KR20170003348A - Updating method and apparatus of registration database for user authentication - Google Patents

Abstract

The present invention relates to a method for updating a registration database for biometric authentication, which comprises the steps of: authenticating an input image by using a registration database; and adaptively updating the registration database based on a time required for authenticating the input image and a pre-calculated result for authenticating the input image.

Description

TECHNICAL FIELD [0001] The present invention relates to a method and an apparatus for updating a registration database for user authentication,

The following embodiments relate to a method and apparatus for updating a registration database for user authentication.

Recently, the importance of security authentication is increasing due to the development of various mobile devices such as smart phones and wearable devices. Biometrics authentication technology authenticates users using fingerprints, irises, voices, faces, blood vessels, and so on. The biometric characteristics used for authentication vary from person to person, not only because they have no inconvenience of possession, but also because they have little risk of theft or imitation, and they do not change well during their lifetime.

Especially, fingerprint recognition technique is most commercialized for convenience, security and economical reasons. For example, the fingerprint recognition technique can enhance the security of the user device and easily provide various application services such as mobile payment.

A method of updating a registration database for biometric authentication according to one side includes: authenticating an input image using the registration database; And adaptively updating the registration database based on a first time required for the authentication of the input image and a result obtained by performing a pre-calculation for authentication of the input image.

The step of adaptively updating the registration database may include: determining similarities based on sizes of overlapping regions between registered images included in the registration database and the input image; And determining whether to register the input image in the registration database, based on the similarities.

Wherein the determining the similarities comprises: determining at least some of the similarities using the pre-calculated results; And determining at least a portion of the similarities as a predetermined minimum value to satisfy a deadline based on the first time.

Wherein the step of determining whether to register in the registration database includes: determining whether the maximum similarity among the similarities is less than or equal to an upper threshold; And determining whether the maximum similarity is greater than or equal to a lower threshold value. The upper threshold value may correspond to a maximum similarity determined based on the overlap area between the registered images.

The pre-calculated result may include at least one of the intra-overlap correlation between the input image and the registered images, and the size of the overlapped areas between the input image and the registered images.

Wherein the step of authenticating the input image comprises: calculating a size of an overlap area between the input image and a registered image included in the registration database; Calculating an intra-overlap correlation between the input image and the registered image; Outputting an accept signal when the correlations are above a threshold; And storing the size of the overlapping area and the degree of correlation in a preset storage space corresponding to the registered image. The step of authenticating the input image may further include storing the first time in the storage space.

Wherein the step of adaptively updating comprises: estimating a second time available for updating the registration database based on the first time; And adaptively updating the registration database based on the second time. The step of estimating the second time may include calculating the second time based on a difference between a third time allowed for the biometric authentication and the first time.

Wherein the step of adaptively updating the registration database comprises: adding the input image to the registration database; And replacing any one of the registered images included in the registration database with the input image. The replacing may include selecting a replacement image based on similarities based on the size of the overlap region between the registered images.

A method of updating a registration database for biometric authentication according to one side includes: receiving an input image including biometric information; Determining a maximum similarity based on the size of an overlapped area between the input image and the registered images; Comparing an upper threshold based on the maximum similarity and the size of the overlap region between the registered images; And registering the input image in the registration database according to the comparison result. The biometric information may include at least one of fingerprint information, blood vessel information, and iris information.

Wherein the step of determining the maximum similarity comprises: calculating similarities between the input image and the registered images based on the size of the overlap region between the input image and the registered images; And determining the similarity having the largest value among the similarities as the maximum similarity.

The calculating of the degrees of similarity may include calculating an intra-overlapping correlation between any one of the registered images and the input image; Calculating a size of an overlap region between any one of the registered images and the input image; And calculating the degree of similarity between any one of the registered images and the input image using the degree of correlation and the size of the overlapping region.

The calculating of the degrees of similarity may include calculating a normalized cross correlation (NCC) using a motion vector and a rotation angle calculated by image registration between any one of the registered images and the input image, ) ≪ / RTI > Calculating a share area ratio between any one of the registered images and the input image; And calculating the degree of similarity between any one of the registered images and the input image using the normalized cross-correlation value and the shared area ratio.

The updating method of the registration database may further include selecting a replacement image among the registered images based on the similarities based on the size of the overlap region between the registered images.

Wherein the selecting of the replacement image comprises: selecting candidate images to be replaced among the registered images based on the similarity based on the size of the overlapping region between the registered images; Accumulating similarity between each of the candidate images and the remaining registered images; And selecting a candidate image having a high degree of similarity among the candidate images as the replacement image.

Wherein the step of registering the input image in the registration database includes: updating the registration database by replacing the replacement image with the input image if the maximum similarity degree is smaller than the upper threshold value; And maintaining the replacement image in the registration database if the maximum similarity is greater than the upper threshold value. The upper threshold value may include a second maximum similarity among the similarities based on the size of the overlap region between the registered images.

The step of registering the input image in the registration database may include comparing the maximum similarity with a preset lower threshold value, And registering the input image in the registration database based on a result of comparison with the lower threshold value.

The step of registering the input image in the registration database based on a result of comparison with the lower threshold may include registering the input image in the database if the maximum similarity degree is larger than the lower threshold value and smaller than the upper threshold value can do.

Wherein the updating method of the registration database comprises: calculating an authentication score between the input image and the registered images; And performing authentication of the input image by comparing the authentication score with an authentication threshold, wherein the step of determining the maximum degree of similarity is a step of calculating, between the input image and the registered images, And determining the maximum similarity based on the similarities.

An apparatus for updating a registration database according to one side includes: a memory for storing a registration database including registered images; And a processor for authenticating the input image using the registration database and adaptively updating the registration database based on a first time period required for the authentication of the input image and a pre- .

An apparatus for updating a registration database according to one side includes: a memory for storing a registration database including the registered images; And determining the maximum similarity based on the size of the overlapping region between the input image and the registered images and comparing the input image with the registered image based on the comparison result of the maximum similarity degree and the upper threshold value based on the size of the overlapping region between the registered images. And registers it in the database.

A database management method according to one aspect includes: receiving an input image including biometric information; Comparing the input image with registered images included in a database and authenticating the registered image; And an update step of determining whether to add the input image to the database or to replace the input image with any one of registered images included in the database.

The sum of the time required for the authentication step and the time required for the update step may be less than or equal to a predetermined maximum time. The biometric information may be fingerprint information.

1 is a view for explaining an operation of acquiring an input image including biometric information according to an embodiment;
2 is a diagram illustrating a recognition rate improved according to an update of a registration database according to an embodiment;
3 is a flowchart illustrating a method of updating a registration database according to an exemplary embodiment;
4 is a flow chart illustrating a method of computing similarity according to one embodiment.
5 is a diagram for explaining a method for determining a maximum similarity according to an embodiment;
6 is a flowchart illustrating a method of updating a registration database according to another embodiment;
FIG. 7 is a flowchart illustrating a method of selecting a replacement image according to an exemplary embodiment. FIG.
8 and 9 are views for explaining a method of selecting a replacement image according to an embodiment.
10 is a flowchart illustrating a method of updating a registration database according to an embodiment;
11 is a view for explaining an authentication process of biometrics according to an embodiment;
12 illustrates an adaptive update of a registration database according to one embodiment;
13 is an operational flow diagram illustrating a method for adaptively updating a registration database according to an embodiment;
14 is a block diagram illustrating a multi-stage authenticator in accordance with one embodiment.
15 illustrates a rough search according to an embodiment;
16 is a view for explaining a fine search according to an embodiment;
17 to 22 are diagrams illustrating an update apparatus for adaptively updating a registration database according to an embodiment.
23 is an operational flow diagram illustrating a method for adaptively updating a registration database according to an embodiment;
24 is a block diagram of an apparatus for updating a registration database according to an embodiment;
25 is a view for explaining a method of recognizing a vein pattern according to an embodiment;

The specific structural or functional descriptions disclosed herein are merely illustrative for the purpose of illustrating embodiments. The disclosed embodiments may be modified and implemented in various other forms, and the scope of the present disclosure is not limited to the disclosed embodiments.

