranges of distance from the skin tone hue value correlated with specific parts of the face.