KR20060096867A - Method and apparatus for setting of comparison area and generating of user authentication information for iris recognition - Google Patents

Abstract

 The present invention relates to a method and apparatus for setting a comparison area and generating user authentication information for iris recognition, the method comprising: (a) obtaining coordinate average points of pixel points of a pupil area in an image image of an eye; (b) setting the concentric circles such that a predetermined portion of the iris region is included around the coordinate average point of the pupil region; (c) obtaining coordinate mean points of pixel points of the excluded concentric region formed by excluding the pupil region in the concentric region; (d) dividing the concentric circles according to the coordinate average points of the pupil regions and the coordinate average points of the excluded concentric circles; And (e) setting one of the divided concentric circles as a comparison region for iris recognition. Accordingly, a method and apparatus for setting a comparison area and generating user authentication information for iris recognition with little false recognition rate according to a user's photographing posture can be provided.

Description

 Method and apparatus for setting of comparison area and generating of user authentication information for iris recognition

1 shows a process of topologically analyzing an iris image according to the present invention.

2A and 2B illustrate a process of setting a data extraction region, which is a partial region that is analyzed topologically among iris images according to the present invention.

3 illustrates a process of obtaining topological analysis data from the data extraction region set in FIG.

4 illustrates a preferred embodiment of a data structure storing original data corresponding to reference luminance values of original registrants according to the present invention.

5 shows a security method using iris recognition in accordance with the present invention,

6 illustrates a process of verifying an identity based on original data and comparative data according to the present invention.

Figure 7a is a block diagram of a preferred embodiment of a security device using iris recognition in accordance with the present invention,

FIG. 7B is a block diagram illustrating the data area setting unit of FIG. 7A in more detail.

FIG. 7C is a block diagram illustrating the iris information generation unit of FIG. 7A in more detail.

FIG. 8 illustrates an embodiment of setting a data extraction region through the process of FIG. 2A according to the present invention.

FIG. 9 illustrates another embodiment of setting a data extraction region through the process of FIG. 2A according to the present invention.

FIG. 10 illustrates an embodiment of setting a data extraction region through the process of FIG. 2B according to the present invention.

FIG. 11 shows another embodiment of setting a data extraction region through the process of FIG. 2B according to the present invention.

 The present invention relates to a method and apparatus for setting a comparison area and generating user authentication information for iris recognition, and more particularly, sets a data extraction area based on coordinates of a pupil area and an iris area, and includes iris image information included therein. The present invention relates to a method and apparatus for setting a comparison area and generating user authentication information for iris recognition by analyzing the topologically.

The security method and apparatus used in the conventional iris recognition system take the image data of the iris part from the pupil of the original registrant and store it as the original data, and compare the identity of the subject of identification to one-to-one with one-to-one comparison with these to determine whether the same person I choose to check. However, in the conventional systems, the original data and the comparative data are compared with each other. Instead of setting a reference according to a certain characteristic of the iris itself and comparing the two according to the reference, the reference is based on the left and right eye tails of a person. By comparing or analyzing them according to a standard that is not related to the iris shape itself, or taking a differential value in quantifying image information based on the noise boundary of the pupil even when the iris shape itself is used as a reference, As a result, it is not efficient in a system requiring real time authentication, and there is a problem in miniaturization of a recognizer. As a result, there is a problem that it cannot be put to practical use.

 The technical problem to be achieved by the present invention is to set a data extraction region that is not significantly affected by the surrounding brightness, face movement and distance and position from the camera, and extracting topological data for identification from the set data extraction region. A method and apparatus for generating a comparison area and user authentication information for iris recognition are provided.

 According to another aspect of the present invention, there is provided a method of setting a comparison area for iris recognition, the method comprising: (a) obtaining coordinate mean points of pixel points of a pupil area in image information of an eye; (b) setting the concentric circles such that a predetermined portion of the iris region is included around the coordinate average point of the pupil region; (c) obtaining coordinate mean points of pixel points of the excluded concentric circle formed by excluding the pupil area in the concentric circle area; (d) dividing the concentric circles according to the coordinate average points of the pupil regions and the coordinate average points of the excluded concentric circles; And (e) setting one of the divided concentric circles as a comparison region for iris recognition.

In addition, the user authentication information generation method for iris recognition of the present invention for solving the technical problem, (a) setting a comparison area to extract the user authentication information from the image information of the eye; (b) extracting pixel points to be compared from pixel points included in the comparison area; (c) extracting topological feature information of pixel point number information, corner number information, and vertex number information of the pixel points to be compared; And (d) generating the user authentication information used as comparative authentication data in the iris recognition system by calculating the extracted topological feature information.

The comparative area setting device for iris recognition according to the present invention for solving the technical problem, the region separating unit for separating the pupil area and the iris region from the image information of the eye; A first coordinate average point calculator for obtaining a coordinate average point of the pupil area with respect to pixel points of the pupil area separated by the area separator; A concentric circle setting unit configured to set a concentric circle area such that a predetermined portion of the iris area is included around the coordinate average point of the pupil area; A second coordinate average point calculator configured to form a concentric circular region excluding the pupil region from the concentric region through the region separation unit, and obtain a coordinate average point of pixel points of the excluding the concentric circular region; A concentric circle division unit dividing the concentric circle area according to a coordinate average point of the pupil area and a coordinate average point of the excluded concentric circle area; And a comparison area setting unit configured to set one of the divided concentric areas as a comparison area for iris recognition.

In addition, the user authentication information generating device for iris recognition of the present invention for solving the technical problem, a comparison area setting unit for setting a comparison area to extract the user authentication information from the image information of the eye; A comparison object extracting unit extracting pixel points to be compared from pixel points included in the comparison area; A topological feature information extraction unit for extracting topological feature information about pixel point number information, corner number information, and vertex number information of the pixel points extracted from the comparison area; And a user authentication information generator for generating user authentication information by calculating the extracted topological feature information.