The terms first or second may be used to describe various elements, but such terms should be understood only for the purpose of distinguishing one element from another. For example, the first component may be referred to as a second component, and similarly, the second component may also be referred to as a first component.

It is to be understood that when an element is referred to as being "connected" or "connected" to another element, it may be directly connected or connected to the other element, . On the other hand, when an element is referred to as being "directly connected" or "directly connected" to another element, it should be understood that there are no other elements in between. Expressions that describe the relationship between components, such as "between" and "between" or "neighboring to" and "directly adjacent to" should be interpreted as well.

The singular expressions include plural expressions unless the context clearly dictates otherwise. In this specification, the terms "comprises ", or" having ", and the like, are used to specify one or more of the features, numbers, steps, operations, elements, But do not preclude the presence or addition of steps, operations, elements, parts, or combinations thereof.

Unless otherwise defined, all terms used herein, including technical or scientific terms, have the same meaning as commonly understood by one of ordinary skill in the art. Terms such as those defined in commonly used dictionaries are to be interpreted as having a meaning consistent with the meaning of the context in the relevant art and, unless explicitly defined herein, are to be interpreted as ideal or overly formal Do not.

Embodiments to be described below may use an input image including biometric information to perform a recognition operation and further update a registration database. For example, a fingerprint recognizer, a vein recognizer, an iris recognizer, or the like may perform a recognition operation to authenticate or identify a user, and use the authenticated or identified biometric information to update an existing registration database.

Embodiments may be implemented in various forms of products, such as personal computers, laptop computers, tablet computers, smart phones, televisions, smart home appliances, intelligent cars, kiosks, wearable devices, For example, embodiments may be applied to the processing of images to authenticate users in smart phones, mobile devices, smart home systems, and the like. In the same manner, the embodiments can be applied to a payment service through user authentication, an intelligent automobile system, and the like. Embodiments can provide a robust user authentication result even when the user's finger is small or the fingerprint input is incorrect in the above-described scenarios. Hereinafter, embodiments will be described in detail with reference to the accompanying drawings. Like reference symbols in the drawings denote like elements.

The size of the nested region Based  Technique to update registration database using similarity

1 is a view for explaining an operation of acquiring an input image including biometric information according to an exemplary embodiment. Hereinafter, it is assumed that the biometric information is a fingerprint for convenience of explanation. However, the embodiments can be equally applied to various biometric information recognizable in the form of images such as veins and irises.

Referring to FIG. 1, a sensor 110 according to an embodiment detects a fingerprint of a user. For example, the sensor 110 may comprise a plurality of sensing elements. The plurality of sensing elements may be arranged in an array or a matrix structure. The sensor 110 may sense a fingerprint input in the form of an analog signal using a plurality of sensing elements. The sensor 110 may convert the sensed analog signal to a digital image using an analog-to-digital converter. Hereinafter, the input image may refer to the converted digital image.

At this time, the size of the sensor 110 may be different from the size of the user's finger 120. For example, if the sensor 110 is mounted on a mobile device, the size of the sensor 110 may be less than the size of the user's finger 120 to satisfy the form factor constraints of the mobile device. have. In this case, the sensor 110 can detect only a part of the fingerprint of the user's finger 120. [ Hereinafter, the input image sensed by the sensor 110 may be a partial image of the user fingerprint.

The sensor standard may be different for each manufacturer of the sensor 110, and the sensor standard may include the size of the sensor array, the resolution of the sensor, and the like. The size of the sensor array may indicate the size of the sensing area formed by the plurality of sensing elements. For example, the size of the sensor array may be 1 cm (length) x 3 cm (width). The resolution of the sensor may represent the number of sensing elements per unit length or unit area. If the size of the sensor array is the same, but the resolution of the sensor is large, the sensor may contain more sensing elements. For example, the resolution of the sensor may be 56 pixels (length) x 144 pixels (width).

Embodiments provide techniques for updating an enrollment database (enrollment database). The registration database may include previously registered images and the registered images may be partial images of the fingerprint of the enrollment user. The update of the registration database may include, for example, an operation of adding a new registered image to the registration database, an operation of replacing some of the registered images included in the registration database with a new registered image, An operation of removing a part, and the like. The registration database may also be referred to as a template.

Referring to FIG. 2, the recognition rate increases in proportion to the registered area. The registered area means the size of the user's fingerprint area covered by the registered images included in the registration database. For example, if the registered images included in the registration database cover the first registration area 130, the recognition rate is 93%. If the registered images included in the registration database cover the second registration area 140, the recognition rate is 99 %. ≪ / RTI >

Embodiments can update the registration database in the direction that the registered area increases. For example, the registered images included in the registration database at the first time point may correspond to the first registered area 130. The first point in time may be the point at which the user first registers the fingerprint to use the electronic device to which the embodiments are applied.

The user can input a fingerprint for user authentication while using the electronic device. Embodiments can increase the registered area by updating the registration database by using the input image that succeeds in user authentication. For example, the registered images included in the registration database at the second time point may correspond to the second registered area 140. The second time point may be a time point at which the registration database is updated using the fingerprint image input for user authentication.

According to the embodiments, user convenience for fingerprint registration can be improved. For example, the embodiments can provide a technique of updating a registration database in a state in which only a predetermined number of registered images are registered in the initial registration step and the user can not recognize in the actual use step.

As will be described in detail below, the embodiments determine whether to register the input image in the registration database using the maximum similarity determined based on the size of the overlapped area between the input image and the registered images, A new fingerprint area can be registered in comparison with the registered image.

Embodiments can acquire more registered images without changing the size of the registration database by replacing the selected replacement image and the input image among the registered images based on the similarities based on the size of the overlap area between the registered images. In addition, embodiments can prevent an outlier from being reflected in a registration fingerprint by registering an input image in a registration database by comparing the maximum similarity with a preset lower threshold value.

In addition, the embodiments can prevent the fingerprint recognition rate from being lowered by excluding the input image containing less fingerprint information from the authentication object. Embodiments can prevent registration of a wrong input fingerprint, a damaged fingerprint, or the like as a registered image by reviewing registration in the registration database with respect to an input image in which user authentication is accepted, and improve the fingerprint recognition rate of the fingerprint recognizer.

In addition, the embodiments can provide a technique of adaptively updating the registration database based on the time taken to authenticate the input image. Embodiments may provide a technique for utilizing the result of the pre-operation to authenticate an input image to update a registration database.

3 is a flowchart illustrating a method of updating a registration database according to an exemplary embodiment of the present invention. The updating method of the registration database may be performed by an updating apparatus (hereinafter referred to as an updating apparatus) of the registration database according to an embodiment. The updating device may be included in the fingerprint authenticator, for example, or may be configured as a separate device.

The updating device may be implemented as a software module and driven by at least one processor. A software module may be written in a form of a program in a memory coupled to the processor. Alternatively, the updating device may be implemented as a hardware module. Alternatively, the updating device may be implemented as a combination of a software module and a hardware module. In this case, the functions implemented by the software module are performed by the processor, and the functions implemented by the hardware module may be performed by the corresponding hardware. Processors and hardware can exchange signals with each other through input / output buses.

Referring to FIG. 3, the updating apparatus receives an input image including biometric information (210). The updating device may receive an input image from a sensor that detects biometric information. The input image may include biometric information such as, for example, fingerprint information, blood vessel information, and iris information.

