Determining the age of a human subject in a digital image

• US 20060045352A1
• Filed: 09/01/2004
• Published: 03/02/2006
• Est. Priority Date: 09/01/2004
• Status: Active Grant

• Current Assignee: Monument Peak Ventures, LLC (Dominion Harbor Enterprises, LLC)
• Original Assignee: Eastman Kodak Company
• Inventors: Gallagher, Andrew C.
• Granted Patent: US 8,000,505 B2
• Time in Patent Office: Days
• US Class Current: 382/224
• CPC Class Codes: G06V 40/178 estimating age from face image; using age information for improving recognition
• CPC Class Codes: G06V 40/193 Preprocessing; Feature extraction

Abstract

The age class of a human subject is ascertained in a digital image. The subject has a redeye defect pair; each defect having one or more defect pixels. In the method, two regions of pixels corresponding to the defects are identified. The distance between the regions is measured. A region size is determined, based upon the size of at least one of the regions. An age class is determined from the distance and region size.

Claims

1. A method for ascertaining the age class of a human subject in a digital image, the human subject having a redeye defect pair, each defect of the redeye defect pair having one or more defect pixels, the method comprising the steps of:

identifying two regions of pixels corresponding to the defects;

measuring a distance between said regions;

determining a region size based upon a size of at least one of said regions; and

determining an age class from said distance and said region size.

2. The method of claim 1 wherein said distance is calculated from corresponding points in each of said regions.

3. The method of claim 1 wherein said measuring of said distance further comprises determining a centroid of each said region and determining a separation of said centroids.

4. The method of claim 1 wherein said measuring of said distance further comprises measuring a separation of said eyes, determining an out-of-plane rotation of the head of the human subject relative to a subject plane defined by said digital image, and correcting said separation for said rotation.

5. The method of claim 1 wherein said region size is a count of the number of pixels in one or both of the redeye defects.

6. The method of claim 1 wherein said determining said region size further comprises measuring a size of each of said regions and determining an average of said sizes of each of said regions.

7. The method of claim 6 wherein the size of each of said regions is the square root of the product of a count of the number of pixels in a respective defect times the inverse of the mathematical constant π

.

8. The method of claim 6 wherein the size of each of said regions is the radius of the smallest circle that inscribes the respective defect.

9. The method of claim 6 wherein said measuring of said distance further comprises determining a centroid of each of said regions;

and the size of each of said regions is the radius of the smallest circle that inscribes the respective defect.

10. The method of claim 6 wherein said measuring of said distance further comprises determining a centroid of each of said regions;

and the size Sc of each of said regions is given by;

$S_C=\underset{i}{\max }<br><br>\left[\sqrt{{\left(x_i-x_C\right)}^2+{\left(y_i-y_C\right)}^2}\right]$

wherein;

(x_i, y_i) are the coordinate locations of the i^th pixel of the respective defect; and

(x_C, y_C) is the centroid of the respective defect.

11. The method of claim 6 wherein said measuring of said distance further comprises determining a centroid of each of said regions;

and the size of each of said regions is the largest circle centered at the centroid of the respective defect that inscribes only defect pixels.

12. The method of claim 1 further comprising inducing pupil dilation in the human subject;

capturing the digital image with a camera during said pupil dilation.

13. The method of claim 1 wherein said camera defines an optical axis and said method further comprises illuminating said human subject, during said capturing, with a light source positioned close to said optical axis of said camera.

14. The method of claim 1 wherein determining said age class further comprises accessing a look-up table value corresponding to said distance and said region size.

15. The method of claim 1 wherein said determining said age class further comprises computing said age class with an age classifier.

16. The method of claim 15 wherein said computing utilizes said size, said distance, and one or more additional features.

17. The method of claim 16 wherein said additional features include at least one of:

wrinkle analysis and hair color analysis.

18. The method of claim 15 wherein said determining said age class further comprises selecting said age class and assigning a probability.

19. The method of claim 18 wherein said age range is selected from the categories:

baby, child, adult, and older adult.

20. The method of claim 15 wherein said age classifier is a Bayesian classifier.

21. The method of claim 15 wherein said determining of said region size further comprises computing a redeye feature F and using said redeye feature F in said age classifier, said redeye feature F being given by:

$F=\mathrm{int}<br><br>\left(\frac{S-B}{R_D-B}<br><br>K\right)+1$

wherein;

S is said region size;

B is a function of pixels of blur in the image;

K is a constant multiplier; and

R_D is said distance.

22. The method of claim 15 wherein RD is in units of pixels and is given by:

R_D=√

{square root over ((x₁−

x₂)²+(y₁−

y₂)²)}

wherein (x₁, y₁) and (x₂, y₂) are the coordinate locations of corresponding said pixels of each defect of the redeye defect pair.

23. The method of claim 21 wherein said determining further comprises measuring a size of each of said regions and determining an average of said sizes of each of said regions.

24. The method of claim 15 wherein said measuring of said distance further comprises determining an angle of out-of-plane rotation of the head of the human subject relative to a subject plane defined by said digital image;

and said determining of said region size further comprises computing a redeye feature F and using said redeye feature F in said age classifier, said redeye feature F being given by;

$F=\mathrm{int}<br><br>\left(\frac{\cos <br><br>\text{\hspace{1em}<br><br>}<br><br>A<br><br>\text{\hspace{1em}<br><br>}<br><br>\left(S-B\right)}{R_D-B}<br><br>K\right)+1$

wherein;

B is a function of pixels of blur in the image;

K is a constant multiplier;

A is said angle of out-of-plane rotation;

S is said region size; and

R_D is said distance.

25. The method of claim 24 wherein B=2 and K=240⁻

1.

26. The method of claim 24 wherein RD is in units of pixels and is given by:

R_D=√

{square root over ((x₁−

x₂)²+(y₁−

y₂)²)}

wherein (x₁, y₁) and (x₂, y₂) are the coordinate locations of corresponding said pixels of each defect of the redeye defect pair.

27. The method of claim 24 wherein said determining further comprises measuring a size of each of said regions and determining an average of said sizes of each of said regions.

28. A system for ascertaining the age class of a human subject in a digital image, the human subject having a redeye defect pair, each defect of the redeye defect pair having one or more defect pixels, the apparatus comprising:

means for identifying two regions of pixels corresponding to the defects;

means for measuring a distance between said regions;

means for determining a region size based upon a size of at least one of said regions; and

means for determining an age class from said distance and said region size.

29. A computer program product for ascertaining the age class of a human subject in a digital image, the human subject having a redeye defect pair, each defect of the redeye defect pair having one or more defect pixels, the computer program product comprising computer readable storage medium having a computer program stored thereon for performing the steps of:

identifying two regions of pixels corresponding to the defects;

measuring a distance between said regions;

determining a region size based upon a size of at least one of said regions; and

determining an age class from said distance and said region size.

30. A photofinishing method comprising:

inputting a photofinishing order having one or more digital images;

ascertaining an age class of a human subject in at least one of said digital images, the human subject having a redeye defect pair, each defect of the redeye defect pair having one or more defect pixels, said ascertaining including the steps of;

identifying two regions of pixels corresponding to the defects,

measuring a distance between said regions,

determining a region size based upon a size of at least one of said regions, and

determining an age class from said distance and said region size; and

modifying fulfillment of said photofinishing order responsive to said ascertained age class.

31. The method of claim 30 wherein said modifying further comprises selecting one of a plurality of advertisements responsive to said ascertaining, and providing said advertisement in said fulfillment.