Hereinafter, with reference to the accompanying drawings will be described in detail a preferred embodiment of the present invention.

1 shows a process of topologically analyzing an iris image according to the present invention. Referring to FIG. 1, first, iris image information obtained by photographing a predetermined iris is obtained (101).

Next, a data extraction region for analyzing the iris image information is set (102). Here, the data extraction region is a region of the iris image information that is not disturbed by eyelashes or eyelids and is not affected by the position of the pupil. It is a geometric realm. Therefore, the topological analysis information in the data extraction region is not significantly influenced by the position of the identification subject.

Next, the user authentication information is generated through the topological feature information in the data extraction region (103).

According to FIG. 2A and the present invention, a process of setting a data extraction region, which is a partial region analyzed topologically, among iris images is shown. Referring to FIG. 2A, first, a pupil area is extracted from iris image information (211). Here, the extracted pupil region is composed of N pixels P (1), P (2), ..., P (N) which are binarized and preprocessed, and the pixel P (i) is a position coordinate (X ( i), P (i) through Y (i) = (X (i), Y (i)).

)을 구한다(212). 보다 구체적으로, 동공영역의 좌표평균점(

)의 위치좌표는 다음 수학식 1에서 보는 바와 같이 구하여진다.Next, the coordinate average point of the pupil area extracted in step 211 (

(212). More specifically, the coordinate mean point of the pupil area (

The positional coordinate of) is obtained as shown in Equation 1 below.

)을 중심으로 홍채영역의 일정부분이 포함되도록 동심원 영역(DC)을 설정한다(213). 동심원 영역(DC)을 설정함에 있어서, 좌표평균 점(

)을 중심으로 홍채영역의 일정부분이 포함되도록 하는 것은 동공반경(r)의 1배에서 2배, 바람직하게는 1.2배 내지 1.5배에 해당하는 영역을 설정하여, 속눈썹이나 눈꺼풀에 의해서 방해받지 않도록 한다. 여기에서, 동공반경(r)은 다음 수학식 2에서 보는 바와 같이 구하여진다.Next, the coordinate mean point (

The concentric circle DC is set to include a portion of the iris area around the center (213). In setting the concentric area DC, the coordinate mean point (

To include a certain portion of the iris area around the center of the iris area is set to 1 to 2 times, preferably 1.2 to 1.5 times the pupil radius (r) so as not to be disturbed by eyelashes or eyelids. do. Here, the pupil radius r is obtained as shown in Equation 2 below.

)을 구한다(214).Next, the coordinate average point of the concentric circles DC set in step 213 (

(214).

)과 좌표평균점(

)을 연결하는 제 1 직선(

) 및 좌표평균점(

)과 좌표평균점(

)의 중점(

)을 지나고 상기 제 1 직선(

)에 수직하는 제 2 직선(

)을 구한다(215).Next, the coordinate mean point (

) And the coordinate mean point (

First straight line connecting the

) And coordinate mean points (

) And the coordinate mean point (

) Emphasis

Past the first straight line (

2nd straight line perpendicular to)

(215).

)과 제 2 직선(

)을 통하여 동심원 영역(DC)를 4개의 영역(DC1, DC2, DC3, DC4)으로 분할하고, 각각의 분할된 영역에 대한 좌표평균점을 구하고, 좌표평균점중 렉시코그래픽 순서가 가장 작은 점을 포함하는 영역을 데이터 추출 영역(DF1)으로 설정한다(216).Next, the first straight line obtained in the step 215 (

) And the second straight line (

Divide the concentric circles (DC) into four regions (DC1, DC2, DC3, DC4), obtain the coordinate mean points for each divided region, and include the points with the smallest lexicographic order among the coordinate mean points. The area to be set is set as the data extraction area DF1 (216).

에 대하여

이

보다 작은 경우를

로 표시하며 다음의 괄호 속과 같이 정의한다.Here, the lexographic order is two points on the plane

about

this

Less than

It is defined as in the following parentheses.

로 나타난다.

Appears.

In addition, obtaining the coordinate average point for each divided region herein is referred to Equation 1 above.

The topological data for the security method using the iris recognition is extracted from the data extraction area DF1 configured as described above.

)과 분할된 영역의 좌표평균점을 지나는 직선을 통하여 2등분하고, 2등분된 영역의 각각에 대한 좌표평균점 중에서 렉시코그래픽 순서가 가장 작은 점을 포함하는 영역을 데이터 추출 영역(DF2)로 다시 설정할 수 있다. 이와 같은 동작을 반복적으로 수행함으로써 데이터 추출 영역(DFn) 의 범위를 좁혀나갈 수 있게 된다. 홍채인식을 이용한 보안 장치별로 해당하는 소정 범위까지 분할하여 그 중 해당하는 데이터 추출 영역을 설정할 수 있게 된다. 즉, 보안 장치별로 속도 및 처리의 정확성을 고려하여 필요한 범위까지 분할하여 데이터 추출 영역을 설정하게 된다.In FIG. 2A, the concentric circle DC is divided into four regions DC1, DC2, DC3, and DC4, and a region including a point having the smallest lexicographic order among coordinate average points for each divided region is included in the data extraction region. It was set to (DF1). Further, the data extraction area DF1 is referred to as the center point (

) Is divided into two through a straight line passing through the coordinate average point of the divided region, and the region including the point with the smallest lexicographic order among the coordinate average points for each of the divided regions is set again as the data extraction region (DF2). Can be. By repeatedly performing such an operation, the range of the data extraction area DFn can be narrowed. It is possible to set a corresponding data extraction area by dividing a predetermined range corresponding to each security device using iris recognition. That is, the data extraction area is set by dividing the security device into necessary ranges in consideration of speed and processing accuracy.

In addition, in FIG. 2A, the concentric circle DC is divided into four areas DC1, DC2, DC3, and DC4 to extract data including a point having the smallest lexicographic order among coordinate average points in each of the divided areas. Although described as setting to the area DF1, the concentric circle DC set in step 213 may be set as the data extraction area. More specifically, the concentric circle DC is shown in FIG. 8, and the concentric circle DC is divided and the corresponding area is shown in FIG. 9.