The updating device determines the maximum similarity based on the size of the overlap region between the input image and the registered images (220). The updating apparatus can calculate the similarities between the input image and the registered images based on the size of the overlap region between the input image and the registered images. For example, if the number of registered images is 15, the updating apparatus calculates the similarity between the input image and the registered image # 1, the similarity between the input image and the registered image # 2, the similarity between the input image and the registered image # 15 Can be calculated. The updating apparatus can determine the similarity having the largest value among the 15 calculated similarities as the maximum similarity.

The degree of similarity according to the embodiments not only reflects how similar the two images are in the overlapped area, but also reflects the size of the overlapped area of the two images. For example, assume that the first case in which the two images are the same in the overlapping region, but the second case in which the overlapping region is not the same in the overlapped region, but the size of the overlapping region is very large have. When considering how similar two images are within the overlapping region, the similarity of the first case can be calculated to be higher than the similarity of the second case. On the other hand, according to the embodiments, the similarity of the second case can be calculated to be higher than the similarity of the first case.

A method for calculating the degree of similarity between any one registered image and an input image will be described with reference to FIG. 4, and a method for determining the maximum similarity will be described with reference to FIG.

The updating device compares the upper threshold based on the maximum similarity and the size of the overlap region between the registered images (230). The upper threshold value may include a second maximum similarity having the largest value among the similarities based on the size of the overlap region between the registered images. For example, the upper threshold value may be a degree of similarity based on the size of the overlap region between two registered images that overlap most of the registered images.

The updating apparatus registers the input image in the registration database according to the comparison result of step 230 (240). The updating device can determine whether the maximum similarity determined between the input image and the registered images is smaller than the upper threshold value. When the maximum similarity degree is smaller than the upper threshold value, the updating apparatus can update the registration database in the direction of increasing the registered area by registering the input image in the registration database.

Here, 'maximum similarity is smaller than upper threshold' means that the degree of similarity based on the overlapping region between the input image and the registered image is lower than the similarity based on the overlapping region between registered images. The 'maximum similarity is smaller than the upper threshold value' can be understood as meaning that the input image includes a new fingerprint area that does not overlap with registered images registered in the registration database at a certain level or more.

The updating device may add an input image to registered images previously registered in the registration database. Alternatively, the update device may select a replacement image to be replaced with the input image among the registered images, and replace the replacement image with the input image to register the input image.

The update apparatus may not update the registration database if the maximum similarity degree is greater than or equal to the upper threshold value. For example, if the maximum similarity is greater than or equal to the upper threshold, the updating apparatus can maintain the replacement image as a registered image in the registration database.

According to the embodiment, the updating apparatus can compare the maximum similarity with a preset lower threshold. The updating apparatus can register the input image in the registration database based on the comparison result with the lower threshold value. The update apparatus can register the input image in the registration database when the maximum similarity degree is larger than the lower threshold value. The 'lower threshold' can be understood as the minimum similarity required for the input image to be recognized as the fingerprint of the user corresponding to the recognition object. Here, 'maximum degree of similarity greater than the lower threshold value' can be understood as meaning that the overlapping area with the registered images that should be displayed by the same user includes the input image at a certain level or more. The update apparatus can prevent the outlier from being reflected in the registration fingerprint by registering the input image in the registration database when the maximum similarity degree is larger than the lower threshold value.

Embodiments can provide a technique of improving the fingerprint recognition rate of a registered user while preventing contamination of the registration database due to fingerprints of other users by updating the registration database only when the maximum similarity degree falls between the upper threshold value and the lower threshold value .

4 is a flowchart illustrating a method of calculating the degree of similarity according to an embodiment.

Referring to FIG. 4, an update apparatus according to an exemplary embodiment may calculate an intra-overlap correlation between a registered image and an input image of the registered images (310). The intra-overlap correlation may be, for example, a normalized cross correlation (ncc) value.

The updating device may calculate the size of the overlap area between any one registered image and the input image (320). The size of the overlapping area can also be referred to as a " shared image area ". The updating device can calculate the shared area ratio between any one registered image and the input image using the shared image area.

The updating apparatus can calculate the similarity between any one registered image and the input image using the degree of correlation and the size of the overlapping region (330). For example, the similarity may be calculated as the product of the normalized cross-correlation value and the shared image area.

FIG. 5 is a diagram for explaining a method of determining the maximum similarity according to an embodiment. Referring to FIG. 5, a registered image 410 and an input image 420 are shown.

The updating apparatus may calculate the similarities between the input image and the registered images, and may determine the similarity having the largest value among the similarities as the maximum similarity. For example, if the number of registered images is 15 and the first registered image among the 15 registered images is a registered image 410, the updating apparatus calculates the similarity between the registered image 410 and the input image 420 Is as follows.

The update apparatus can calculate the motion vector and the rotation angle between the registered image 410 and the input image 420 through image registration between the registered image 410 and the input image 420. [ Image matching is a processing technique that transforms different images and displays them in one coordinate system. Through image matching, we can see how images obtained through different measurement methods correspond. The image matching can be classified into a spatial domain scheme and a frequency domain scheme.

The spatial domain method is a method of matching the pattern or characteristic of the pixel intensity in the image space. For example, image deformation between two images can be detected using a random sample consensus (RANSAC) technique when the number of feature points included in the two images is greater than the minimum number required for the image transformation. In addition, the frequency domain method is a method of directly detecting parameters required for transformation between two images in the frequency domain. The parameters required for the transformation that can be detected at this time are, for example, image movement, rotation angle, size change, and the like.

The updating device can calculate the normalized cross-correlation value using the calculated motion vector and rotation angle. For example, a normalized cross-correlation value between two images I ₁ and I ₂ can be calculated using the following equation (1).

Here, W is the distance between the image I ₁ and the image I ₂ Represents an overlap area. i represents the X axis coordinate of the image, j represents the Y axis coordinate of the image, x represents the X axis movement value, and y represents the Y axis movement value. I ₁ (i, j) denotes a pixel value in the image I ₁ (i, j) coordinates, and the image I ₂ (x + i, y + j) is the image I ₂ (x + i, y + j ) ≪ / RTI > coordinates. Image I ₁ is registered image 410, image I ₂ is It can be understood as an input image 420.

The updating device can calculate the size of the overlap area between the registered image 410 and the input image 420. [ The size of the overlapping area may be a shared image area. The updating apparatus can calculate the shared area ratio between the registered image 410 and the input image 420 using the shared image area. The share area ratio can be obtained, for example, by the following equation (2).

The updating device can calculate the shared image area for the total image area of the registered image 410 and the input image 420 as a shared area ratio. According to an embodiment, the updating apparatus may calculate a shared area ratio by weighting a shared image area.

The update apparatus can calculate the similarity between the registered image 410 and the input image 420 using the normalized cross-correlation value and the shared area ratio. The similarity between the registered image 410 and the input image 420 can be obtained by the following equation (3).

The updating apparatus can calculate the similarity between the registered image 410 and the input image 420 by multiplying the normalized cross-correlation value and the shared area ratio.

The update apparatus can calculate the similarity between the 14 registered images and the input image 410 except for the registered image 410 in which the similarity is calculated, in the same manner as described above. The updating apparatus can determine the similarity having the largest value among the 15 calculated similarities as the maximum similarity.

6 is a flowchart illustrating a method of updating a registration database according to another embodiment. Referring to FIG. 6, the updating apparatus receives an input image including biometric information (510). The update apparatus calculates the similarities between the input image and the registered images based on the size of the overlap region between the input image and the registered images (520), and determines the similarity having the largest value among the calculated similarities as the maximum similarity (530). The update device may select a replacement image among the registered images based on the similarities based on the size of the overlap area between the registered images (540). The method by which the update apparatus selects the replacement image will be described with reference to FIGS. 7 to 9. FIG.

