KR102508789B1 - Method of authorizing identity using an iris and system thereof - Google Patents

Abstract

A method for authenticating a user includes obtaining an encoded pattern using an authentication device, wherein the encoded pattern includes a structure representing data corresponding to a first iris code representing a first region that is part of an iris image of a first user. , generating a first iris code from an encoded pattern, acquiring an iris image of a second user, receiving information about a positional relationship between a first region and an entire iris image of the first user, Generating a second iris code from an iris image of a user, wherein a second region corresponding to the first region is extracted from the iris image of the second user by using positional relationship information, and the second region is extracted from the second region. 2 Generating iris codes, comparing the first iris code with the second iris code, and authenticating whether the first user and the second user are the same person based on the comparison result.

Description

Iris authentication method and iris authentication system {METHOD OF AUTHORIZING IDENTITY USING AN IRIS AND SYSTEM THEREOF}

The present disclosure relates to an iris authentication method and system for acquiring an iris image and comparing it with pre-registered information to determine whether or not it is the same person. An iris authentication method and system for extracting feature information and generating and using a code corresponding thereto.

In general, in a method of recognizing a specific individual using the iris, iris characteristic information is coded and stored from an iris image in an image signal obtained by photographing a person's eye, and by comparing pre-stored codes in a database, the result of the comparison to authenticate the identity of a specific individual. An iris authentication method is disclosed in United States Patent No. 5,291,560 to John G. Daugman.

In the Dogman patent, after finding a circle that best approximates the boundary line between the iris and the pupil and approximating the boundary between the iris and the sclera as a circle in the same principle, the area between them is set as the iris area and this area is proportional to the principle An annular selection area partitioned by a plurality of concentric circles is set by , and iris feature information extracted from each selection area through Gabor transformation is used as an iris code to build a database. During authentication, iris feature information extracted from the human eye is compared with registered iris feature information for each selected area to authenticate the identity of a specific individual.

In US Patent No. 6,247,813 registered by the present applicant, the iris code is extracted from the fibrous tissue of the iris obtained from the iris image, the reaction of the pupil and the autonomic ring by light, the shape of the autonomic ring, and the location and shape of the tear A technique for distinguishing a specific individual using this has been disclosed.

Registered U.S. Patent No. 8,009,876 of the present applicant discloses a technique for increasing iris recognition by dividing an iris image into a plurality of regions, selectively merging each unit divided region, and generating an iris code from the region.

U.S. Patent Registration No. 5,291,560 U.S. Patent Registration No. 6,247,813 U.S. Patent Registration No. 8,009,876

According to the present disclosure, accuracy of iris recognition may be improved by dividing an iris region into a plurality of regions and extracting an iris code from each region.

According to one aspect of the present disclosure, in a method for generating an iris code from an image of an iris, the step of dividing at least a part of a region between an inner boundary and an outer boundary of the iris into a plurality of divided regions, the plurality of divisions. the region includes a first segmented region that includes at least a portion of the autonomic ring of the iris and a second segmented region that does not include the autonomic ring of the iris; extracting a plurality of first features by applying a first transform to the first divided region and extracting at least one second feature by applying a second transform to the second divided region; and generating an iris code based on the plurality of first characteristics and the at least one second characteristic.

According to an embodiment, the first transform and the second transform may include a wavelet transform, and a decomposition level of the first transform may have a higher order than a decomposition level of the second transform. there is.

According to an embodiment, the method may further include dividing at least one divided area among the first divided area and the second divided area into a plurality of unit divided areas.

According to one embodiment, the plurality of divided regions further comprises a third divided region not including the autonomic ring, wherein the first divided region is adjacent to the second divided region, but the second divided region is the first divided region. After the step of dividing at least a part of the region between the first divided region and the third divided region and between the inner boundary and the outer boundary of the iris into a plurality of divided regions, a third transformation is applied to the third divided region, The method may further include extracting at least one third feature, and a decomposition level of the first transform may have a higher order than a decomposition level of the third transform.

According to an embodiment, the decomposition level of the second transform may have a higher order than the decomposition level of the third transform.

According to an embodiment, prior to dividing at least a portion of the region between the inner boundary and the outer boundary of the iris into a plurality of divided regions, at least a portion of the region between the inner boundary and the outer boundary of the iris is converted into a rectangular region. A mapping step may be further included.

According to an embodiment, after the step of mapping the plurality of divided regions into rectangular regions, further dividing at least one of the first divided region and the second divided region into a plurality of unit divided regions. can include more.

According to an embodiment, the further dividing into a plurality of unit divided regions may include dividing a part of the unit divided region that is likely to be covered by the upper eyebrow into a first unit divided region, and dividing the unit divided region into a lower eyebrow. Dividing a portion likely to be occluded into a second unit divided region, and dividing a portion unlikely to be occluded into a third unit divided region, wherein the size of the first unit divided region < the second unit The size of the partitioned area < the size of the third unit partitioned area -.

According to an embodiment, the first transform and the second transform include a Haar transform, and the plurality of first features and the at least one second feature include a low frequency component. can do.