FIG. 8 illustrates an embodiment of setting a data extraction region through the process of FIG. 2A according to the present invention. Referring to FIG. 8, reference image 801 denotes an eyebrow line as an image of a human eye. Therefore, the upper portion of the line indicated by the reference number 801 is a portion that is not photographed in the actual image. Reference numeral 802 denotes the pupillary line, reference numeral 803 denotes the autonomic nerve line, reference numerals 804, 805, 806, 807 and 808 denote the hot collinearity present in the iris, and reference numeral 809 denotes the pupil of the pupil. An edge is shown and reference numeral 810 denotes an area set as a data extraction area. Reference numeral 811 denotes a concentric circle area DC. In FIG. 8, it can be seen that the entire concentric region DC is set as the data extraction region.

9 illustrates another embodiment of setting a data extraction region through the process of FIG. 2A according to the present invention. Referring to FIG. 9, reference numeral 901 denotes an eyebrow line as an image of a human eye. Therefore, the upper portion of the line indicated by reference numeral 901 is a portion which is not photographed in the actual image. Reference numeral 902 denotes a pupillary line, reference numeral 803 denotes an autonomic nerve line, reference numerals 904, 905, 906, 907, and 908 denote hot air lines present in the iris, and reference numeral 909 denotes a pupil of the pupil. An edge is shown and reference numeral 910 denotes an area set as a data area. Reference numeral 911 denotes a concentric circle area DC. In FIG. 9, it can be seen that a region including a point having the smallest lexicographic order among the coordinate average points of each of the divided regions by dividing the concentric circles DC is set as the data extraction region.

2B illustrates a process of setting a data extraction region, which is a partial region analyzed topologically, among iris images according to the present invention. Referring to FIG. 2B, first, a pupil area is extracted from iris image information (221). Here, the extracted pupil area is composed of pixels P (i) which are binarized and preprocessed, where P (i) is a P containing position coordinates (X (i), Y (i)) and pixel point brightness density 0. (i) = (X (i), Y (i), 0)

)을 구한다(222). 보다 구체적으로, 동공영역의 좌표평균점(

)은 위치좌표가 다음 수학식 3에서 보는 바와 같다.Next, the coordinate mean point of the extracted pupil area (

(222). More specifically, the coordinate mean point of the pupil area (

) Is as shown in Equation 3 below.

)을 중심으로 홍채영역의 일정부분이 포함되도록 동심원 영역(DC)을 설정한다(223). 동심원 영역(DC)을 설정함에 있어서, 좌표평균점()을 중심으로 홍채영역의 일정부분이 포함되도록 하는 것은 동공반경(r)의 1배 이상 2배 이하, 바람직하게는 1.2배 내지 1.5배에 해당하는 영역을 설정하여, 속눈썹이나 눈꺼풀에 의해서 방해받지 않도록 한다. 여기에서, 동공반경(r)은 다음 수학식 4에서 보는 바와 같이 구하여진다.Next, the coordinate mean point (

The concentric circle (DC) is set to include a certain portion of the iris area around the center (223). In setting the concentric circles DC, the coordinate mean point ( In order to include a certain portion of the iris area around the center, the area corresponding to 1 or more and 2 times or less, preferably 1.2 to 1.5 times the pupil radius (r) is not disturbed by eyelashes or eyelids. Do not Here, the pupil radius r is obtained as shown in Equation 4 below.

Next, in operation 224, the concentric circle DC excluding the pupil, which is an area excluding the pupil area, is set in the concentric circle DC.

Next, an appropriate bending point is found through a histogram smoothed to a 5-unit average value of pixel point brightness concentrations in the concentric region DC ′ excluding the pupil (225).

Next, the concentration of the pixel points is binarized based on the pixel concentration of the bending point (226). More specifically, the density of pixel points that are less than or equal to the pixel concentration of the bending point is binarized to 1 (black point), and the concentration of pixel points whose pixel point brightness concentration is equal to or greater than the pixel concentration of the bending point is binarized to 0 (white point).

)을 구한다(227). Next, the coordinate average point (P) of the pixel point of the concentric circle DC 'excluding the pupil (

(227).

)과 동공 제외된 동심원 영역의 좌표평균점(

)을 연결하는 제 1 직선(

) 및 좌표평균점(

)과 좌표평균점(

)의 중점(

)을 지나고 상기 제 1 직선(

)에 수직하는 제 2 직선(

)을 구한다(228).Next, the coordinate mean point of the pupil area (

) And the coordinate mean point of the concentric

First straight line connecting the

) And coordinate mean points (

) And the coordinate mean point (

) Emphasis

Past the first straight line (

2nd straight line perpendicular to)

(228).

)과 제 2 직선(

)을 통하여 동공 제외된 동심원 영역(DC')을 4개의 영역(DC'1, DC'2, DC'3, DC'4)으로 분할하고, 각각의 분할된 영역에 대한 좌표평균점을 구하고, 좌표평균점중 렉시코그래픽 순서가 가장 작은 점을 포함하는 영역을 데이터 추출 영역(DF'1)으로 설정한다(229). 여기에서, 각각의 분할된 영역에 대한 좌표평균점을 구하는 것은 상기 수학식 3을 참조로 한다.Next, the first straight line obtained in the step 228 (

) And the second straight line (

The concentric circle DC 'excluding the pupil is divided into four regions DC'1, DC'2, DC'3, and DC'4, and the coordinate average point for each divided region is obtained. An area including the point with the smallest lexicographic order among the average points is set as the data extraction area DF'1 (229). Here, to obtain the coordinate average point for each divided region, refer to Equation 3 above.

The topological data for the security method using the iris recognition is extracted from the data extraction region DF'1 configured as described above.

