Method of verifying the presence of an eye in a close-up image

• US 6,028,949 A
• Filed: 12/02/1997
• Issued: 02/22/2000
• Est. Priority Date: 12/02/1997
• Status: Expired due to Term

• Current Assignee: Raymond A. Mckendall
• Original Assignee: Raymond A. Mckendall
• Inventors: McKendall, Raymond A.
• Primary Examiner(s): Boudreau, Leo H.
• Assistant Examiner(s): Cage, Shawn B
• Application Number:
  US08/982,365
• Time in Patent Office: 812 Days
• Field of Search: 382/115, 382/117, 351/204, 351/205, 351/206
• US Class Current: 382/117
• CPC Class Codes: G06V 40/168 Feature extraction; Face representation

Abstract

A method for verifying that an eye is present and accurately located in an image that is used for identifying an individual using iris identification techniques begins with an image in which the eye is believed to be present at a selected location. The image is then subjected to a series of tests to confirm that an eye is present at the selected location.

Claims

1. A method for verifying that an eye is present in an image comprised of pixels each having an x, y coordinate location in the image and determining its location comprised of the steps of:

a. selecting an image which is believed to contain an eye having an iris comprised of a set of pixels and a pupil comprised of a second set of pixels, a radius of the pupil being given as r_p and a radius of the iris being given as r_i ; and

the eye is believed to be located at a selected location;

b. computing r_p /r_i

c. if r_p /r_i is less than 0.5 then determining a centerpoint x_i ;

y_i of the iris and a centerpoint x_p ;

y_p of the pupil and calculating

space="preserve" listing-type="equation">D=((x.sub.i -x.sub.p).sup.2 +(y.sub.i,-y.sub.p).sup.2).sup.1/2

d. if D<

0.1 r_i reporting an eye image is present at the selected location; and

e. if r_p /r_i is greater than or equal to 0.5 or D is greater than or equal to 0.1 r_i then concluding that an eye is not present at the selected location.

2. The method of claim 1 also comprising the steps of;

a. computing a mean M of values of all pixels in a region corresponding to the pupil in the image;

b. if M is not less than 85 then computing a standard deviation S of values of all pixels in a region corresponding to the pupil in the image; and

c. if M is greater than 85 or if S is greater than 0.1M then concluding that an eye is not present at the selected location.

3. The method of claim 1 also comprising the steps of:

f. growing an expanded pupil to include all pixels adjacent the pupil pixels which have a value V within the range M-2S<

V<

M+2S;

g. computing a centroid of the expanded pupil; and

h. repeating steps b through g for the expanded pupil.

4. The method of claim 1 also comprising the step of comparing at least a portion of the image to a portion of a file image believed to be a match.

5. The method of claim 1 wherein all steps are performed by a computer.

6. The method of claim 1 also comprising the step of storing the image which is selected.

7. A method for verifying that an eye is present in an image comprised of pixels each having an x, y coordinate location in the image and determining its location comprised of the steps of:

a. selecting an image which is believed to contain an eye having an iris comprised of a set of pixels and a pupil comprised of a second set of pixels, a radius of the pupil being given as r_p and a radius of the iris being given as r_i ; and

the eye is believed to be located at a selected location;

b. computing r_p /r_i

c. if r_p /r_i is less than 0.5 then determining a centerpoint x_i ;

y_i of the iris and a centerpoint x_p ;

y_p of the pupil and calculating

space="preserve" listing-type="equation">D=((x.sub.i -x.sub.p).sup.2 +(y.sub.i -y.sub.p).sup.2).sup.1/2

d. if D<

0.1 r_i reporting an eye image is present at the selected location;

e. if r_p /r_i is greater than or equal to 0.5 or D is greater than or equal to 0.1 r_i then concluding that an eye is not present at the selected location;

f. if r_p /r_i is less than 0.5 and D is less than 0.1 r_i then computing a mean M of values of all pixels in a region corresponding to the pupil in the image;

g. if M is not less than 85 then computing a standard deviation S of values of all pixels in a region corresponding to the pupil in the image;

h. if M is greater than or equal to 85 or if S is greater than or equal to 0.1M then concluding that an eye is not present at the selected location;

i. if M is less than 85 and if S is less than 0.1M selecting each pixel adjacent the pupil pixels and for each selected pixel determining if a value for that pixel V is within the range M-2S<

V<

M+2S;

j. growing an expanded pupil to include all pixels adjacent the pupil pixels which have a value V within the range M-2S<

V<

M+2S;

k. computing a centroid of the expanded pupil; and

l. repeating steps b through k for the expanded pupil.

8. The method of claim 7 also comprising the step of comparing at least a portion of the image to a portion of a file image believed to be a match.

9. The method of claim 7 wherein all steps are performed by a computer.

10. The method of claim 7 also comprising the step of storing the image which is selected.

11. An improved method for identifying a particular individual of the type in which an image of an eye of a human having a pupil and an iris is acquired, the eye is determined to be present at a selected location in the image, a radius of the pupil being given as r_p and a radius of the iris being given as r_i ;

, an iris portion of the eye is isolated, and an iris code is generated and compared to a previously generated iris code wherein the improvement comprises verifying that an eye is present in the image at a selected location prior to generating an iris code in which verifying is comprised of;

b. computing r_p /r_i

c. if r_p /r_i is less than 0.5 then determining a centerpoint x_i ;

y_i of the iris and a centerpoint x_p ;

y_p of the pupil and calculating

space="preserve" listing-type="equation">D=((x.sub.i -x.sub.p).sup.2 +(y.sub.i -y.sub.p).sup.2).sup.1/2

d. if D<

0.1 r_i reporting an eye image is present at the selected location; and

e. if r_p /r_i is greater than or equal to 0.5 or D is greater than or equal to 0.1 r_i then concluding that an eye is not present at the selected location.

12. The method of claim 11 also comprising the steps of;

a. computing a mean M of values of all pixels in a region corresponding to the pupil in the image;

b. if M is not less than 85 then computing a standard deviation S of values of all pixels in a region corresponding to the pupil in the image; and

c. if M is greater than 85 or if S is greater than 0.1M then concluding that an eye is not present at the selected location.

13. The method of claim 11 also comprising the steps of:

f. growing an expanded pupil to include all pixels adjacent the pupil pixels which have a value V within the range M-2S<

V<

M+2S;

g. computing a centroid of the expanded pupil; and

h. repeating steps b through g for the expanded pupil.

14. The method of claim 11 also comprising the step of comparing at least a portion of the image to a portion of a file image believed to be a match.

15. The method of claim 11 wherein all steps are performed by a computer.

16. The method of claim 11 also comprising the step of storing the image which is selected.