Iris identification
An iris identification system for confirming identity of a person and method thereof are disclosed. The iris identification system comprises a mode converter for selecting one of registration and identification modes, an image input means, a luminance adjustment module, a registration module, an identification module, a reference iris image storage, and a main control unit. The registration module captures iris images when the registration mode is selected, preprocesses the captured iris image, defines the preprocessed iris image as main, auxiliary, negative main and compensation data, encodes the defined data, and registers the encoded data as data of the reference iris images; The identification module captures an iris image when the identification mode is selected, preprocesses the captured iris image, defines the preprocessed iris image as main, auxiliary, negative main and compensation data, encodes the defined data, and analyzes similarity both data.
1. An iris identification system for confirming an identity of a person, comprising:
a mode converter that selects one of registration and identification modes;
an image inputter that inputs an iris image;
a luminance adjustment module that sequentially selects predetermined luminance levels when the registration mode is selected, and irradiates light of the selected luminance level together with an infrared ray for iris registration;
a registration module that captures a plurality of reference iris images input when the light of the selected luminance level is irradiated, preprocesses each captured reference iris image to determine pupil boundary, pupil center, iris size, and class, categorizes each preprocessed reference iris image as at least one of main, auxiliary, negative main and compensation data, encodes the categorized data of each reference iris image, and registers the encoded data of the plurality of reference iris images as registered reference iris images;
an identification module that captures an iris image for identification input when the identification mode is selected, preprocesses the captured iris image for identification to determine pupil boundary, pupil center, iris size, and class, categorizes the preprocessed iris image for identification as at least one of main, auxiliary, negative main and compensation data, encodes the categorized data of the iris image for identification, and analyzes similarity between the encoded data of the iris image for identification and the registered reference iris images;
a reference iris image storage that stores the data of the registered reference iris images; and
a main control unit that controls the image inputter, the luminance adjustment module, the registration module, the identification module and the reference iris image storage according to the mode of the mode converter;
wherein each iris image captured by the registration module and the identification module is horizontally divided so as to form a plurality of horizontal bands, the horizontal bands are divided by a perpendicular passing through a center of the pupil such that a plurality of blocks are symmetrically formed, each block is divided into a plurality of rectangular sub-blocks and each rectangular sub-block is divided into a plurality of units;
wherein each band has a priority and a weighted value to be considered in the analysis; and
wherein each unit in a rectangular sub-block is classified as at least one of auxiliary data, main data, negative main data, and compensation data according to the luminance level of the unit.
2. The iris identification system of claim 1 ,
wherein units having luminance less than the average luminance of the rectangular sub-block are categorized as auxiliary data;
wherein, among units categorized as auxiliary data, units having luminance less than average luminance of auxiliary data and length larger than a predetermined length are categorized as main data, units having luminance greater than the average luminance of the rectangular sub-block and length larger than the predetermined length are categorized as negative main data; and
wherein units corresponding to 9/10˜41/10 of the average luminance of auxiliary data are categorized as compensation data.
3. The iris identification system of claim 1 ,
wherein a center of the pupil is calculated by obtaining a plurality of random pupil centers I i , extracting candidate pupil centers from the random pupil centers, and calculating a final pupil center T p (x p , y p ) using the candidate pupil centers.
4. The iris identification system of claim 3 ,
wherein the random pupil center I i is obtained by randomly selecting two points, S(x 1 , y 1 ) and E(x 2 , y 2 ), on the pupil boundary of the iris image, defining a segment SE by a line connecting the points S and E, defining a perpendicular to the segment SE extending from a center of the segment SE such that the perpendicular crosses the pupil boundary at a point C(x 3 , y 3 ), and calculating the random pupil center I i (x 0 , y 0 ) by the following relationships:
a
=
1
2
(
x
1
-
x
2
)
2
+
(
y
1
-
y
2
)
2
,
c
=
1
2
(
x
1
-
x
2
-
2
x
3
)
2
+
(
y
1
-
y
2
-
2
y
3
)
2
d
=
1
2
c
(
a
2
-
c
2
)
,
D
=
tan
-
1
(
y
1
-
y
2
x
1
-
x
2
)
-
π
2
,
x
0
=
d
·
cos
D
+
1
2
(
x
1
+
x
2
)
y
0
=
-
(
d
·
sin
D
+
1
2
(
y
1
+
y
2
)
)
wherein, a is half of the length of segment SE, c is a distance from the center of the segment SE to point C, and d is a distance from the center of the segment SE to the random pupil center I i (x 0 , y 0 ),
wherein the candidate pupil centers have radii that are in whole class range B, wherein, B is the ratio of the pupil radius to the iris radius and satisfies
1
5
≤
β
≤
4
5
;
and
wherein the final pupil center T p (x p , y p ) is obtained by the following equations:
x
p
=
1
n
∑
x
0
i
,
y
p
=
1
n
∑
y
0
i
wherein, n is number of the candidate centers, and x 0i , y 0i are respectively x-axis and y-axis coordinates of a center of each candidate.