The update device may determine whether the maximum similarity is less than the upper threshold (550). The upper threshold value may be, for example, a second maximum similarity degree having the largest value among the similarities based on the size of the overlap region between the registered images. If it is determined in step 550 that the maximum similarity is smaller than the upper threshold, the updating apparatus can update the registration database by replacing the replacement image with the input image (560). As a result of the determination in step 550, if the maximum similarity degree is greater than or equal to the upper threshold value, the update apparatus can maintain the replacement image in the registration database (570).

FIG. 7 is a flowchart illustrating a method of selecting a replacement image according to an exemplary embodiment. According to an exemplary embodiment, the update apparatus may firstly select candidate images to duplicate and store the largest amount of information, and finally select a candidate image having more information redundant with the remaining registered images of the candidate images as a replacement image have.

Referring to FIG. 7, the update apparatus may select candidate images to be replaced among the registered images based on the similarity based on the size of the overlap region between the registered images (610). In one embodiment, the registered images having high similarity are selected as candidate images to be replaced because the duplicated registered images are eliminated because there is no need to duplicate substantially the same or almost similar registered images in the registration database.

The update apparatus can find, for example, registered images corresponding to the maximum similarity degree among the similarities based on the size of the overlap region between registered images. The update apparatus can select registered images corresponding to the maximum similarity as candidate images to be replaced. Since the similarity based on the size of the overlapping region is calculated between two different images, the candidate images corresponding to the maximum similarity may be two registered images.

The update device may accumulate the similarity between each of the candidate images selected in step 610 and the remaining registered images (620). Here, the 'remaining registered images' can be understood as the remaining registered images excluding the candidate images to be replaced among the registered images. For example, the updating apparatus can calculate the cumulative similarity between the candidate image and the remaining registered images using Equation (4).

Here, s _overlap (i, j) is an i-th registered image and a j- . The jth registered image may be one of the two candidate images. M is the number of registered images. According to another embodiment, when i = j, s _overlap (i, j) is calculated to be the same value (for example, maximum value) irrespective of j, The condition may be omitted.

For example, suppose that the number of registered images is 18, and the candidate image to be replaced is registered image 13 and registered image 18. The updating device may calculate the accumulated similarity s (13) between the registered image 13 and the remaining registered images and calculate the accumulated similarity s (18) between the registered image 18 and the remaining registered images.

The update apparatus may select a candidate image having a high degree of similarity among the candidate images as a replacement image (630). The updating apparatus can select a candidate image having a high value among the cumulative similarity s (13) and the accumulated similarity s (18) as a replacement image.

8 and 9 are views for explaining a method of selecting a replacement image according to an embodiment. 8 and 9, there is shown a graph 730, 810 of projecting the similarity between registered images 710 and registered images 710 onto a three-dimensional matrix. In the graphs 730 and 810, the X axis and the Y axis represent the indexes of the registered images, and the Z axis represents the similarity between the registered images.

For example, it is assumed that the degree of similarity between the registered image 713 and the registered image 716 among the registered images 710 is the maximum similarity. The update apparatus can find registered images having the maximum similarity by finding registration images corresponding to the points 815 shown in the graph 810 of FIG.

The graph 830 may represent the cumulative similarity of the candidate images. The graph 830 may be a one-dimensional projected result of the graph 810 of FIG. 9 projected onto a three-dimensional matrix. In the graph 830, the X-axis represents the index of the registered images, and the Y-axis represents the cumulative similarity.

For example, when the index of the candidate image is 13, 18, the updating device calculates the accumulated similarity s (18) corresponding to the registered image 18 and the accumulated similarity s (13) corresponding to the registered image 13, Can be compared. The 'higher accumulated similarity' means more overlap with other registered images, so more duplicated registered images can be selected as replacement images. The update apparatus can select the registered image No. 18 having the higher cumulative similarity as the replacement image. The registered image No. 13 is more likely to cover the new area than the registered image No. 18, so it can be maintained in the registration database.

10 is a flowchart illustrating a method of updating a registration database according to an embodiment.

Referring to FIG. 10, an update apparatus according to an exemplary embodiment receives an input image (910) and extracts an effective image from an input image (920). The updating device may determine 930 whether the size of the valid image is larger than a preset first threshold value. If it is determined in operation 930 that the size of the valid image is larger than a predetermined first threshold value, the updating apparatus may calculate a verification score (940). If the size of the effective image is smaller than or equal to a predetermined first threshold value, the updating device rejects the authentication and provides the user with a message to guide the user to newly input the input image to receive a new input image (910).

In step 940, the updating device may calculate the authentication score in various manners. For example, the updating apparatus may calculate the authentication score considering the size of the overlapping area, or may calculate the authentication score without considering the size of the overlapping area. The update device can divide the input image into small blocks, calculate the score for each block, and calculate the authentication score by integrating the scores for each block.

The update device may determine in step 950 whether the authentication score is greater than a preset authentication threshold. The preset authentication threshold may be, for example, an authentication score between registered images. The updating device can perform user authentication on the input image by comparing the authentication score and the authentication threshold. As a result of the determination in step 950, if the authentication score is less than or equal to a preset authentication threshold, the updating apparatus can determine that the user authentication is failed by the input image. As a result of the determination in step 950, if the authentication score is greater than a preset authentication threshold, the update device may accept the user authentication. It can be considered to update the registered image with respect to the input image for which authentication has been accepted.

The updating device may determine the maximum similarity Smax based on the similarities calculated between the input image and the registered images in which the user authentication is accepted (970). The similarities calculated in step 970 may be similarities based on the size of the overlap region. Hereinafter, an input image can be understood as an input image in which user authentication is accepted.

The update device may determine 980 whether the maximum similarity determined in step 970 is greater than the lower threshold and less than the upper threshold. The upper threshold value may be the maximum similarity degree based on the size of the overlap region between the registered images included in the registration database 905. [ For example, the updating apparatus can detect a registered image pair having the highest degree of similarity between registered images and define the similarity of detected registered image pairs as the second maximum similarity. The second maximum similarity may be used as an upper threshold. In one embodiment, when the input image and the registered images overlap with each other over a predetermined area, that is, when the maximum similarity is greater than the lower threshold, the input image is registered in the registration database to prevent the outlier from being reflected in the registration fingerprint have.

If the maximum similarity is greater than the lower threshold and less than the upper threshold, the updating device may update the registration database 905 at step 960. The updating device may add the input image to the registration database 905. [ Alternatively, it may be replaced with any one of registered images of the registration database 905. [ In this case, the updating device may select a replacement image among the registered images previously registered in the registration database 905 based on the similarities based on the size of the overlap area between the registered images (960). If the maximum similarity is less than the lower threshold or greater than the upper threshold, then the updating device may not update the registration database 905 at step 960.

A technique of adaptively updating the registration database according to the time required for authentication

11 is a view for explaining an authentication process of biometrics according to an embodiment. Referring to FIG. 11, the authentication process of the biometrics according to an exemplary embodiment may include an authentication process and an adaptive update process.

In the authentication step, authentication of the input image can be performed. For example, in the authentication step, the registered images included in the registration database and the input image are compared, and a score can be calculated. If the calculated score is greater than the predetermined authentication threshold, the authentication is successful, otherwise the authentication may fail.

In the adaptive update step, operations for updating the registration database can be performed. For example, in the adaptive update step, a technique of updating the registration database using the similarity based on the size of the overlap area described with reference to FIGS. 1 to 10 may be performed.

Although not shown in the figure, the authentication process of the biometrics may further include a pre-process step before the authentication step. The preprocessing operations for authentication of the input image may be performed in the preprocessing step. The preprocessing operations may include an operation of enhancing an input image, an operation of detecting an effective area in an input image, and the like.