According to one embodiment, the step of dividing at least a portion of the region between the inner boundary and the outer boundary of the iris into a plurality of divided regions may further include determining a region where the autonomic ring is located.

According to one embodiment, the step of determining the region where the autonomic ring is located may determine the region where the autonomic ring is located based on the distance from the pupil to the inner boundary and the outer boundary of the iris.

According to another aspect of the present disclosure, an iris authentication system includes an acquisition module configured to acquire an iris image; An iris region is divided from an iris image, the iris region is divided into a plurality of divided regions including a first divided region and a second divided region, and a first transformation having a first decomposition level is applied to the first divided region. extracts a plurality of first features, extracts at least one second feature by applying a second transform having a second decomposition level smaller than the first decomposition level to the second split region, and an encoding module for generating an iris code based on the first feature and the at least one second feature; and a matching module receiving and comparing the iris code with the iris code from the database, and performing iris authentication based on the comparison result.

According to one embodiment, the first segmented area may include an autonomic ring and the second segmented area may not include an autonomic ring.

According to an embodiment, the first transform and the second transform include a Haar transform, and the plurality of first features and the at least one second feature include a low frequency component. can do.

According to another aspect of the present disclosure, in a method for generating an iris code from an image of an iris, the step of dividing at least a part of a region between an inner boundary and an outer boundary of the iris into a plurality of divided regions, the plurality of divisions. the region includes a first divided region including a first iris pattern of a first density and a second divided region including a second iris pattern of a second density smaller than the first density; extracting a plurality of first features by applying a first transform to the first divided region and extracting at least one second feature by applying a second transform to the second divided region; and generating an iris code based on the plurality of first characteristics and the at least one second characteristic.

According to an embodiment, the first transform and the second transform may include a wavelet transform, and a decomposition level of the first transform may have a higher order than a decomposition level of the second transform. there is.

According to another aspect of the present disclosure, a method of generating an iris code from an iris image includes converting at least a part of a region between an inner boundary and an outer boundary of the iris into a rectangular region coordinate system; Dividing the region between the inner boundary and the outer boundary of the iris converted into the square region coordinate system into a plurality of divided regions, wherein the plurality of divided regions include a first divided region where the autonomic ring is estimated to be located and an autonomic ring includes a second partition area estimated not to be located; extracting a plurality of first features by applying a first transform to the first split region and extracting at least one second feature by applying a second transform to the second split region; the level has a higher order than the decomposition level of the second transform; and generating an iris code based on the plurality of first characteristics and the at least one second characteristic.

In one embodiment, the method of generating an iris code from an iris image may further include dividing at least one of the first divided region and the second divided region into a plurality of unit divided regions. .

In one embodiment, the first divided region may further include a region where a pupillary sphincter is located, and the second divided region may be located outside the first divided region.

It is possible to improve accuracy in iris recognition by dividing the iris region into a plurality of regions and adding information of several frequency bands to each region.

1 is an example of an image of one eye including the entire iris.
FIG. 2 is an example of an iris image acquired from the image of one eye of FIG. 1 .
3 is an example of dividing an iris image into a plurality of regions.
4 is an example of dividing an iris region into a plurality of regions.
5 is an example of dividing an iris region into a plurality of regions.
6A to 6C are exemplary diagrams of decomposing an image into a plurality of levels.
7 is an example of a binary tree when a signal is decomposed into 3 levels.
8A is an example of converting an iris region into a coordinate system of a rectangular region.
8B is a conceptual diagram obtained by wavelet transforming the coordinate system of the rectangular region of FIG. 8A.
9A is another example of converting an iris region into a coordinate system of a rectangular region.
FIG. 9B is a conceptual diagram obtained by performing wavelet transformation on the rectangular region coordinate system of FIG. 9A.
10 is a conceptual diagram of an iris authentication system according to an embodiment of the present disclosure.
11 is a flowchart of an authentication method according to an embodiment of the present disclosure.

Hereinafter, the present disclosure will be described in detail with reference to the drawings. In describing the present disclosure, if it is determined that a detailed description of a related known function or configuration may unnecessarily obscure the gist of the present disclosure, a detailed description thereof will be omitted. In addition, the following embodiments may be modified in many different forms, and the scope of the technical idea of the present disclosure is not limited to the following embodiments. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the spirit of the disclosure to those skilled in the art.

It should be understood that the techniques described in this disclosure are not intended to be limited to specific embodiments, and include various modifications, equivalents, and/or alternatives of the embodiments of this disclosure.

In connection with the description of the drawings, like reference numerals may be used for like elements.

In the present disclosure, expressions such as “has,” “can have,” “includes,” or “can include” indicate the presence of a corresponding feature (eg, numerical value, function, operation, or component such as a part). , which does not preclude the existence of additional features. In this disclosure, expressions such as “A or B,” “at least one of A and/and B,” or “one or more of A or/and B” may include all possible combinations of the items listed together. . For example, “A or B,” “at least one of A and B,” or “at least one of A or B” (1) includes at least one A, (2) includes at least one B, Or (3) may refer to all cases including at least one A and at least one B.