)과 분할된 영역의 좌표평균점을 지나는 직선을 통하여 2등분하고, 2등분된 영역의 각각에 대한 좌표평균점중 렉시코그래픽 순서가 가장 작은 점을 포함하는 영역을 데이터 추출 영역(DF'2)로 다시 설정할 수 있다. 이와 같은 동작을 반복적으로 수행함으로써 데이터 추출 영역(DF'n)의 범위를 좁혀나갈 수 있게 된다. 홍채인식을 이용한 보안 장치별로 해당하는 소정 범위까지 분할하여 그 중 해당하는 데이터 추출 영역을 설정할 수 있게 된다. 즉, 보안 장치별로 속도 및 처리의 정확성을 고려하여 필요한 범위까지 분할하여 데이터 추출 영역을 설정하게 된다.In FIG. 2B, the lexical order of the coordinate mean points of the divided regions is divided by dividing the concentric circle DC 'excluding the pupil into four regions DC'1, DC'2, DC'3, and DC'4. Has shown that the region containing the smallest dot is set as the data extraction region DF'1. Further, the data extraction area DF'1 is referred to as the midpoint (

) And the region containing the smallest lexicographic order among the coordinate average points for each of the divided areas as the data extraction area (DF'2) through a straight line passing through the coordinate mean points of the divided areas. You can reset it. By repeatedly performing such an operation, the range of the data extraction area DF'n can be narrowed. It is possible to set a corresponding data extraction area by dividing a predetermined range corresponding to each security device using iris recognition. That is, the data extraction area is set by dividing the security device into necessary ranges in consideration of speed and processing accuracy.

In addition, in FIG. 2B, the concentric circle DC 'excluding the pupil is divided into four regions DC'1, DC'2, DC'3, and DC'4. Although the region including the points having the smallest order is set as the data extraction region DF'1, the pupil-excluded concentric region DC 'set in step 223 may be set as the data extraction region. More specifically, it is shown in FIG. 10 that the pupil-excluded concentric region DC 'is a data extraction region, and the corresponding region is divided into the data extraction region by dividing the pupil-excluded concentric region DC'. It is shown in FIG.

FIG. 10 illustrates an embodiment of setting a data extraction region through the process of FIG. 2B according to the present invention. Referring to FIG. 10, reference numeral 1001 denotes an eyebrow line as an image of a human eye. Therefore, the upper portion of the line indicated by the reference number 1001 is a portion that is not photographed in the actual image. Reference numeral 1002 denotes a pupillary line, reference numeral 1003 denotes an autonomic nerve line, reference numerals 1004, 1005, 1006, 1007, and 1008 denote hot air lines present in the iris, and reference numeral 1009 denotes a pupil of the pupil. An edge is shown, and reference numeral 1010 shows an area set as a data extraction area. Reference numeral 1011 denotes a concentric circle area DC. In FIG. 10, it can be seen that the pupil-excluded concentric region DC ′ in which the pupil portion is excluded from the concentric region DC is set as the data extraction region.

FIG. 11 shows another embodiment of setting a data extraction region through the process of FIG. 2B according to the present invention. Referring to FIG. 11, an image of a human eye is denoted by reference numeral 1101 to denote an eyebrow line. Therefore, the upper portion of the line indicated by the reference number 1101 is a portion that is not photographed in the actual image. Reference numeral 1102 denotes a pupillary line, reference numeral 1103 denotes an autonomic nerve line, reference numerals 1104, 1105, 1106, 1107, and 1108 denote hot air lines present in the iris, and reference numeral 1109 denotes a pupil of the pupil. An edge is shown and reference numeral 1110 denotes an area set as a data area. Reference numeral 1111 denotes a concentric circle area DC. In FIG. 11, it is shown that the concentric circle DC 'excluding the pupil is divided, and an area including a point having the smallest lexicographic order among the coordinate average points for each of the divided concentric circle areas is set as the data extraction area. Can be.

FIG. 3 illustrates a process of generating topological analysis information from the data extraction region set in FIG. 2. Referring to FIG. 3, first, pixel points constituting a data extraction region set from input iris image information are decomposed (301).

Next, in operation 302, topological feature information about pixel points obtained by decomposition is obtained. Here, the topological feature information includes number information on the decomposed pixel point, pixel point edge information on an edge surrounding the decomposed pixel point, and pixel point vertex on a vertex appearing on the decomposed pixel point edge. It's about information.

More specifically, the case in which n pixel points exist in the set data extraction region, the number information for the pixel points is determined as P (1), P (2), ..., P (n). The pixel point boundary information sets each boundary line to B (1), B (2), ..., B (4n), and the pixel point vertex information sets each vertex to V (1), V ( 2), ..., V (4n).

In this case, the boundary lines B (1), B (2), ..., B (4n) and the vertices V (1), V (2), ..., V (4n) have respective overlapping times. do. In other words, if a specific boundary B (i) appears only once in B (1), B (2), ..., B (4n), the degree of redundancy is 0, and the degree of redundancy is 1. In the same way, if vertex V (i) shows only one time among V (1), V (2), ..., V (4n), the degree of redundancy will appear as 0 and the degree of redundancy will appear once and the third time. Is 2 and 4 times, the redundancy is 3.

Here, the topological model function for generating the topological feature information is obtained by multiplying the first function A (D), which is the total number of pixel points, and the result of the first function 4 by the number of pixel points. A second function P (D) is obtained by subtracting a sum of the redundancies of the boundary lines obtained through the pixel point boundary information and multiplying by two, and the overlapping of the boundary lines obtained through the pixel point boundary information to the result of the first function. After the sum of the figures is added, the third function E (D) subtracts the sum of the degree of redundancy of the vertices obtained through the pixel vertex information.

That is, the first function A (D), the second function P (D), and the third function E (D) are as shown in Equations 5, 6, and 7, respectively.