12 is a diagram illustrating an adaptive update of a registration database according to an embodiment. Referring to FIG. 12, the authentication process of the biometrics may have a maximum allowable time limit (hereinafter referred to as a maximum allowed time). If the maximum allowed time is determined, the authentication phase and the adaptive update phase should be performed within the maximum allowed time.

According to an embodiment, a technique of performing authentication by reducing a search candidate group through a multi-stage authentication technique may be applied to the authentication step. The multi-stage authentication scheme may be composed of two steps of coarse search and fine search, or may be composed of three or more steps. In this case, the search candidates are reduced through a minimum operation in each stage, and precise matching can be performed in the final stage.

The multi-stage authentication technique can stop the authentication or search operation and determine an early accept if the operation result is a certain score or more even in the stage before the final stage. Thus, the authentication speed can be improved while ensuring the authentication accuracy.

When the multi-stage authentication technique is used, the time required in the authentication step may differ from case to case. For example, the time required for authentication in Case A may be longer than the time required for authentication in Case D.

If the maximum allowed time is specified, the longer the time required for authentication, the shorter the time allowed for adaptive updating. For example, the time allowed for adaptive update in case A may be shorter than the time allowed for adaptive update in case D.

Embodiments can provide a technique for selecting the best registration database update technique by predicting the time spent in the authentication step and predicting the future available time. Embodiments can ensure authentication performance while minimizing authentication time.

For example, when the registration database is updated only when the time for updating the registration database is sufficiently secured as in Case D, the performance of the fingerprint recognizer may not be improved. This is because the input image accepted early is likely to be very similar to the registered image. As described with reference to Fig. 2, the recognition rate of the fingerprint recognizer increases in proportion to the registered area. Since the input image, which is very similar to the registered image, contains little information other than the information contained in the registered image, the registered area is not increased even if the input image is added to the registration database.

The embodiments can be applied not only in cases where the time for updating the registration database is sufficiently secured as in Case D but also in the case where the time for updating the registration database is not sufficient as in Case A, Quot; is updated.

13 is an operation flowchart showing a method of adaptively updating a registration database according to an embodiment. Referring to FIG. 13, an updating apparatus according to an embodiment collects biometric data in step 1310. For example, the updating device may receive an input image including fingerprint information from a fingerprint sensor. The update device performs authentication at step 1320. [ Optionally, the update device may perform pre-processing prior to step 1320. [

If it is determined in step 1330 that the authentication is acceptable, the updating device can estimate the time taken for authentication of the input image in step 1340. For example, the updating device can measure the number of clocks required for authentication of an input image using a timer, and estimate the time required for authentication of the input image from the measured clock number. In addition, the updating device can estimate the usable time for the adaptive update from the estimated time.

The update device may adaptively update the registration database based on the estimated time in step 1350. [ The update device may control the amount of computation for the adaptive update step so that both the authentication and adaptive update steps can be performed within the maximum allowed time.

14 is a block diagram illustrating a multi-stage authenticator 1400 in accordance with one embodiment. Referring to FIG. 14, a multi-stage authenticator 1400 according to one embodiment includes a plurality of authenticators corresponding to a plurality of stages. For example, the multi-stage authenticator 1400 may include a first authenticator 1410 and a second authenticator 1420. Although not shown in the figure, the multi-stage authenticator 1400 may include three or more stages.

The first authenticator 1410 may perform authentication in a coarse search manner. The first authenticator 1410 may compare the input image and the registered images with each other through a search method, and determine whether a matching score equal to or greater than the first authentication threshold is derived. The first authenticator 1410 may accept the authentication and terminate the authentication operation in accordance with the determination that a matching score equal to or greater than the first authentication threshold has been derived. In this case, an acceptance signal can be output at time t ₁ .

The first authenticator 1410 may pass the reduced search space to the second authenticator 1420 if a match score of a predetermined threshold or more is not derived while performing the coarse search. For example, the first authenticator 1410 extracts a predetermined number of candidates from the registered images included in the registration database in order of excellent matching scores, and transmits information related to the extracted candidate group to the second authenticator 1420 .

15 is a diagram illustrating a coarse search according to an embodiment. Referring to FIG. 15, the first authenticator may calculate matching scores by comparing input images and registered images on a one-to-one basis. As an example, the first authenticator may use the normalized cross-correlation value of Equation (1) as a matching score. As another example, the first authenticator may use the similarity based on the size of the overlap region of Equation (3) as a matching score.

Referring again to FIG. 14, the second authenticator 1420 may perform authentication in a fine search manner in a reduced search space. The second authenticator may compare the input image and the registered images in a fine search manner to determine whether a matching score above the second authentication threshold is derived. The second authenticator 1420 may accept the authentication and terminate the authentication operation upon determining that a matching score that is greater than or equal to the second authentication threshold has been derived. In this case, the acceptance signal may be output at time t _2. Time t ₂ is a time later than the time t _1.

If the second authenticator 1420 determines that a matching score above the second authentication threshold is not derived in the reduced search space, the second authenticator 1420 may reject the authentication and terminate the authentication operation.

16 is a view for explaining a fine search according to an embodiment. Referring to FIG. 16, the second authenticator may divide the input image into a plurality of blocks. The second authenticator can calculate the block matching score through block pattern matching. The second authenticator may determine whether the input image is authenticated by using the sum K of the block matching scores or the average value.

17 to 22 are diagrams illustrating an update apparatus for adaptively updating a registration database according to an embodiment. 17, an update device according to one embodiment includes an authentication module 1710, an update module 1720, a time estimation module 1730, and a registration database 1740. [

The authentication module 1710 authenticates the input image and outputs an acceptance signal or a rejection signal. The time estimation module 1730 estimates the time required for authentication of the input image in response to the acceptance signal or the time available for the adaptive update. The update module 1720 updates the registration database 1740. The update module 1720 may control the amount of computation to update the registration database 1740 based on the time estimated by the time estimation module 1730. [

The authentication module 1710 may include a candidate selector 1711, a first authenticator 1712, and a second authenticator 1713. Referring to FIG. 18, the candidate selector 1711 receives an input image. The input image may be the result of preprocessing. The candidate selector 1711 can extract the first candidate group among the registered images included in the registration database 1740. The candidate selector 1711 can extract the first candidate group in various ways.

The registration database 1740 stores 20 registered images corresponding to the first identifier id1, 20 registered images corresponding to the second identifier id2, and 20 registered images corresponding to the third identifier id3 Lt; / RTI > The first identifier, the second identifier, and the third identifier may correspond to different fingers of the same user, respectively. The 60 registered images may be partial images including a part of the fingerprint of the corresponding finger.

The candidate selector 1711 can extract the first candidate group by processing the input image and the registered images in the frequency domain. The candidate selector 1711 may generate a third image indicating the phase correlation between the input image and the registered image in the Fourier space and may calculate the variance of the phase correlation degree in the third image. If the calculated variance is greater than a predetermined threshold value, the candidate selector 1711 may classify the registered image as a first candidate group.

The candidate selector 1711 may communicate information associated with the first candidate group to the first authenticator 1712. As will be described in detail below, the results computed in the authentication operation can be utilized to minimize the amount of computation in the adaptive update operation. The candidate selector 1711 may store the first candidate corpus-related information in advance in the provided memory 1750 for future adaptive update operations.

Referring to FIG. 19A, the first authenticator 1712 receives information related to the first candidate group. The first authenticator 1712 may authenticate the input image. The operation described with reference to FIG. 15 may be applied to the first authenticator 1712. However, instead of using all 60 registered images in the registration database 1740 to authenticate the input image, the first authenticator 1712 can use only the first candidate group.