Expressions such as "first," "second," "first," or "second," used in the present disclosure may modify various elements regardless of order and/or importance, and may refer to one element as It is used only to distinguish it from other components and does not limit the corresponding components.

A component (e.g., a first component) is "(operatively or communicatively) coupled with/to" another component (e.g., a second component); When referred to as "connected to", it should be understood that the certain component may be directly connected to the other component or connected through another component (eg, a third component). On the other hand, when an element (eg, a first element) is referred to as being “directly connected” or “directly connected” to another element (eg, a second element), the element and the above It may be understood that other components (eg, a third component) do not exist between the other components.

The expression “configured to (or configured to)” as used in this disclosure means, depending on the situation, for example, “suitable for,” “having the capacity to.” ," "designed to," "adapted to," "made to," or "capable of." The term "configured (or set) to" may not necessarily mean only "specifically designed to" hardware. Instead, in some contexts, the phrase "device configured to" may mean that the device is "capable of" in conjunction with other devices or components. For example, the phrase “processor configured (or configured) to perform A, B, and C,” “module configured (or configured) to perform A, B, and C” refers to a dedicated processor ( Example: embedded processor), or a general-purpose processor (eg, CPU or application processor) capable of performing corresponding operations by executing one or more software programs stored in a memory device.

U.S. Patent Nos. 5,291,560, 6,247,813 and 8,009,876 are incorporated herein by reference. Parts briefly described in this disclosure may be replaced with those disclosed in US Patent Registration Nos. 5,291,560, 6,247,813 and 8,009,876.

According to an embodiment, authentication using iris recognition detects an iris from an original image of an eye to obtain an iris image, localizes the iris image, and segments the pupil area and the iris area. . The iris region is normalized to a fixed dimension and converted to a rectangular region interpretable in row and column format. This is also called mapping. The iris code is extracted by encoding the rectangular region. US Patent Registration No. 8,009,876 discloses a method of dividing an iris region into a plurality of various regions using a variable multi-sector method and extracting an iris code from each region. Authentication is performed by comparing the extracted iris code with a previously stored template.

Hereinafter, a method of extracting an iris code from an iris region will be described with reference to the drawings.

1 is an example of an image 100 of a user's eye according to an embodiment. 2 is an example in which an image of a user's pupil is enlarged.

Referring to FIGS. 1 and 2 , an eye image 100 includes a pupil 120 and an iris 110 surrounding the pupil 120 . The iris 110 includes an autonomic nerve wreath 130 . In general, the pattern of the iris is formed by the pupillary sphincter muscle (not shown), dilator muscle (not shown), collagenous fiber (not shown), autonomic ring 130, etc. in the iris do. The overlapping region between the pupillary sphincter and dilated pupil is sometimes called the iris ring or the iris collarette, and the autonomic ring 130 and the iris divider ring are sometimes equated. In general, the autonomic ring 130 is located at the ⅓ point on the pupil 120 side, and the pupillary sphincter muscle exists on the inside and the pupillary dilatation muscle on the outside. In general, the pupillary sphincter region has a complex first pattern (radial furrows 134), and the pupillary dilatation muscle region has relatively few second patterns (138). The spokes 132 may be connected to the autonomic nerve ring 130. Patterns present in the iris 110, such as the autonomic ring 130, the spokes 132, the first pattern 134, and the second pattern 138, are also referred to as iris patterns. In FIG. 2 , the first pattern 134 and the second pattern 138 are conceptual, and the actual shapes of the pupillary sphincter (first pattern) and dilated pupil 138 may be different from these.

11 is a flowchart of an authentication method according to an embodiment of the present disclosure.

An iris image is acquired (S1110). The iris image may be obtained by localizing and segmenting an eye image obtained through a charge-coupled device (CCD) sensor or a complementary metal-oxide semiconductor (CMOS) sensor. The acquired iris image is divided into a plurality of regions (S1120). For example, the boundary between the iris and the sclera may be detected, the iris region may be segmented with the pupil as the center, and segmentation may also be performed through radial segmentation. The divided region may include a first divided region including the autonomic ring and a second divided region not including the autonomic ring. The division of the region may use a method described in the present disclosure and references. A plurality of first features are extracted by applying a first transform to the first divided region, and at least one second feature is extracted by applying a second transform to the second divided region (S1130). The first transform and the second transform may include a wavelet transform, and a decomposition level of the first transform may have a higher order than a decomposition level of the second transform. An iris code is generated based on the extracted features (S1140).

Hereinafter, iris segmentation and feature extraction will be described in detail.

3 is an example of dividing an iris region into a plurality of regions. Although the pupil 120 and the iris 110 are illustrated as concentric circles for convenience, it is apparent in the art that the actual pupil 120 and the iris 110 may not have exactly circular shapes.

Referring to FIG. 3 , the iris region 110 may be divided into an upper portion R1 , a middle portion R2 , and a lower portion R3 of the pupil 120 . In one embodiment, the iris region 110 is divided into a first region C1, a second region C2, a third region C3, and a fourth region C4 in a radial direction about the pupil 120. It can be. Here, the radiation regions C1 to C4 divided in the radial direction are exemplary, and the number of regions divided in the radial direction is not limited to four. Each of the radiation regions C1 to C4 may be divided into a plurality of unit division regions.