 A (D) = total number of pixel points at D

P(D) = 4 * A(D) - 2 *

P (D) = 4 * A (D)-2 *

Where b (i) = redundancy of B (i)

E(D) = - + A(D)

E (D) =-+ A (D)

Where b (i) = B (i) redundancy, v (i) = V (i) redundancy

More specifically, for example, when the result values of the first function, the second function, and the third function are obtained with respect to the pixel points having three horizontal and three vertical pixels, they are as follows. The result value of the first function is 9, which is the total number of pixel points according to Equation 5. According to Equation 6, the second function result value is 12 minus the value obtained by multiplying 2, which is the degree of redundancy of B (i), by multiplying 9, which is the first function result value, by 4. The result of the third function is 1 by subtracting 20, which is the degree of redundancy of V (i), and 12, which is the degree of redundancy of B (i), plus 9, the result of the first function.

More specifically, for example, the result of the first function, the second function, and the third function is obtained as to the case where the pixel point among the pixel points to which three horizontal and three vertical pixels are connected is empty. The result of the first function is 8, which is the total number of pixel points according to Equation 5. According to Equation 6, the second function result value is 16 minus the value obtained by multiplying 8, which is the degree of redundancy of B (i), by 2, by multiplying 8, which is the first function result, by 4. The third function result value is 0 by subtracting 16, which is the degree of redundancy of V (i), from 8, which is the degree of redundancy of B (i), and adding 8, the result of the first function.

Next, the topological feature information obtained in step 302 is input to a user authentication function for generating user authentication information to generate user authentication information used as comparison data in the iris recognition system (303).

Here, before generating the user authentication information, a fourth function result value is obtained by inputting the fourth function S (D) through the first function result value and the second function result value, and the obtained fourth function result value. A preprocessing process for setting the topological analysis information for determining whether or not it is within this predetermined range may be further performed. Here, the fourth function S (D) is a function of multiplying the result of the first function by multiplying it by 16 after multiplying the result of the second function.

That is, the fourth function S (D) is shown in Equation 8.

S (D) = (P (D) * P (D)) / (16 * A (D))

In a more specific example as described above, the fourth function result is 1 divided by the square of 12, which is the second function result, and multiplied by 9, which is the first function result.

More specifically, in another example as described above, the fourth function result value is 2 obtained by multiplying the second function result value 16 by multiplying the first function result value 8 by 16.

In this manner, first, it is determined whether the comparison copy falls within a predetermined range (for example, within an error range of 40%) of the original through the fourth function S (D), and when it is determined to fall within the predetermined range. A process of generating user authentication information is performed. On the other hand, if it is determined that it does not fall within a certain range is terminated without performing the process of generating the user authentication information.

When it is determined to fall within a predetermined range in the preprocessing process for generating user authentication information as shown in Equation 8, the first function result value, the second function result value and the third function result value are input to a predetermined user authentication function. Set the result value as user authentication information for iris recognition.

The predetermined user authentication function is as shown in equation (9). In Equation 9, an arbitrary constant value x serves to give a weight.

Where x is any constant value

In Equation 9, a result value generated by inputting a result value of the first function A (D), the second function P (D), and the third function E (D) into Equation 9 is used. It will be used as authentication information. In Equation 9, x is preferably 0, 1 and 2, that is, F (0), F (1) and F (2) as a predetermined user authentication function.

4 illustrates a preferred embodiment of a data structure storing original data corresponding to reference luminance values for each original registrant according to the present invention. Referring to Figure 4, according to the present invention shows a preferred embodiment of a data structure for storing the original data corresponding to the reference luminance value for each original registrant. The original data includes the topological analysis information obtained from the data extraction region set in FIGS. 2A and 2B, but only a part of the topological analysis information may be selected and used as necessary. In other words, it can be appropriately selected in consideration of the accuracy or processing speed of authentication.

5 shows a security method using iris recognition in accordance with the present invention. Referring to FIG. 5, first, original data corresponding to a reference luminance value having predetermined values is generated and stored (501).

Next, comparative data of topological geometric information corresponding to the iris image information of the person to be identified is generated (502).

Next, the identity is verified by comparing the original data with the comparative data and confirming whether the person whose identity is confirmed is a user who is already registered in the security device using iris recognition (503).

In this process, the iris image of the person to be identified is obtained twice at different luminosities, and the number of pupil pixels of the iris image information obtained above is compared, and if there is no difference between the two iris images, the person to be identified is regarded as an inanimate object. The process of failing the above verification can be further included.

6 illustrates a process of verifying an identity based on original data and comparative data according to the present invention. Referring to FIG. 6, first, it is determined whether there is any reference luminance value of the original data having the same value as that of the comparative data.

)을 선정한다(602). 그리고 위 기준값들 중에서 비교본 데이터의 광도값보다 작고 가장 근접되는 기준값(

)을 설정한다(603). Next, when there is no identical value as a result of the determination in step 601, the reference value that is larger than the luminance value of the comparative data and is closest to the reference value (

(602). Among the above reference values, the reference value that is smaller than the luminance value of the comparative data and is closest

(603).

)및 기준값(

)에 대응되는 원본 데이터들과 위 비교본 데이터 간의 차이값들을 구한다(604).Next, the reference value of the predetermined registrant among the original registrants who registered the original data (

) And reference value (

Difference values between the original data corresponding to the reference data) and the comparison data are obtained (604).

)및 기준값(

)에 대응되는 원본 데이터간의 차이보다 큰지 여부를 판단한다(605). 상기 605 단계의 판단결과 참인 경우에는 모든 원본 등록자에 대하여 위 참조번호 604 단계 및 605 단계를 수행하였는지 여부를 판단한다(606). 상기 참조번호 606 단계의 판단결과 모든 원본 등록자들에 대하여 상기 참조번호 604 단계 및 605 단계를 수행한 것이 아니면 상기 참조번호 604 단계로 이동한다. 한편, 상기 참조번호 606 단계의 판단결과 모든 원본 등록자에 대하여 위 참조번호 604 단계 및 605 단계를 수행한 것이면 위 비교본 데이터에 대응되는 신원확인 신청자가 원본 등록자 중에 포함되지 않은 것으로 판단(즉, 인증이 실패된 것으로 처리)한다(607).Next, the difference values are based on the registrant's reference value (

) And reference value (

In step 605, it is determined whether the difference is greater than the difference between the original data corresponding to the " If the determination result of step 605 is true, it is determined whether the reference numerals 604 and 605 are performed for all original registrants (606). If it is determined in step 606 that all the original registrants have not performed steps 604 and 605, the process moves to step 604. On the other hand, if the determination result of step 606 refers to the steps 604 and 605 of all the original registrants, it is determined that the applicant for identity verification corresponding to the comparative data is not included among the original registrants (that is, authentication Process as failed (607).