The first authenticator 1712 outputs an acceptance signal when a matching score equal to or greater than the first authentication threshold is derived during authentication. The output acceptance signal is transmitted to the time estimation module 1730. The results computed by the first authenticator 1712 may be stored in memory 1750 for future adaptive update operations.

If no matching score above the first authentication threshold is derived, the first authenticator 1712 may extract a second candidate group corresponding to a predetermined number of higher matching scores. The first authenticator 1712 may forward information associated with the second candidate group to the second authenticator 1713. The information associated with the second candidate group and the results computed by the first authenticator 1712 may be stored in memory 1750 for future adaptive update operations.

Referring to FIG. 19B, matching information of the registration unit and the number of operations or time until authentication can be stored in the memory 1750. s is ncc according to Equation 1 or s _{overlap according} to Equation (3), and area can be the shared area ratio or the size of the overlap region according to Equation (2). For example, the information stored in the first row and the first column of the in-memory table may be a result of comparing the registered image having the index 1 of the registered images corresponding to the thumb with the identifier and the input image.

Referring to FIG. 20, the second authenticator 1713 receives information related to the second candidate group. The second authenticator 1713 can authenticate the input image. The operation described with reference to FIG. 16 may be applied to the second authenticator 1713. The second authenticator 1713 outputs an acceptance signal when a matching score equal to or higher than the second authentication threshold is derived during authentication. The output acceptance signal is transmitted to the time estimation module 1730. The results computed by the second authenticator 1713 may be stored in memory 1750 for future adaptive update operations. If no matching score above the first authentication threshold is derived, the first authenticator 1712 outputs a reject signal.

Referring to FIG. 21, the time estimation module 1730 receives an acceptance signal. The time estimation module 1730 may estimate the time taken for authentication of the input image or the available time for adaptive update in response to the acceptance signal. The update module 1720 may update the registration database 1740 based on the time estimated by the input image and time estimation module 1730. [

For example, the update module 1720 may calculate the maximum similarity based on the size of the overlap region between the input image and the registered image. At this time, instead of determining all the similarities between the 60 registered images and the input image included in the registration database 1740, the update module 1720 updates only the similarities between the 20 registered images corresponding to the accepted identifiers and the input image Can be determined. When adding an input image to a registration database, 20 times of score calculation is required. When an input image is replaced with one of 20 registered images, 19 scoring operations are required except for a scoring operation with a registered image to be replaced .

At this time, the update module 1720 can utilize the pre-computed results stored in the memory 1750. 22, the update module 1720 uses the preliminary computation results stored in the memory 1750 to determine similarities 2000 between the 20 registered images corresponding to the accepted identifier and the input image A portion 2210 of similarities can be determined.

The number of score operations for determining whether or not the update module 1720 is updated is 20 or 19 times. If the score between the input image and the registered image has already been calculated by the authentication module 1710 and the calculation result is stored in the memory 1750, the update module 1720 can reduce the amount of computation. For example, if the first similarity between the first registered image and the input image has already been calculated by the first authenticator, the updating module 1720 may obtain the first similarity from the memory 1750. [ The above-described operation may be referred to as a score reuse method.

The update module 1720 may calculate similarities 2220 that have not yet been determined with best effort based on the time estimated by the time estimation module 1730. [ For example, the first authenticator 1712 may calculate registration information. The registration information may include a motion vector and an angle of rotation through image matching. The first authenticator 1712 may calculate the normalized cross-correlation with the overlap area ratio by rotating and moving the image.

If there is no similar overlap area in the registration step, the first authenticator 1712 may skip the subsequent calculation process and proceed to the next registered image. The average authentication time is reduced through the intermediate termination operation, but there is no score for the intermediate terminated registration image. In this case, the update module 1720 must directly calculate the similarity. At this time, the update module 1720 can calculate the similarities to the extent that time is allowed within the maximum allowable time.

If there is insufficient time to calculate all the similarities, the updating module 1720 may set the remaining similarities 2230 to a predetermined minimum value. The predetermined minimum value may be a value smaller than the lower threshold value shown in Figs. The operation described above may be referred to as a selective score reuse method.

The update module 1720 may process the registered images in the index order or random order of the registered images to calculate the similarities 2000 for a given time. Alternatively, the update module 1720 may preferentially process the registered images selected for the first candidate group or the second candidate group.

The operation of the first authenticator 1712 may be summarized by the algorithm of Table 1.

Rough navigation
A. Registration (movement vector and rotation angle detection)
-> If the detection is difficult, skip the corresponding image
B. Score operation by weighting ncc and overlap area
C. If the score is above a certain level, the acceptance declaration and authentication operation ends.
D. Repeat steps A to C for the next image if it is not the last one.
E. End if last video

The operation of the second authenticator 1713 may be summarized by the algorithm of Table 2. [

Fine search
A. Resurfacing in blocks
B. Block Score Operation
C. If the block score is above a certain level, the acceptance declaration and authentication operation ends.
D. Repeat steps A to C for the next image if it is not the last one.
E. End if last video

Scoring
A. Sort block scores calculated in fine search
B. Top K Average Score Operations
C. Determine acceptance or rejection by comparing average score and authentication threshold
D. Exit

The operation of the time estimation module 1730 may be summarized by the algorithm of Table 3. [

1. Identify the steps that were accepted and how many images were processed
2. Estimate the time required for authentication using the unit processing time (T1) in the first authentication step and the unit processing time (T2) in the second authentication step
(For example, assuming that the recognizer with M registered images has accepted the result of the third operation of the second stage of authentication, it takes time T1 * M + T2 * 3)
3. Estimate the number of recyclable scores and estimate the time available for future calculations considering the maximum allowed time

When the input image is replaced with one of the registered images, the operation of the update module 1720 can be summarized by the algorithm of Table 4. [

1. Reuse if there is a pre-computed result by the authentication module.
2. If there is no pre-computed result by the authentication module, review the remaining time and calculate the score corresponding to the allowed time.
- Estimate the estimated time in the time estimation module and estimate the maximum number of scoring operations allowed
- Calculate the score by comparing the allowable number of maximum score operation with the required number of score operation
3. If there is no remaining time, the score for the remaining images is set to a predetermined minimum value (for example, 0.3)
4. Repeat steps 1 ~ 3 for all image pairs except input image and replacement candidate
5. When replacing the replacement candidate with the input image based on the calculated scores, if the maximum score is lowered, replace the replacement candidate and the input image, and update the score matrix storing the score between the previously registered images
6. Exit

23 is an operational flowchart illustrating a method for adaptively updating a registration database according to an embodiment. Referring to FIG. 23, in step 2311 of the authentication step 2310, the authentication score s is calculated. If it is determined that the authentication score s calculated in step 2312 is greater than the authentication threshold Th1, the authentication is accepted. If authentication is accepted, pre-computed results for authentication are stored in memory 2330.

The area score S 'is updated within the time limit in steps 2321 and 2322 of the update step 2320. [ In step 2321, an individual area score s 'for calculating the area score S' may be determined. If there is information previously stored in the memory 2330, the individual area score s' may be loaded from the memory 2330. For example, the individual region scores s' of the first registration image, the second registration image, and the 19th registration image corresponding to the thumb may be loaded from the memory 2330. If there is no pre-stored information in the memory 2330, the individual area score s' may be calculated based on the authentication score s and the overlap area. For example, the individual region scores of the third to the eighteen registered images corresponding to the thumb may be calculated by f (s, area). Where f (s, area) = s x area.

It may be determined in step 2322 whether the current time t is within the T _{limit time} . If the current time t is within the T _{limit time} , the process returns to step 2321 and the individual area score s' may be calculated. If the current time t has passed the T _{limit time} , the individual area scores that have not yet been calculated can be determined to a predetermined value. For example, the individual area score of the 20th registered image corresponding to the thumb may be set to a predetermined minimum value (for example, 0.1).