In one embodiment, the upper portion R1 may be an iris region at least partially covered by the upper eyebrow or eyelid. The lower part R3 may be an iris region at least partially covered by the lower eyebrow or eyelid. In one embodiment, the upper portion R1 may be about 60% of the distance from the center of the pupil 120 or the center of the iris 110 to the outer boundary of the iris 110 . The lower portion R3 may be about 40% of the distance from the center of the pupil 120 to the outer boundary of the iris 110 . In one embodiment, the upper part (R1) and lower part (R3) may not include the autonomic ring 130. In the process of mapping the iris 110 region, accurate iris coordinates (center, distance to the outer border of the iris) and pupil coordinates (center, distance to the pupil border) may be used.

In the human eye, the center of the pupil and the center of the iris actually do not coincide completely, but are in similar positions. Since the center of the pupil and the center of the iris can be considered to be approximately the same, what is referred to as the center of the pupil in this specification may actually mean the center of the iris.

4 is an example of dividing an iris region into a plurality of regions. In Fig. 4, the iris pattern is omitted. The region segmentation shown in FIG. 5(b) of US Patent Registration No. 8,009,876 can be obtained by mapping the iris region. A detailed description is disclosed in U.S. Patent No. 8,009,876 and incorporated herein. Region division will be additionally described later in this specification.

Referring to FIG. 4 , the size of the unit divided region included in the upper portion R1 close to the upper eyebrow may be smaller than the size of the unit divided region included in the lower portion R3 close to the lower eyebrow. In an embodiment, a size S11 of at least one unit divided region included in the upper portion R1 may be smaller than a size S21 of at least one unit divided region included in the lower portion R3. A size S21 of at least one unit partitioned area included in the lower portion R3 may be smaller than sizes S31 , S32 , S33 , and S34 of at least one unit divided area included in the middle portion R2 .

In one embodiment, the size of the unit segmentation region may increase toward the inner side of the iris. For example, in the upper part R1 , the size S11 of the unit segmentation region located outside the iris 110 may be smaller than the size S12 of the unit segmentation region located inside the iris. In the lower part R3 , the size S21 of the unit segmentation region located outside the iris 110 may be smaller than the size S22 of the unit segmentation region located inside the iris. In another embodiment, the size of the unit segmentation region may vary according to the pattern density of the iris 110 . For example, a transform with a higher degree of decomposition level can be applied to a region with high density.

5 is an example of dividing an iris region into a plurality of unit division regions. In Fig. 5, the iris pattern is omitted. Also, only a part of the unit partition area is shown in FIG. 5 for explanation. As shown in FIG. 3 , the iris region 110 is divided into a first region C1, a second region C2, a third region C3, and a fourth region C4 in a radial direction with the pupil 120 as the center. It can be. Here, the radiation regions C1 to C4 divided in the radial direction are exemplary, and the number of regions divided in the radial direction is not limited to four. For convenience of description, the autonomic ring 130 and the spokes 132 are shown to be included in the second region C2, but the positions of the ring 130 and the spokes 132 are not limited thereto. For example, the autonomic nerve ring 130 and/or spokes 132 may be positioned to span at least two areas among the first to fourth areas.

In one embodiment, the mapping calculates the coordinate values (centre, distance to the boundary line) of the figure by obtaining the most approximate circle or ellipse for each of the iris 110 and the pupil 120, and polar coordinates mapping. ), the donut-shaped iris region can be collectively converted into a rectangular shape.

In one embodiment, the iris region 110 may be mapped to a rectangular coordinate system, and then radial direction segmentation and unit region segmentation may be performed.

In one embodiment, after dividing the iris region 110 with the pupil 120 as the center in a radial direction, the iris region 110 may be mapped to a rectangular coordinate system, and then divided into unit division regions.

In one embodiment, the iris region 110 may be radially divided with the pupil 120 as the center, divided into unit division regions, and then mapped to a rectangular coordinate system.

An iris code is extracted from the segmented unit segmentation region. Various wavelet transforms can be used for iris code extraction. In the present disclosure, the Haar transform is described as an example, but the wavelet transform is not limited to the Haar transform.

In the present disclosure, when an iris code is extracted, it may be decomposed into different levels according to an iris pattern included in a unit segmentation region (or segmentation region).

In one embodiment, different levels of wavelet transform may be applied according to the density of the iris pattern. For example, a wavelet transform with a relatively high level of decomposition may be applied to an area where the density of iris patterns is (relatively) high, and a wavelet transform with a relatively low level of decomposition may be applied to an area where the density of iris patterns is (relatively) low. can

In one embodiment, the region with high density of the iris pattern may include a region where the autonomic ring 130 is located. Since the position of the area where the autonomic ring 130 exists may be different for each person, the area where the autonomic ring is located is estimated, and the area between the inner and outer borders of the iris is the area where the autonomic ring is estimated to be located. It can be divided into regions where the autonomic ring is not presumed to be located. In general, it can be estimated that the autonomic ring 130 is located at ⅓ of the pupil 120 side. Alternatively, as shown in FIG. 5 , based on the distances to the inner boundary and the outer boundary of the iris region 110, the region where the autonomic ring 130 is estimated to be located and the autonomic ring 130 are not located area can be divided.

