Method and apparatus for identifying individuals

• US 6,144,754 A
• Filed: 03/25/1998
• Issued: 11/07/2000
• Est. Priority Date: 03/28/1997
• Status: Expired due to Fees

• Current Assignee: OKI Electric Industry Company Limited
• Original Assignee: OKI Electric Industry Company Limited
• Inventors: Okano, Kenji, Kuno, Yuji
• Primary Examiner(s): MEHTA, BHAVESH M
• Application Number:
  US09/047,482
• Time in Patent Office: 958 Days
• Field of Search: 382/117, 396/51, 351/210, 351/209, 351/212
• US Class Current: 382/117
• CPC Class Codes: G06V 40/168 Feature extraction; Face representation
• CPC Class Codes: G06V 40/18 Eye characteristics, e.g. of the iris

Abstract

An individual identification system and method for identifying individuals having elliptical pupils and elliptical irises. In the individual identification system, an image storage unit stores eye images, an the ellipse preparation unit prepares a plurality of ellipses, and a pixel addition unit adds up pixels laid along circumferences of the ellipses. An ellipse selection unit selects ellipses representative of the pupil and the iris among the plurality of ellipses based upon the sums of the pixels given by the pixel addition unit, whereby the parameters of the selected ellipses are output.

Claims

1. An individual identification system which identifies individuals by comparing irises thereof, comprising:

a first closed curved-line preparation circuit preparing a first closed curved-line similar to the iris;

a first similarity calculation circuit calculating a first similarity between the iris and the first closed curved-line;

a first judgment circuit judging whether or not the first similarity satisfies a first predetermined threshold similarity value;

a first decision circuit deciding that the first closed curved-line is representative of the outline of the iris upon judgment that the first similarity satisfies the first predetermined threshold similarity value;

a second closed curved-line preparation circuit preparing a second closed curved-line similar to the iris upon judgment that the first similarity does not satisfy the first predetermined threshold similarity value;

a second similarity calculation circuit calculating a second similarity between the iris and the second closed curved-line;

a second judgment circuit judging whether or not the second similarity satisfies a second predetermined threshold similarity value; and

a second decision circuit deciding that the second closed curved-line is representative of the outline of the iris upon judgment that the second similarity satisfies the second predetermined threshold similarity value, said second decision circuit selecting a second closed curved-line that is the most similar to the iris among a plurality of second closed curvedlines upon judgment that no second similarity satisfies the second predetermined threshold similarity value, and deciding that the selected second closed curved-line is representative of the outline of the iris.

2. The iris position detecting system claimed in claim 1, wherein the first similarity calculation circuit and/or the second similarity calculation circuit calculates the Hamming distance of a first iris code with respect to a second iris code, said Hamming distance being defined as the number of all the bits of the first iris code and the number of bits by which the first and second iris codes disagree with each other.

3. The iris position detecting system claimed in claim 1 wherein the outline of the iris is a boundary between a pupil and the iris.

4. The iris position detecting system claimed in claim 1 wherein the outline of the iris is a boundary between the iris and a sclera.

5. An individual identification system which identifies individuals by comparing irises thereof, comprising:

a first closed curved-line preparation circuit preparing a first closed curved-line similar to the iris;

a first similarity calculation circuit calculating a first similarity between the iris and the first closed curved-line;

a first judgment circuit judging whether or not the first similarity satisfies a first predetermined threshold similarity value;

a first decision circuit deciding that the first closed curved-line is representative of the outline of the iris upon judgment that the first similarity satisfies the first predetermined threshold similarity value;

a second closed curved-line preparation circuit preparing a second closed curved-line similar to the iris upon judgment that the first similarity does not satisfy the first predetermined threshold similarity value;

a second similarity calculation circuit calculating a second similarity between the iris and the second closed curved-line;

a second judgment circuit judging whether or not the second similarity satisfies a second predetermined threshold similarity value; and

a second decision circuit deciding that the second closed curved-line is representative of the outline of the iris upon judgment that the second similarity satisfies the second predetermined threshold similarity value, and wherein said first closed curved-line is an ellipse similar to the iris, and the second closed curved-line is an ellipse similar to the first closed curved-line,

wherein the second closed curved-line prepared by the second closed curved-line preparation circuit has changing parameters of an ellipse, and

wherein a parameter of said ellipse is a ratio of a line of apsides to a minor axis of said ellipse.

6. The iris position detecting system claimed in claim 5, wherein the first similarity calculation circuit and/or the second similarity calculating circuit calculates the Hamming distance of a first iris code with respect to a second iris code, said Hamming distance being defined as the number of all the bits of the first iris code and the number of bits by which the first and second iris codes disagree with each other.