)및 기준값(

)에 대응되는 원본 데이터간의 차이보다 작은 경우에는 위 비교본에 대응되는 신원 확인 신청자가 위 원본 데이터들에 대응되는 원본 등록자와 동일인인 것으로 판단(즉, 인증이 성공한 것으로 처리)한다(608).On the other hand, as a result of the determination in step 605 the difference values are the reference value of the registrant (

) And reference value (

If less than the difference between the original data corresponding to the), the identity verification applicant corresponding to the comparison is determined to be the same as the original registrant corresponding to the original data (ie, the authentication is successful) (608).

On the other hand, if there is a reference luminance value of the original data that is the same as the luminance value of the comparison copy as a result of the determination in step 601, whether the original data corresponding to the reference luminance value matches the above comparison data. Decisions are made (609, 610). If the result of the determination at step 609 is the same, the process moves to reference numeral 608.

Figure 7a is a block diagram of a preferred embodiment of a security device using iris recognition in accordance with the present invention. Referring to FIG. 7A, first, the input unit 708 determines a predetermined operation command from a user, distinguishing information on whether the iris image information captured by the photographing unit 701 is original data or comparative data, and the iris image. When the information is original data, identification information (for example, personal identification information such as a name, a social security number, an address, etc.) for identifying the original registrant corresponding thereto is received. Here, the predetermined operation command may be input through a predetermined operation button, a predetermined sensor, or the like.

The photographing unit 701 is operated by receiving the operation command from the input unit 708, and generates iris image information by photographing the iris of the original registrant corresponding to the original data or the identification applicant corresponding to the comparative copy data. In addition, by measuring the luminous intensity at the time of the above shooting to generate luminous intensity value information. In addition, when the distinguishing information is received from the input unit 708, and the distinguishing information is the above iris image information, the luminance value corresponding to the above iris image information and the above iris image information is applied to the iris image information. The second iris image information generated by photographing again in an environment having a brightness value different from that of the second iris image is generated.

The data area setting unit 702 receives the upper iris image information and a luminance value corresponding thereto from the photographing unit 701. In addition, when analyzing the distinguishing information provided from the input unit 708, when the iris image information is recognized as the comparative data, the second iris image information is additionally provided from the photographing unit 701, and the input unit ( 708) the above identification information is provided. A portion of the iris image of the iris image information and the second iris image information is set as the data region. The data region setting unit 702 will be described in more detail with reference to FIG. 7B.

The iris information generating unit 703 may set the second iris when the distinguishing information, the brightness value, the iris image information and the corresponding data region setting information, and the distinguishing information are set as the comparative data from the data region setting unit 702. Image information, data area setting information, and identification information corresponding thereto are provided. And based on the iris image information and the data setting information, the characteristic information of the iris image is generated. The feature information of the iris image includes topological analysis data obtained from the data area. When the distinguishing information is set as the comparative copy data, feature information corresponding to the second iris image information and the data setting information is generated.

In addition, if the iris image information is the original data by analyzing the distinguishing information, the identification information, the brightness value, and the iris image information are transmitted to the original data management unit 704, and when the iris image information is the comparative data, The luminance value, the iris characteristic information, and the second iris characteristic information are transmitted to the living body identification unit 705.

The original data management unit 704 receives the identification information, the brightness value, and the iris image characteristic information from the iris information generation unit 703 and stores it in the original data storage unit 709 which is a database.

The living body identification unit 705 receives the brightness value, the iris characteristic information, and the second iris characteristic information from the iris information generation unit 703, and compares the iris characteristic information with the second iris characteristic information when both are the same. In the identification of the identification applicant corresponding to the above comparative data as an inanimate object, the authentication failure information is generated as authentication result information and provided to the user through the output unit 707. On the other hand, if the comparison is not the same as the above results, the luminance value, the iris characteristic information is provided to the authenticator 706.

The authenticator 706 receives the luminance value and the iris image information from the living organism identification unit 705 and searches the original data storage unit 709 for a reference luminance value corresponding to the luminance value, and searches for the reference luminance value. Search for the original data corresponding to the identification applicant corresponding to the topological geometric analysis information (comparative) among the original data corresponding to the iris characteristic information according to the original brightness of the original registrant. If the search result is successful, the identity verification applicant is reported to the searched original registrant, the authentication success result information is generated, and if the search fails, the authentication failure result information is generated and provided to the output unit 707.

Here, the method of searching for the original registrant who can be seen as the same person as the identity verification applicant, if there is a reference luminance value equal to the above luminance value, searches for and matches the topological geometric analysis information among the original data corresponding to the reference luminance value. The original registrant corresponding to the original data shall be deemed to be the same person as the applicant. If there is no reference luminance value that is the same as the above luminance value, the two reference luminance values that are larger and smaller than the above luminance values are obtained, and the original data of any same original registrant corresponding to the two reference luminance values is searched. Compares the difference between the original data and the difference between the original data and the topological analysis information. Does the difference between the original data and the topological analysis information be smaller than the difference between the original data? The data storage unit 709 is searched. If the search is successful, the identification applicant corresponding to the topological analysis information and the original registrant corresponding to the original data are regarded as the same person, and the authentication success result information is generated. Generate.