In one embodiment, the dense region of the iris pattern may include the region where the pupillary sphincter 134 is present. Alternatively, the region where the iris pattern has a high density may include the region where the autonomic ring 130 and the pupillary sphincter 134 are located. For example, the pupillary sphincter 134 may be located in the C1 region. That is, the decomposition level of the wavelet transform applied to the area where the autonomic ring 130) and/or the pupillary sphincter 134 are located or the area estimated to be located (regions C1 and/or C2) is The decomposition level of the wavelet transform applied to the region where the sphincter muscle 134 is located or to the region C3 and/or C4 where it is estimated that the sphincter muscle 134 is not located may be relatively higher.

When the Haar transform is used as the wavelet transform, the unit segmentation region may be decomposed into different levels according to the iris pattern included in the segmentation region. Accordingly, the matching success rate can be increased for the iris region, which is deformed in various sizes according to lighting conditions. In one embodiment, the decomposition level applied when converting the divided region C21 where the autonomic ring 130 is located and the divided regions C31 and C41 where the autonomic ring 130 is not located may be different. For example, the decomposition level of the transformation applied to the segmented region C21 where the autonomic ring 130 is located is an order of magnitude higher than the decomposition level applied to the segmented regions C31 and C41 where the autonomic ring 130 is not located. ) can have. In addition, the order of the decomposition level may decrease as the distance from the segmentation region C21 where the autonomic ring 130 is located increases.

In one embodiment, a region in which the autonomic ring 130 of the iris and/or at least a part of the pupillary sphincter is located may be preset. The pupil region and the iris region are segmented by localization of the iris image, and a predetermined location region of the autonomic ring 130 and/or the pupillary sphincter may be determined from the segmented iris region. The iris region is divided into a plurality of segmented regions collectively according to a given rule, and the region containing (or likely to be included) the autonomic ring 130 and/or the pupillary sphincter is included in the autonomic ring 130 and/or the pupillary sphincter You can set it to where it exists.

For example, if D is the distance from the inner border of the iris 110 to the outer border of the area where the autonomic ring 130 is located, the area where the autonomic ring 130 is located is from the inner border of the iris 110. A region having a distance of 20% (D1) or more of D and a region having a distance (D2) of 40% or less of D from the outer boundary may be set as regions excluded from the iris 110 . The area where the pupillary sphincter is located may include an area having a distance D1 within 20% of the inner boundary of the iris 110 . These figures may vary due to individual differences.

6A to 6C are exemplary diagrams of decomposing an image into a plurality of levels. Wavelet transform or Haar transform analyzes a signal by taking various wavelet functions as basic functions and scaling and shifting the wavelet basic functions. As the scale value increases, low-frequency characteristics are exhibited, and as the scale value decreases, high-frequency characteristics are exhibited. In this way, the signal can be decomposed into an approximate component and a detail component. At this time, the approximate value is a low frequency component and the detailed value is a high frequency component. An approximate value is also called a vector of approximate coefficients, and a detail value is also called a vector of detail coefficients. In iris authentication, a low-frequency component is characterized to generate and store an iris code. For example, a low-frequency component may be stored as an iris code, or an iris code may be generated based on the low-frequency component.

6A is an exemplary diagram in which a 2D image is decomposed to a first level using wavelet transform. LL ₁ represents approximate values, and HL ₁ , LH ₁ and HH ₁ represent detailed values. FIG. 6B is an exemplary diagram in which LL ₁ of FIG. 6A is decomposed using wavelet transform once more (second level). LL ₂ represents approximate values, and HL ₂ , LH ₂ and HH ₂ represent detailed values. FIG. 6C is an exemplary view in which LL ₂ of FIG. 6B is decomposed using wavelet transform once more (third level). LL ₃ represents approximate values, and HL ₃ , LH ₃ and HH ₃ represent detailed values. It has been described above that the approximate value is a low-frequency component and the detailed value is a high-frequency component. From this, it can be seen that more feature values can be extracted as the wavelet transformation is repeated (as the decomposition level increases or the decomposition order increases).

7 is an example of a binary tree when the signal S is decomposed into three levels. Referring to FIG. 7 , the signal S is decomposed into a first level approximate value A ₁ and a first level detailed value D ₁ . The first level approximation value (A ₁ ) is decomposed into a second level approximation value (AA ₂ ) and a second level detail value (DA ₂ ). The second level approximation value AA ₂ is decomposed into a third level approximation value AAA ₃ and a third level detail value DAA ₃ . From this, it can be seen that three low-frequency components can be obtained by decomposing the signal into three levels.

As can be seen from FIGS. 6A to 6C and FIG. 7 , as the decomposition level increases, more approximate values (low frequency components) can be obtained.