5. The iris identification system of claim 1 ,
wherein the pupil boundary is determined by the following equation:
when I min <I b <I ma ,
I
mb
=
1
N
b
∑
I
b
wherein,
I
ma
=
1
N
a
∑
I
a
,
I a is luminance of a pixel in the iris image, I ma is an average luminance of the iris image, N a is number of pixels in the iris image, and I min is a minimum luminance limit.
6. An iris identification method for confirming an identity of a person, comprising:
selecting one of registration and identification modes;
inputting an iris image;
sequentially selecting predetermined luminance levels when the registration mode is selected, and irradiating light of the selected luminance level together with an infrared ray for iris registration;
capturing a plurality of reference iris images input when the light of the selected luminance level is irradiated, preprocessing each captured reference iris image to determine pupil boundary, pupil center, iris size, and class, categorizing each preprocessed reference iris image as at least one of main, auxiliary, negative main and compensation data, encoding the categorized data of each reference iris image, and registering the encoded data of the plurality of reference iris images as registered reference iris images;
storing the data of the registered reference iris images; and
capturing an iris image for identification input when the identification mode is selected, preprocessing the captured iris image for identification to determine pupil boundary, pupil center, iris size, and class, categorizing the preprocessed iris image for identification as at least one of main, auxiliary, negative main and compensation data, encoding the categorized data of the iris image for identification, and analyzing similarity between the encoded data of the iris image for identification and the registered reference iris images;
wherein each captured iris image is horizontally divided so as to form a plurality of horizontal bands, the horizontal bands are divided by a perpendicular passing through a center of the pupil such that a plurality of blocks are symmetrically formed, each block is divided into a plurality of rectangular sub-blocks and each rectangular sub-block is divided into a plurality of units;
wherein each band has a priority and a weighted value to be considered in the analysis; and
wherein each unit in a rectangular sub-block is classified as at least one of auxiliary data, main data, negative main data, and compensation data according to the luminance level of the unit.
7. The iris identification method of claim 6 ,
wherein units having luminance less than the average luminance of the rectangular sub-block are categorized as auxiliary data;
wherein, among units categorized as auxiliary data, units having luminance less than average luminance of auxiliary data and length larger than a predetermined length are categorized as main data, units having luminance greater than the average luminance of the rectangular sub-block and length larger than the predetermined length are categorized as negative main data; and
wherein units corresponding to 9/10˜11/10 of the average luminance of auxiliary data are categorized as compensation data.
8. The iris identification method of claim 6 ,
wherein a center of the pupil is calculated by obtaining a plurality of random pupil centers I i , extracting candidate pupil centers from the random pupil centers, and calculating a final pupil center T p (x p , y p ) using the candidate pupil centers.
9. The iris identification method of claim 8
wherein the random pupil center I i is obtained by random selecting two points, S(x 1 , y 1 ) and E (x 2 , y 2 ), on the pupil boundary of the iris image, defining a segment SE by a line connecting the points S and E, defining a perpendicular to the segment SE extending from a center of the segment SE such that the perpendicular crosses the pupil boundary at a point C(x 3 , y 3 ), and calculating the random pupil center I i (x 0 , y 0 ) by the following relationships:
a
=
1
2
(
x
1
-
x
2
)
2
+
(
y
1
-
y
2
)
2
,
c
=
1
2
(
x
1
-
x
2
-
2
x
3
)
2
+
(
y
1
-
y
2
-
2
y
3
)
2
d
=
1
2
c
(
a
2
-
c
2
)
D
=
tan
-
1
(
y
1
-
y
2
x
1
-
x
2
)
-
π
2
,
x
0
=
d
·
cos
D
+
1
2
(
x
1
+
x
2
)
y
0
=
-
(
d
·
sin
D
+
1
2
(
y
1
+
y
2
)
)
wherein, a is half of the length of segment SE, c is a distance from the center of the segment SE to point C, and d is a distance from the center of the segment SE to the random pupil center I i (x 0 , y 0 ),
wherein the candidate pupil centers have radii that are in whole class range β, wherein, β is the ratio of the pupil radius to the iris radius and satisfies
1
5
≤
β
≤
4
5
;
and
wherein the final pupil center T p (x p , y p ) is obtained by the following equation:
x
p
=
1
n
∑
x
0
i
,
y
p
=
1
n
∑
y
0
i
wherein, n is number of the candidate centers, and x 0i , y 0i are respectively x-axis and y-axis coordinates of a center of each candidate.