The output unit 707 receives the authentication result information from the living body confirmation unit 705 and the authentication unit 706, and outputs the result to the user.

Meanwhile, in the present invention, the secret code is received from the users through the input unit 708, and the secret code is registered together with the original data in the original data storage unit 709, and then received from an identity verification target person. The method may further include determining whether the secret code is the same as any one of the registered secret codes. In this case, through the topographic code analysis information stored in the original data storage unit 709 and the photographing unit 701 corresponding to the identity verification subject identified through the secret code, Only when the topological analysis information obtained from the iris image information obtained by photographing the identification subject is consistent, the authentication of the identification subject can be treated as successful.

FIG. 7B is a block diagram illustrating the data area setting unit of FIG. 7A in more detail. Referring to FIG. 7B, an area separator 710 may be configured to separate the pupil area and the iris area from the image information of the eye; A first coordinate average point calculator 711 for obtaining a coordinate average point of the pupil area with respect to pixel points of the pupil area separated by the area separator 710; A concentric circle setting unit (712) for setting a concentric circle area so that a predetermined portion of the iris area is included around the coordinate average point of the pupil area; A second coordinate average point calculator 713 for forming a concentric circle in which the pupil is excluded from the concentric circle through the region separation unit 710, and obtaining a coordinate average point for pixel points of the excluded concentric circle region; ); A concentric circle segmentation unit 714 for dividing the concentric circle region according to a coordinate mean point of the pupil region and a coordinate average point of the excluded concentric circle region; And a comparison area setting unit 715 for setting one of the divided concentric circle areas as a comparison area for iris recognition. Parts not described in FIG. 7B will be described with reference to FIGS. 1 to 6.

FIG. 7C is a block diagram illustrating in detail the iris information generator of FIG. 7A. Referring to FIG. 7C, a comparison object extracting unit 720 extracting pixel points to be compared from pixel points included in the data area setting unit 702; A topological feature information extracting unit 721 for extracting topological feature information on pixel point number information, corner number information, and vertex number information of the pixel points extracted from the comparison area; A user authentication information generation unit 722 for generating user authentication information by calculating the extracted topological feature information; And a storage unit 723 for storing a topological model function for extracting the topological feature information and a user authentication function for generating user authentication information from the topological feature information. Here, the topological model function stored in the storage unit 723 is a first function that adds all the number of pixel points through the pixel point number information, and does not contact other pixel points through the pixel point boundary information. A second function that adds all the number of corners, the first function result value, the second function result value, and the pixel point vertex information, wherein the feature information on the number of empty pixel points in a region formed by contacting a plurality of pixel points And the user authentication function is a function for generating user authentication information using the first function result value, the second function result value, and the third function result value as input values. Although the iris information generation unit 703 in FIG. 7C does not include the comparison area setting unit (refer to FIG. 7B), the comparison setting of the comparison area for extracting the user authentication information from the eye image information is performed. The area setting unit may be included together. Parts not described in FIG. 7C will be described with reference to FIGS. 1 to 6.

The invention can also be embodied as computer readable code on a computer readable recording medium. The computer-readable recording medium includes all kinds of recording devices in which data that can be read by a computer system is stored. Examples of computer-readable recording media include ROM, RAM, CD_ROM, magnetic tape, floppy disks, and optical data storage, and may also include those implemented in the form of carrier waves (eg, transmission over the Internet). . The computer readable recording medium can also be distributed over network coupled computer systems so that the computer readable code is stored and executed in a distributed fashion.

As described above, optimal embodiments have been disclosed in the drawings and the specification. Although specific terms have been used herein, they are used only for the purpose of describing the present invention and are not intended to limit the scope of the present invention as defined in the claims or the claims. Therefore, those skilled in the art will understand that various modifications and equivalent other embodiments are possible from this. Therefore, the true technical protection scope of the present invention will be defined by the technical spirit of the appended claims.

According to the present invention, the data extraction area is set based on the coordinates of the pupil area and the iris area and the pixel point brightness concentration, and the iris is recognized by analyzing the iris image information included therein, so as to recognize the iris. There is an effect that the recognition rate is hardly affected.

In addition, since the original data corresponding to the various reference luminance values are provided and the identity is confirmed using the topological analysis information corresponding to the luminance values of the comparative image, the recognition rate can be significantly lowered even when there is a change in luminance during the shooting. There is an effect that can be installed and used in various places where the brightness changes.

In addition, since the image of the iris can be analyzed as a topological feature and used as characteristic information of the iris, there is an effect of reducing the false recognition rate even when there is a change in the size of the iris image according to the shooting distance.

Claims (19)

(a) obtaining coordinate mean points of pixel points of the pupil area in the image information of the eye; (b) setting the concentric circles such that a predetermined portion of the iris region is included around the coordinate average point of the pupil region; (c) obtaining coordinate mean points of pixel points of the excluded concentric region formed by excluding the pupil region in the concentric region; (d) dividing the concentric circles according to the coordinate average points of the pupil regions and the coordinate average points of the excluded concentric circles; And (e) setting one of the divided concentric circle areas as a comparison area for iris recognition, comprising: a comparison area setting method for iris recognition. The method of claim 1, wherein step (e) Obtaining a coordinate mean point for each of the divided concentric circles; And And setting a region including a point having the smallest lexographical order among the coordinate average points of the divided concentric circles as the comparison region. (Here, the lexographic order is that two coordinate points on the plane