8A is an example of converting an iris region into a coordinate system of a rectangular region. 8B is a conceptual diagram obtained by wavelet transforming the coordinate system of the rectangular region of FIG. 8A. In FIG. 8A, the iris pattern is omitted for convenience. In FIG. 8B, for convenience, only the leftmost unit partition regions D1, ANW1, B1, and A1 are shown after performing wavelet transformation.

Referring to FIG. 8A , regions C1', C2', C3', and C4' are included in the first region C1, second region C2, third region C3, and fourth region C4 of FIG. correspond to each It is assumed that the autonomic ring 130 (see FIG. 3) is located in the C2' region and the autonomic ring is not located in the C1', C3' and C4' regions. Wavelet transform is applied to the unit partition regions D1, ANW1, B1, and A1 of regions C1', C2', C3', and C4' with different decomposition level orders.

In one embodiment, the deepest or most decomposition level is applied to the unit segmentation region ANW1 of region C2' where the autonomic ring is located. Different decomposition levels can be applied to the unit slices (D1, B1, A1) of the regions (C1', C3', C4') that do not contain the autonomic ring and the unit slice (ANW1) of the C2' region where the autonomic ring is located. can

In one embodiment, the level of degradation may vary depending on the distance from the autonomic ring. For example, the level of degradation may decrease with distance from the autonomic ring. As shown in FIG. 8B, when the unit segmentation area (ANW1) of the C2' region where the autonomic ring is located is decomposed into three levels, the unit segmentation regions of C1' and C3' regions immediately adjacent to the C2' region where the autonomic ring is located. (D1, B1) can be decomposed into two levels, and the unit division area A1 of area C4' can be decomposed into one level. Here, the order of the level (1, 2, 3) is exemplary for description, and the order of the level is not limited thereto. In an embodiment, decomposition levels applied to the unit divided regions D1, B1, and A1 other than the unit divided region ANW1 of region C2' may be different or the same. In addition, the same decomposition level may be applied to the adjacent unit partition regions D1, D2/ANW1, ANW2/B1, B2/A1, and A2 of the first to fourth regions C′1 to C4, or different decomposition levels may be applied. this may apply.

Alternatively, although not shown, in one embodiment, the decomposition level of the C1' and C3' regions immediately adjacent to the C2' region where the autonomic ring is located may be the same as that of the C4' region.

9A is another example of converting an iris region into a coordinate system of a rectangular region. FIG. 9B is a conceptual diagram obtained by performing wavelet transformation on the rectangular region coordinate system of FIG. 9A. In FIG. 9A, the iris pattern is omitted for convenience. Referring to FIG. 9A , regions C1', C2', C3', and C4' are included in the first region C1, second region C2, third region C3, and fourth region C4 of FIG. correspond to each It is assumed that the autonomic ring 130 (see FIG. 3) is located in the C2' region and the autonomic ring is not located in the C1', C3' and C4' regions. Wavelet transform may be applied to regions C1', C2', C3', and C4' at different decomposition levels without dividing regions C1', C2', C3', and C4' into a plurality of unit partitioned regions.

In one embodiment, the deepest or most decomposition level is applied to the unit segmentation area (ANW) of region C2' where the autonomic ring is located. A different decomposition level can be applied to regions (C1', C3', C4') that do not contain the autonomic ring than the C2' region where the autonomic ring is located. As shown in FIG. 9B, when the C2' region where the autonomic ring is located is decomposed into three levels, the C1' and C3' regions immediately adjacent to the C2' region where the autonomic ring is located are decomposed into two levels, and the C4' region is It can be disassembled in 1 level. Alternatively, the same decomposition level may be applied to the regions C1', C3', and C4' that do not include the autonomic ring.

In one embodiment, a combination of FIGS. 8A and 9A is also possible. For example, at least one of C1', C3', and C4' may be divided into unit partitions, without dividing the C2' region where the autonomic ring is located.

In one embodiment, as shown in FIG. 8A, the first to fourth regions C1 to C4 are converted into a rectangular coordinate system, and then, as shown in FIG. 9A, the regions C1', C2', C3', and C4' are converted into a plurality of units. It can be divided into partitioned areas.

Referring back to FIG. 7 , the low frequency component varies according to the division level. For example, when an image is decomposed into three levels as shown in FIG. 7 , three low frequency components (A ₁ , AA ₂ , AAA ₃ ) are generated. An iris code may be generated and used based on the generated low-frequency components. Therefore, it can be seen that different numbers of low-frequency components can be extracted and different iris codes can be generated according to the decomposition level applied to each segmented region. In one embodiment, each iris region or segmented region in which the autonomic ring is located may be decomposed to have more low-frequency components than the iris region or segmented region in which the autonomic ring is not located. The low-frequency component (or the iris code generated based on the low-frequency component) extracted from the iris region or segmented region where the autonomic ring is located can be used as a feature in the subsequent matching process.

10 is a conceptual diagram of an iris authentication system 1000 according to an embodiment of the present disclosure. The iris authentication system 1000 includes an acquisition module 1010, a segmentation module 1020, a normalization module 1030, a division module 1035, and an encoding module 1040. ) and a matching module (1050). The iris authentication system 1000 may further include a database 1060 . The iris authentication system 1000 is configured to perform iris authentication using the methods described in this disclosure and references.