10. The iris identification method of claim 6 ,
wherein the pupil boundary is determined by the following equation:
when I min <I b <I ma ,
I
mb
=
1
N
b
∑
I
b
wherein,
I
ma
=
1
N
a
∑
I
a
,
I a is luminance of a pixel in the iris image, I ma is an average luminance of the iris image, N a is number of pixels in the iris image, and I min is a minimum luminance limit.
11. A computer readable medium for storing a program that implements iris identification to confirm an identity of a person, comprising:
a mode selecting segment that selects one of registration and identification modes;
an inputting segment that inputs an iris image;
a luminance level segment that sequentially selects predetermined luminance levels when the registration mode is selected, and irradiates light of the selected luminance level together with an infrared ray for iris registration;
an image capturing segment that captures a plurality of reference iris images input when the light of the selected luminance level is irradiated, preprocesses each captured reference iris image to determine pupil boundary, pupil center, iris size, and class, categorizes each preprocessed reference iris image as at least one of main, auxiliary, negative main and compensation data, encodes the categorized data of each reference iris image, and registers the encoded data of the plurality of reference images as registered reference iris images;
a data storing segment that stores the data of the registered reference iris images; and
an image capturing segment that captures an iris image for identification input when the identification mode is selected, preprocesses the captured iris image for identification to determine pupil boundary, pupil center, iris size, and class, categorizes the preprocessed iris image for identification as at least one of main, auxiliary, negative main and compensation data, encodes the categorized data of the iris image for identification, and analyzes similarity between the encoded data of the iris image for identification and the registered reference iris images;
wherein each captured iris image is horizontally divided so as to form a plurality of horizontal bands, the horizontal bands are divided by a perpendicular passing through a center of the pupil such that a plurality of blocks are symmetrically formed, each block is divided into a plurality of rectangular sub-blocks and each rectangular sub-block is divided into a plurality of units;
wherein each band has a priority and a weighted value to be considered in the analysis; and
wherein each unit in a rectangular sub-block is classified as at least one of auxiliary data, main data, negative main data, and compensation data according to the luminance level of the unit.
12. The computer readable storage medium of claim 11 ,
wherein units having luminance less than the average luminance of the rectangular sub-block are categorized as auxiliary data;
wherein, among units categorized as auxiliary data, units having luminance less than average luminance of auxiliary data and length larger than a predetermined length are categorized as main data, units having luminance greater than the average luminance of the rectangular sub-block and length larger than the predetermined length are categorized as negative main data; and
wherein units corresponding to 9/10˜11/10 of the average luminance of auxiliary data are categorized as compensation data.
13. The computer readable storage medium of claim 11 ,
wherein a center of the pupil is calculated by obtaining a plurality of random pupil centers I i , extracting candidate pupil centers from the random pupil centers, and calculating a final pupil center T p (x p , y p ) using the candidate pupil centers.
14. The computer readable storage medium of claim 13 ,
wherein the random pupil center I i , is obtained by randomly selecting two points, S(x 1 , y 1 ) and E(x 2 , y 2 ), on the pupil boundary of the iris image, defining a segment SE by a line connecting the points S and E, defining a perpendicular to the segment SE extending from a center of the segment SE such that the perpendicular crosses the pupil boundary at a point C(x 3 , y 3 ), and calculating the random pupil center I i (x 0 , y 0 ) by the following relationships:
a
=
1
2
(
x
1
-
x
2
)
2
+
(
y
1
-
y
2
)
2
,
c
=
1
2
(
x
1
-
x
2
-
2
x
3
)
2
+
(
y
1
-
y
2
-
2
y
3
)
2
d
=
1
2
c
(
a
2
-
c
2
)
D
=
tan
-
1
(
y
1
-
y
2
x
1
-
x
2
)
-
π
2
,
x
0
=
d
·
cos
D
+
1
2
(
x
1
+
x
2
)
y
0
=
-
(
d
·
sin
D
+
1
2
(
y
1
+
y
2
)
)
wherein, a is half of the length of segment SE, c is a distance from the center of the segment SE to point C, and d is a distance from the center of the segment SE to the random pupil center I i (x 0 , y 0 ),
wherein the candidate pupil centers have radii that are in whole class range β, wherein, β is the ratio of the pupil radius to the iris radius and satisfies
1
5
≤
β
≤
4
5
;
wherein the final pupil center T p (x p , y p ) is obtained by the following equation:
x
p
=
1
n
∑
x
0
i
,
y
p
=
1
n
∑
y
0
i
wherein, n is number of the candidate centers, and x 0i , y 0i are respectively x-axis and y-axis coordinates of a center of each candidate.
15. The computer readable storage medium of claim 11 ,
wherein the pupil boundary is determined by the following equation:
when I min <I b <I ma ,
I
mb
=
1
N
b
∑
I
b
wherein,
I
ma
=
1
N
a
∑
I
a
,
I a is luminance of a pixel in the iris image, I ma is an average luminance of the iris image, N a is number of pixels in the iris image, and I min is a minimum luminance limit.