There is whether it can be determined that the area between the score (S ') in step (2323) a predetermined first threshold value (Th _{AREA _ 1)} and a second threshold (Th _{AREA_2).} The area score S 'can be calculated by Equation (5).

Here, M is a predetermined natural number, and sort _decend (s') represents the result of sorting individual region scores in descending order. According to Equation (5), the area score S 'can be calculated as an average of M individual area scores having a high value.

If M is 1, the area score S 'corresponds to the maximum similarity, and in this case, step 2323 may correspond directly to step 980 of FIG.

If it is determined that the area between the score (S ') in step (2323) a predetermined first threshold value (Th _{AREA _ 1)} and a second threshold (Th _{AREA_2),} the registration database is updated. For example, if the area score S 'satisfies the condition of step 2323, a registered image may be added to the registration database.

For example, the number of registered images included in the registration database may be changed to the following scenarios.

1. When registering the first user, register 18 images in the registration database

2. In the process of using fingerprint authentication, add a new fingerprint image that satisfies the update condition of the registration database to the registration database and register up to 25 registered images

3. When the number of registered images is 25, new fingerprint images satisfying the renewal conditions of the registration database are replaced with previously registered images, and the registered images are retained in 25

The operation of removing the image registered in the registration database is disabled to add the fingerprint image to the registration database or the operation of removing the image registered in the registration database in order to replace the fingerprint image with the previously registered image may be activated have.

A first threshold value (Th _{_ AREA 1)} is the second threshold value (Th _{AREA_2)} and corresponds to the upper threshold value in step 2323 corresponds to the lower threshold. Here, the upper threshold value may be a score of the replacement candidate having the most redundant information among the previously registered images. Although not shown in the drawing, an upper threshold value may not be considered when a fingerprint image is added to the registration database. In this case, only whether or not the area score S 'is larger than the lower threshold value is determined in step 2323, and if the area score S' is larger than the lower threshold value, the corresponding fingerprint image can be added to the registration database.

24 is a block diagram of an apparatus for updating a registration database according to an embodiment.

24, an updating apparatus according to an embodiment includes a sensor 2410, a memory 2420, a processor 2430, and a receiver 2440. [ The update device may further include a template manager 2450. [ The sensor 2410, the memory 2420, the processor 2430, the receiver 2440, and the template manager 2450 can communicate with each other via the bus 2460.

The sensor 2410 may be the fingerprint sensor 110 shown in Fig. The sensor 2410 may capture a fingerprint image in a well-known manner (e.g., a manner of converting an optical image into an electrical signal, etc.). The image is output to the processor 2430 via the receiver 2440.

As another example, the sensor 2410 may include a sensor that recognizes the vein pattern of the user. The sensor 2410 can extract a vein pattern from the skin of the user's hand. The sensor 2410 can obtain an image including a vein pattern after maximizing the brightness contrast of the blood vessel to the skin using infrared light and a filter. The processor 2430 can update the registration database using the input image corresponding to the vein pattern and the images of the vein pattern previously registered.

Referring to FIG. 25, the technique of authenticating the fingerprint pattern through FIGS. 1 to 23 can be directly applied to the technique of authenticating the vein pattern. For example, the user's smart watch 2510 may sense the user's vein pattern 2520. The input image 2530 generated by sensing the user's vein pattern 2520 may be authenticated via coarse search 2550 and fine search 2580. The candidate group 2560 of the registered images included in the result registration database 2540 of the rough search 2550 may be extracted. The input image 2530 may be blocked 2570 for the fine search 2580. [

As another example, the sensor 2410 may include a sensor that recognizes a user's iris pattern. The sensor 2410 may scan or capture an iris pattern between the pupil of the user and the sclera (the white area of the eye). The processor 2430 can update the registration database using the input image corresponding to the iris pattern and the images of the previously registered iris pattern.

The memory 2420 stores a registration database including pre-registered images. The memory 2420 may store an input image including biometric information captured by the sensor 2410, a newly registered input image, and the like. The memory 2420 may be a volatile memory or a non-volatile memory.

The processor 2430 determines the maximum similarity based on the size of the overlap region between the input image and the registered images. The processor 2430 registers the input image in the registration database according to the result of the comparison of the upper threshold value based on the maximum similarity degree and the size of the overlap region between the registered images. The processor 2430 may calculate the similarities between the input image and the registered images based on the size of the overlap region between the input image and the registered images and determine the similarity having the largest value among the similarities as the maximum similarity.

The processor 2430 may calculate the degree of correlation in the overlapping region between any one of the registered images and the input image and calculate the size of the overlapping region between any one registered image and the input image. The processor 2430 can calculate the degree of similarity between any one registered image and the input image using the degree of correlation and the size of the overlapping region. In addition, the processor 2430 may perform at least one of the methods described above with respect to Figures 1-10.

The processor 2430 can execute the program and control the update device. The program code executed by the processor 2430 may be stored in the memory 2420. [ The update device is connected to an external device (e.g., a personal computer or a network) through an input / output device (not shown) and can exchange data.

The receiving unit 2440 receives an input image including biometric information sensed by the sensor 2410.

The template manager 2450 can select a replacement image among the registered images based on the similarities based on the size of the overlap area between the registered images. The template manager 2450 can select candidate images to be replaced among the registered images based on the similarity based on the size of the overlap area between the registered images. The template manager 2450 accumulates the similarities between each of the candidate images and the remaining registered images, and selects a candidate image having a high degree of similarity accumulated among the candidate images as a replacement image. According to the embodiment, the template manager 2450 is not constituted as a separate element as shown in Fig. 24, and the operation of the template manager 2450 may be performed by the processor 2430. Fig.

The update device can be used in a variety of electronic systems such as mobile devices, smart phones, PDAs, tablet computers, mobile devices such as laptop computers, computing devices such as personal computers, tablet computers, netbooks, or electronic devices such as televisions, smart televisions, Lt; / RTI >

The embodiments described above may be implemented in hardware components, software components, and / or a combination of hardware components and software components. For example, the devices, methods, and components described in the embodiments may be implemented within a computer system, such as, for example, a processor, a controller, an arithmetic logic unit (ALU), a digital signal processor, such as an array, a programmable logic unit (PLU), a microprocessor, or any other device capable of executing and responding to instructions. The processing device may execute an operating system (OS) and one or more software applications running on the operating system. The processing device may also access, store, manipulate, process, and generate data in response to execution of the software. For ease of understanding, the processing apparatus may be described as being used singly, but those skilled in the art will recognize that the processing apparatus may have a plurality of processing elements and / As shown in FIG. For example, the processing unit may comprise a plurality of processors or one processor and one controller. Other processing configurations are also possible, such as a parallel processor.

The software may include a computer program, code, instructions, or a combination of one or more of the foregoing, and may be configured to configure the processing device to operate as desired or to process it collectively or collectively Device can be commanded. The software and / or data may be in the form of any type of machine, component, physical device, virtual equipment, computer storage media, or device , Or may be permanently or temporarily embodied in a transmitted signal wave. The software may be distributed over a networked computer system and stored or executed in a distributed manner. The software and data may be stored on one or more computer readable recording media.

The method according to an embodiment may be implemented in the form of a program command that can be executed through various computer means and recorded in a computer-readable medium. The computer-readable medium may include program instructions, data files, data structures, and the like, alone or in combination. The program instructions to be recorded on the medium may be those specially designed and configured for the embodiments or may be available to those skilled in the art of computer software. Examples of computer-readable media include magnetic media such as hard disks, floppy disks and magnetic tape; optical media such as CD-ROMs and DVDs; magnetic media such as floppy disks; Magneto-optical media, and hardware devices specifically configured to store and execute program instructions such as ROM, RAM, flash memory, and the like. Examples of program instructions include machine language code such as those produced by a compiler, as well as high-level language code that can be executed by a computer using an interpreter or the like. 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.