If,

However, if

Back side

If The method of claim 1, wherein step (d) Obtaining a first straight line connecting the coordinate mean point of the pupil area and the coordinate mean point of the excluded concentric circle area; Obtaining a second straight line perpendicular to the first straight line and passing through a midpoint between a coordinate average point of the pupil area and a coordinate average point of the excluded concentric circle area; And And dividing the concentric region through the first straight line and the second straight line. The method of claim 3, wherein And dividing the concentric region divided by the first straight line and the second straight line repeatedly within the range required for iris recognition. The method of claim 1, wherein after step (e) Obtaining a coordinate mean point of the divided concentric circles set as the comparison regions; Obtaining a coordinate average point of the divided concentric circle set as the comparison area and a third straight line passing through the midpoint; Dividing the divided concentric circles again through the third straight line; And And setting one of the concentric circles divided through the third straight line as a comparison area for iris recognition. The method of claim 1, wherein step (b) Obtaining a radius of the pupil area; And And setting a concentric circle area having a radius of a predetermined multiple larger than a radius of the pupil area as a center point of the coordinate average point of the pupil area. 2. The method of claim 1, wherein step (c) Setting a pupil excluded concentric region formed by excluding the pupil region from the concentric region; Obtaining a reference value for the pixel point in the concentric excluded region; Converting pixel points in the concentric excluded region into binarized pixel points according to the reference value; And And obtaining a coordinate mean point of the concentric circle region excluding the pupil from the coordinates of the binarized pixel points in the pixel point converting step. The method of claim 7, wherein And the reference value is obtained by comparing the pixel point brightness concentration with respect to the bending point of the histogram with respect to the pixel point brightness concentration in the concentric circle excluding the pupil. (a) setting a comparison area to extract user authentication information from eye image information; (b) extracting pixel points to be compared from pixel points included in the comparison area; (c) extracting topological feature information of pixel point number information, corner number information, and vertex number information of the pixel points to be compared; And (d) generating the user authentication information used as comparative authentication data in the iris recognition system by calculating the extracted topological feature information; and generating user authentication information for iris recognition. The method of claim 9, wherein in the step (c) the topological feature information, Obtaining a first function result value of adding all the pixel point numbers through the pixel point number information; Obtaining a second function result value which adds both the first function result value and the number of corners not contacting other pixel points through the pixel point boundary information; And Obtaining a third function result value which is feature information on the number of empty pixel points in an area formed by contacting a plurality of pixel points through the first function result value, the second function result value, and the pixel point vertex information; Method for generating user authentication information for iris recognition, characterized in that consisting of. The method of claim 10, The second function is a function of multiplying the sum of the redundancy of the boundary lines obtained through the pixel point boundary line information by multiplying the result value of the first function by 4 and multiplying by 2, The third function is a function of subtracting the total sum of redundancy of vertices obtained through the pixel vertex information after adding the sum of redundancy of the boundary lines obtained through the pixel point boundary information to the result of the first function. User authentication information generation method for iris recognition. The method of claim 10, In step (d), the first function result value, the second function result value, and the user authentication function may be assigned to each of the first function result value, the second function result value, and the third function result value. And generating a user authentication information by inputting a third function result value. The method of claim 12, Further comprising; a pre-processing step for performing the user authentication before generating the user authentication information in step (d), The preprocessing step may include: obtaining a fourth function result value obtained by dividing the user squared second function result value by the product of the first function result value multiplied by a predetermined constant; and presetting the fourth function result value. Determining whether the fourth function result value is not within a preset range, and generating user authentication information for iris recognition, wherein the user authentication information for iris recognition is terminated. How to produce. The method of claim 9, wherein step (a) Obtaining coordinate mean points of pixel points of the pupil area in the image information of the eye; Setting a concentric circle area to include a predetermined portion of the iris area around the coordinate average point of the pupil area; Obtaining a coordinate mean point of pixel points of the excluded pupil concentric region formed by excluding the pupil region from the concentric region; Dividing the concentric circles according to a coordinate average point of the pupil area and a coordinate average point of the excluded concentric circle areas; And And setting one of the divided concentric circles as a comparison area for iris recognition. The method of claim 9, wherein step (b) Obtaining a bending point of the histogram with respect to the pixel point brightness concentration in the comparison area; And And converting pixel points in the comparison area into binarized pixel points based on pixel point brightness concentrations of the bend points of the histogram. An area separator for separating the pupil area and the iris area from the image information of the eye; A first coordinate average point calculator for obtaining a coordinate average point of the pupil area with respect to pixel points of the pupil area separated by the area separator; A concentric circle setting unit configured to set a concentric circle area such that a predetermined portion of the iris area is included around the coordinate average point of the pupil area; A second coordinate average point calculator configured to form a concentric region excluding the pupil region from the concentric region through the region separation unit, and to obtain coordinate average points of pixel points of the excluding the concentric circle region; A concentric circle division unit dividing the concentric circle area according to a coordinate average point of the pupil area and a coordinate average point of the excluded concentric circle area; And And a comparison area setting unit configured to set one of the divided concentric circle areas as a comparison area for iris recognition. The method of claim 16, wherein the concentric region divider Obtain a first straight line connecting the coordinate average point of the pupil area and the coordinate average point of the excluded concentric circle, the center of the coordinate average point of the pupil region and the coordinate average point of the excluded pupil region orthogonal to the first straight line Comprising a second straight line passing through the comparison region setting apparatus for iris recognition, characterized in that to divide the concentric region through the first straight line and the second straight line. A comparison area setting unit for setting a comparison area to extract user authentication information from eye image information; A comparison object extracting unit extracting pixel points to be compared from pixel points included in the comparison area; A topological feature information extraction unit for extracting topological feature information about pixel point number information, corner number information, and vertex number information of the pixel points extracted from the comparison area; And And user authentication information generation unit for generating user authentication information by calculating the extracted topological feature information. The method of claim 18, Further comprising a storage unit for storing a topological model function for extracting the topological feature information and a user authentication function for generating user authentication information from the topological feature information, The topological model function is a first function that adds all the number of pixel points through the pixel point number information, and a second function that adds all the number of corners that do not contact other pixel points through the pixel point boundary information. A third function that is characteristic information about the number of empty pixel points in a region formed by contacting a plurality of pixel points through a function result value, a second function result value, and the pixel point vertex information, and the first function result value, 2. The apparatus for generating user authentication information for iris recognition, comprising: a user authentication function for generating user authentication information using the function result value and the third function result value as input values.

Images