The acquisition module 1010 is configured to acquire an image of an eye. The segmentation module 1020 is configured to detect the iris from the image of the eye to obtain an iris image, and localize the iris image to distinguish between a pupil region and an iris region. The normalization module 1030 is configured to apply a geometric normalization scheme to transform the segmented iris regions into interpretable rectangular regions in row and column format. The segmentation module 1035 is configured to divide the iris region into multiple divided regions and/or unit divided regions or divide the rectangular region into multiple divided regions and/or unit divided regions. The encoding module 1040 is configured to extract features from the iris region converted into rectangular regions, segmented regions, and/or unit segmented regions using feature extraction rules and generate iris codes. The encoding module 1040 may transmit the generated iris code to the database 1060 . The matching module 1050 is configured to receive the iris code from the database 1060 and match it with the iris code generated by the encoding module. The matching module 1050 determines iris authentication based on the matching result. Database 1060 is configured to store iris codes and transmit iris codes to matching module 1050 .

The devices and methods described above may be implemented as hardware components, software components, and/or a combination of hardware components and software components. For example, devices and components described in the embodiments may include, for example, a processor, a controller, an arithmetic logic unit (ALU), a digital signal processor, a microcomputer, a field programmable array (FPA), It may be implemented using one or more general purpose or special purpose computers, such as a programmable logic unit (PLU), microprocessor, or any other device capable of executing and responding to instructions. The processing device may run an operating system (OS) and one or more software applications running on the operating system. A processing device may also access, store, manipulate, process, and generate data in response to execution of software. For convenience of understanding, there are cases in which one processing device is used, but those skilled in the art will understand that the processing device includes a plurality of processing elements and/or a plurality of types of processing elements. It can be seen that it can include. For example, a processing device may include a plurality of processors or a processor and a controller. Other processing configurations are also possible, such as parallel processors.

Software may include a computer program, code, instructions, or a combination of one or more of the foregoing, which configures a processing device to operate as desired or processes independently or collectively. You can command the device. Software and/or data may be any tangible machine, component, physical device, virtual equipment, computer storage medium or device, intended to be interpreted by or to provide instructions or data to a processing device. , or may be permanently or temporarily embodied in a transmitted signal wave. Software may be distributed on networked computer systems and stored or executed in a distributed manner. Software and data may be stored on one or more computer readable media.

The method according to the embodiment may be implemented in the form of program instructions that can be executed through various computer means and recorded on a computer readable medium. Computer readable media may include program instructions, data files, data structures, etc. alone or in combination. Program instructions recorded on the medium may be specially designed and configured for the embodiment, or may be known and usable to a person in charge of computer software. Examples of computer-readable recording media include magnetic media such as hard disks, floppy disks and magnetic tapes, optical media such as CD-ROMs and DVDs, and magnetic media such as floptical disks. - optical media (magneto-optical media), and specially configured to store and execute program instructions such as ROM, RAM, flash memory, etc.

Hardware devices are included. Examples of program instructions include high-level language codes that can be executed by a computer using an interpreter, as well as machine language codes such as those produced by a compiler. The hardware devices described above may be configured to operate as one or more software modules to perform the operations of the embodiments, and vice versa.

As described above, although the embodiments have been described with limited drawings, those skilled in the art can apply various technical modifications and variations based on the above. For example, the described techniques may be performed in an order different from the method described, and/or components of the described system, structure, device, circuit, etc. may be combined or combined in a different form than the method described, or other components may be used. Or even if it is replaced or substituted by equivalents, appropriate results can be achieved.

Therefore, other implementations, other embodiments, and equivalents of the claims are within the scope of the following claims.

Iris features are extracted by applying transformations of different orders to regions with many iris patterns and regions with few iris patterns. Accordingly, it is possible to provide an iris authentication method and system with increased discrimination of iris recognition.

100: eye image 110: iris
120: pupil 130: autonomic ring
1010: acquisition module 1020: segmentation module
1030: normalization module 1035: division module
1040: encoding module 1050: matching module
1060: database

Claims (20)