Although the preferred embodiments of the present invention have been disclosed for illustrative purposes, those skilled in the art will appreciate that various modifications, additions and substitutions are possible, without departing from the scope and spirit of the invention as disclosed in the accompanying claims. For example, it is to be understood that the techniques described may be performed in a different order than the described methods, and / or that components of the described systems, structures, devices, circuits, Lt; / RTI > or equivalents, even if it is replaced or replaced. Therefore, other implementations, other embodiments, and equivalents to the claims are also within the scope of the following claims.

Claims (32)

A method for updating a registration database for biometric authentication,
Authenticating an input image using the registration database; And
Adaptively updating the registration database based on a first time required for the authentication of the input image and a result obtained by performing a pre-calculation for authentication of the input image
And updating the registration database. The method according to claim 1,
The step of adaptively updating the registration database
Determining similarities based on sizes of overlapping regions between registered images included in the registration database and the input image; And
Determining whether to register the input image in the registration database, based on the similarities;
And updating the registration database. 3. The method of claim 2,
The step of determining the similarities
Determining at least some of the similarities using the pre-calculated result; And
Determining at least some of the similarities to a predetermined minimum value to satisfy the deadline based on the first time,
The method comprising the steps of: 3. The method of claim 2,
The step of determining whether to register in the registration database
Determining whether the maximum similarity among the similarities is less than or equal to an upper threshold value; And
Determining whether the maximum similarity is greater than or equal to a lower threshold value
The method comprising the steps of: 5. The method of claim 4,
The upper threshold
Corresponding to the maximum similarity determined based on the overlapping area between the registered images. The method according to claim 1,
The pre-calculated result is
Overlapping areas between the input image and the registered images, and sizes of overlapping areas between the input image and the registered images. The method according to claim 1,
The step of authenticating the input image
Calculating a size of an overlap area between the input image and a registered image included in the registration database;
Calculating an intra-overlap correlation between the input image and the registered image;
Outputting an accept signal when the correlations are above a threshold;
Storing the size of the overlapping area and the degree of correlation in a preset storage space corresponding to the registered image
And updating the registration database. The method according to claim 1,
The step of authenticating the input image
Calculating a size of an overlap area between the input image and a registered image included in the registration database;
Calculating an intra-overlap correlation between the input image and the registered image;
Outputting an accept signal when the correlations are above a threshold;
Storing the size of the overlapping area and the degree of correlation in a preset storage space corresponding to the registered image
And updating the registration database. The method according to claim 1,
The step of authenticating the input image
Authenticating the input image using a multi-stage authentication technique
And updating the registration database. The method according to claim 1,
The step of adaptively updating
Estimating a second time available for updating the registration database based on the first time; And
Adaptively updating the registration database based on the second time
And updating the registration database. 11. The method of claim 10,
The step of estimating the second time
Calculating the second time based on a difference between a third time allowed for the biometric authentication and the first time,
And updating the registration database. The method according to claim 1,
The step of adaptively updating the registration database
Adding the input image to the registration database; And
Replacing any one of the registered images included in the registration database with the input image
The method comprising the steps of: 13. The method of claim 12,
The replacing step
Selecting a replacement image based on the similarities based on the size of the overlap area between the registered images
And updating the registration database. A method for updating a registration database for biometric authentication,
Receiving an input image including biometric information;
Determining a maximum similarity based on the size of an overlapped area between the input image and the registered images;
Comparing an upper threshold based on the maximum similarity and the size of the overlap region between the registered images; And
Registering the input image in the registration database according to the comparison result
And updating the registration database. 15. The method of claim 14,
The step of determining the maximum similarity comprises:
Calculating similarities between the input image and the registered images based on sizes of overlapping regions between the input image and the registered images; And
Determining the similarity having the largest value among the similarities as the maximum similarity
And updating the registration database. 16. The method of claim 15,
The step of calculating the similarities
Calculating an intra-overlap correlation between any one of the registered images and the input image;
Calculating a size of an overlap region between any one of the registered images and the input image; And
Calculating a degree of similarity between any one of the registered images and the input image using the degree of correlation and the size of the overlap region,
And updating the registration database. 16. The method of claim 15,
The step of calculating the similarities
Calculating a normalized cross correlation (NCC) value using a motion vector and a rotation angle calculated by image registration between any one of the registered images and the input image;
Calculating a share area ratio between any one of the registered images and the input image; And
Calculating a degree of similarity between any one registered image and the input image using the normalized cross-correlation value and the shared area ratio
And updating the registration database. 15. The method of claim 14,
Selecting a replacement image among the registered images based on the similarities based on the size of the overlap area between the registered images
Further comprising the step of updating the registration database. 19. The method of claim 18,
The step of selecting the replacement image
Selecting candidate images to be replaced among the registered images based on the similarity based on the size of the overlap region between the registered images;
Accumulating similarity between each of the candidate images and the remaining registered images; And
Selecting a candidate image having a high degree of similarity among the candidate images as the replacement image
And updating the registration database. 19. The method of claim 18,
The step of registering the input image in the registration database
Updating the registration database by replacing the replacement image with the input image if the maximum similarity is smaller than the upper threshold; And
If the maximum similarity is greater than the upper threshold, maintaining the replacement image in the registration database
And updating the registration database. 15. The method of claim 14,
The upper threshold
And a second maximum similarity of similarities based on the size of the overlap region between the registered images. 15. The method of claim 14,
The step of registering the input image in the registration database
Comparing the maximum similarity with a preset lower threshold; And
Registering the input image in the registration database based on a result of the comparison with the lower threshold value
And updating the registration database. 23. The method of claim 22,
And registering the input image in the registration database based on a result of the comparison with the lower threshold value
Registering the input image in the database if the maximum similarity is greater than the lower threshold and less than the upper threshold;
And updating the registration database. 15. The method of claim 14,
Calculating an authentication score between the input image and the registered images; And
Comparing the authentication score with an authentication threshold to perform user authentication on the input image
Further comprising:
The step of determining the maximum similarity comprises:
Determining a maximum similarity based on the similarities calculated between the input image and the registered images in which the user authentication is accepted,
And updating the registration database. 15. The method of claim 14,
The biometric information
Fingerprint information, blood vessel information, and iris information. 26. A computer program stored in a medium for executing the method of any one of claims 1 to 25 in combination with hardware. Receiving an input image including biometric information;
Comparing the input image with registered images included in a database and authenticating the registered image; And
An update step of adding the input image to the database or determining whether to replace the input image with any one of the registered images included in the database
Lt; / RTI >
Wherein the sum of the time required for the authentication step and the time required for the update step is less than or equal to a preset maximum time. 28. The method of claim 27,
Wherein the biometric information is fingerprint information. 28. The method of claim 27,
The updating step
Determining similarities based on sizes of overlapping regions between the registered images included in the database and the input image; And
Based on the similarities, determining whether to register the input image in the database
The database management method comprising: 30. The method of claim 29,
The step of determining the similarities
Determining at least some of the similarities using the pre-computed results for authentication of the input image; And
Determining at least some of the similarities as a predetermined minimum value to satisfy a deadline based on the time required for the authentication step
The database management method comprising: 30. The method of claim 29,
The step of determining whether to register in the database
Determining whether the maximum similarity among the similarities is less than or equal to an upper threshold value; And
Determining whether the maximum similarity is greater than or equal to a lower threshold value
The database management method comprising: 28. The method of claim 27,
The updating step
Selecting a replacement image based on the similarities based on the size of the overlap area between the registered images
The database management method comprising:

Images