7. An individual identification system which identifies individuals by comparing irises thereof, comprising:

a first closed curved-line preparation circuit preparing a first closed curved-line similar to the iris;

a first similarity calculation circuit calculating a first similarity between the iris and the first closed curved-line;

a first judgment circuit judging whether or not the first similarity satisfies a first predetermined threshold similarity value;

a first decision circuit deciding that the first closed curved-line is representative of the outline of the iris upon judgment that the first similarity satisfies the first predetermined threshold similarity value;

a second closed curved-line preparation circuit preparing a second closed curved-line similar to the iris upon judgment that the first similarity does not satisfy the first predetermined threshold similarity value;

a second similarity calculation circuit calculating a second similarity between the iris and the second closed curved-line;

a second judgment circuit judging whether or not the second similarity satisfies a second predetermined threshold similarity value; and

a second decision circuit deciding that the second closed curved-line is representative of the outline of the iris upon judgment that the second similarity satisfies the second predetermined threshold similarity value, and wherein said first closed curved-line is an ellipse similar to the iris, and the second closed curved-line is an ellipse similar to the first closed curved-line,

wherein the second closed curved-line prepared by the second closed curved-line preparation circuit has changing parameters of an ellipse, and

wherein a parameter of said ellipse is an inclination of a line of apsides to an X coordinate axis of said ellipse.

8. The iris position detecting system claimed in claim 7 wherein a parameter of said ellipse is the center coordinate thereof.

9. An apparatus for identifying an individual from the texture of the iris of the individual'"'"'s eye, comprising:

an image storage unit for storing images of a plurality of human individuals'"'"' eyes, each of said eyes having a pupil and an iris therein;

a closed curved-line preparation unit including;

a circle preparation unit for generating circles as the closed curved-lines; and

an ellipse preparation unit for generating ellipses as the closed curved lines;

an iris analysis unit coupled to said image storage unit and said closed curved-line preparation unit for analyzing irises stored in said image storage unit based upon the closed curved-lines generated by said closed curved-line preparation unit, said iris analysis unit including;

a circle analysis unit for selecting first and second circles from said circle preparation unit that correspond most closely to the outlines of a pupil and an iris stored in said storage unit, said circle analysis unit further establishing a polar coordinate system based upon said first and second circles; and

an ellipse analysis unit for selecting first and second ellipses from said ellipse preparation unit that correspond most closely to the outlines of the pupil and the iris stored in said storage unit, said ellipse analysis unit further establishing a polar coordinate system based upon said first and second ellipses;

an iris storage unit for storing a plurality of iris codes, each of said iris codes corresponding to the texture of an iris forming an image stored in said image storage unit;

a comparison result output unit; and

an iris comparison unit coupled to said iris analysis unit, said closed curved-line preparation unit, said iris storage unit and said comparison result output unit for comparing an iris code stored in said image storing unit with the plurality of iris codes stored in said iris storage unit, said iris comparison unit including;

a similarity calculation unit for calculating the similarity between an iris code to be identified, stored in said image storing unit, and the iris codes stored in said iris storage unit; and

a similarity judgment unit for determining whether the iris code to be identified agrees or does not agree with an iris code stored in said iris storage unit, whereby if said codes agree said similarity judgment unit outputs a signal indicating coincidence between said iris codes, and if said codes do not agree said similarity judgment unit outputs a signal identifying an iris code most similar to the iris code to be identified,

wherein said similarity calculating unit calculates the similarity between the iris code to be identified which is stored in said image storing unit, and the iris codes stored in said iris storage unit by calculating the Hamming distance between the iris code to be identified and each of the iris codes stored in said iris storage unit, and

wherein said similarity judgment unit determines whether the lowest Hamming distance calculated is less than a predetermined Hamming distance, and wherein if the lowest Hamming distance calculated is less than said predetermined Hamming distance, said similarity judgment unit outputs a signal denoting a coincidence between the iris code to be identified and an iris code stored in said iris storage unit, and if the lowest Hamming distance calculated is not less than said predetermined Hamming distance, said similarity judgment unit outputs a signal denoting a lack of coincidence between the iris code to be identified and an iris code stored in said iris storage unit.

10. A method of identifying an individual by comparing the texture of the iris of the individual'"'"'s eye with a plurality of ellipses, the ellipse being prepared by steps comprising:

(a) generating a first ellipse having a center Xc, Yc and a line of apsides A having a ratio β

thereof to a minor axis of the ellipse, a circumference of the ellipse having a plurality of pixels each designated by a pixel number M;

(b) incrementing the value of M up to a variable pixel sum PS;

(c) deciding whether the incremented value of M is equal to or greater than a maximum value Mmax, if a pixel sum PS is not equal to or greater than Mmax repeating steps (a) and (b), and if a pixel sum PS is equal to or greater than Mmax storing said pixel sum, the stored pixel sums PS corresponding to ellipses representing the boundary of the iris and a pupil of the individual'"'"'s eye;

(d) incrementing said line of apsides A in discrete steps;

(e) deciding whether the incremented value obtained by step (d) is greater than a maximum value Amax, if the incremental value obtained by step (d) is not greater than Amax repeating steps (a), (b) and (c), and if the incremental value obtained by step (d) is greater than Amax storing said value of A; and

(f) searching said values of A to determine which of said values results in the largest difference between a first pixel sum PS corresponding to first value of A and a second pixel sum corresponding to a second value of A, whereby an ellipse is identified which is most similar to the boundary between the iris and the pupil of the individual'"'"'s eye.