A method for generating an iris code from an image of an iris,
Dividing at least a portion of a region between the inner boundary and the outer boundary of the iris into a plurality of divided regions, wherein the plurality of divided regions include a first divided region including at least a part of the autonomic ring of the iris and the autonomic ring. including a second partitioned area that does not include;
extracting a plurality of first features by applying a first transform to the first divided region and extracting at least one second feature by applying a second transform to the second divided region; and
generating an iris code based on the plurality of first characteristics and the at least one second characteristic;
The plurality of divided regions further includes a third divided region not including the autonomic ring, and the first divided region is adjacent to the second divided region, but the second divided region comprises the first divided region and the first divided region. Located between the 3 partitions,
Further comprising extracting at least one third feature by applying a third transform to the third divided region after dividing at least a part of the region between the inner boundary and the outer boundary of the iris into a plurality of divided regions. do,
wherein the decomposition level of the first transform has a higher order than the decomposition level of the third transform. According to claim 1,
wherein the first transform and the second transform include a wavelet transform, and a decomposition level of the first transform has a greater order than a decomposition level of the second transform. . According to claim 1,
The method of generating an iris code further comprising the step of further dividing at least one of the first divided region and the second divided region into a plurality of unit divided regions. delete According to claim 1,
wherein the decomposition level of the second transform has a higher order than the decomposition level of the third transform. According to claim 1,
Prior to the step of dividing at least a part of the region between the inner boundary and the outer boundary of the iris into a plurality of divided regions,
The method of generating an iris code further comprising mapping at least a portion of a region between an inner boundary and an outer boundary of the iris to a rectangular region. According to claim 6,
After mapping the plurality of divided regions into rectangular regions,
The method of generating an iris code further comprising the step of further dividing at least one of the first divided region and the second divided region into a plurality of unit divided regions. According to claim 7,
Further dividing into the plurality of unit division areas,
Dividing a part of the unit divided region with a relatively high possibility of being covered by the upper eyebrow into a first unit divided region, and dividing a part of the unit divided region with a relatively high possibility of being covered by the lower eyebrow into a second unit divided region; , Dividing a portion with a relatively small possibility of being occluded into a third unit divided region, wherein the size of the first unit divided region < the size of the second unit divided region < the size of the third unit divided region - How to generate an iris code. According to claim 1,
generating an iris code, wherein the first transform and the second transform include a Haar transform, and wherein the plurality of first features and the at least one second feature include a low frequency component; How to. According to claim 1,
Dividing at least a portion of the region between the inner boundary and the outer boundary of the iris into a plurality of divided regions further comprises determining a region in which the autonomic ring is located. According to claim 10,
Determining the region where the autonomic ring is located is
Based on the distance from the pupil to the inner boundary and the outer boundary of the iris, determining the region where the autonomic ring is located, iris code generating method. In the iris authentication system,
an acquisition module configured to acquire an iris image;
a segmentation module configured to segment an iris region from an iris image;
a division module configured to divide the iris region into a plurality of divided regions including a first divided region including the autonomic ring and a second divided region not including the autonomic ring and adjacent to the first divided region;
A plurality of first features are extracted by applying a first transform having a first decomposition level to the first split region, and a second transform having a second decomposition level smaller than the first decomposition level is applied to the second split region. an encoding module for extracting at least one second feature by applying and generating an iris code based on the plurality of first features and the at least one second feature; and
A matching module for receiving and comparing the iris code with the iris code from the database and performing iris authentication based on the comparison result,
The segmentation area further includes a third segmentation area not including the autonomic ring, the first segmentation area is adjacent to the second segmentation area, and the second segmentation area is adjacent to the first segmentation region and the first segmentation region. Located between the 3 partitions,
The encoding module extracts at least one third feature by applying a third transform to the third partitioned region, and the plurality of first features, the at least one second feature, and the at least one third feature are extracted. An iris authentication system based on which an iris code is generated. delete According to claim 12,
wherein the first transform and the second transform include a Haar transform, and wherein the plurality of first features and the at least one second feature include a low frequency component. According to claim 12,
A normalization module converting the iris region into a rectangular region interpretable in row and column format;
The segmentation module divides the rectangular area into the plurality of divided areas, iris authentication system. A method for generating an iris code from an image of an iris,
Dividing at least a portion of a region between an inner boundary and an outer boundary of the iris into a plurality of divided regions, wherein the plurality of divided regions include a first divided region including a first iris pattern of a first density and the first divided region. including a second segmented region including a second iris pattern having a second density smaller than the density;
extracting a plurality of first features by applying a first transform to the first divided region and extracting at least one second feature by applying a second transform to the second divided region; and
generating an iris code based on the plurality of first characteristics and the at least one second characteristic;
The first divided region includes a region where at least one of the autonomic ring and pupillary sphincter is located, and the second divided region does not include a region where the autonomic ring and pupillary sphincter are located;
wherein the first transform and the second transform include a wavelet transform, and a decomposition level of the first transform has a greater order than a decomposition level of the second transform. . delete A method for generating an iris code from an image of an iris,
converting at least a part of a region between an inner boundary and an outer boundary of the iris into a rectangular region coordinate system;
Dividing the region between the inner boundary and the outer boundary of the iris converted into the square region coordinate system into a plurality of divided regions, wherein the plurality of divided regions include a first divided region where the autonomic ring is estimated to be located and an autonomic ring includes a second partition area estimated not to be located;
extracting a plurality of first features by applying a first transform to the first split region and extracting at least one second feature by applying a second transform to the second split region; the level has a higher order than the decomposition level of the second transform; and
generating an iris code based on the plurality of first characteristics and the at least one second characteristic;
The second split area is divided into a plurality of split areas, the plurality of split areas include a second split area adjacent to the first split area and a second split area not adjacent to the first split area;
Characterized in that the first divided region, the second divided region adjacent to the first divided region, and the second divided region not adjacent to the first divided region have different or identical decomposition levels, How to generate an iris code. delete According to claim 18,
The method of generating an iris code, wherein the first segmented area further includes an area where a pupillary sphincter is located, and wherein the second segmented area is located outside the first segmented area.

Images