KR101959892B1 - Fingerprint authentication method and apparatus - Google Patents

Abstract

According to an embodiment of the present invention, there is provided a method of generating a fingerprint image, comprising the steps of: (a) acquiring feature points based on a change in a grayscale value of a specific pixel and a neighboring pixel on a fingerprint image acquired through a fingerprint sensor; (b) comparing the feature points with previously registered feature points; (c) comparing the authentication gain value applied to the output value from the fingerprint sensor with the registered gain value applied for fingerprint image registration when the similarity degree is determined to be equal to or greater than the threshold value; And (d) determining that the fingerprint authentication fails if the difference between the authentication gain value and the registration gain value is out of the reference range.

Description

[0001] FINGERPRINT AUTHENTICATION METHOD AND APPARATUS [0002]

The present invention relates to a fingerprint authentication method and apparatus, and more particularly, to a fingerprint authentication method and apparatus that extract fingerprint variation points in a fingerprint image using feature points to perform fingerprint pattern authentication, The present invention relates to a method and apparatus for enabling accurate fingerprint authentication and forgery fingerprint detection by determining that fingerprint authentication is successful only when the output value from the pixels is within the reference range.

User authentication through fingerprint recognition is not only easy to use, but also excellent in security and economy, and is now widely used. However, even if the fingerprint of another person is copied by using a material such as silicone, rubber, or polyvinyl, and is used for authentication, a fingerprint authentication error may occur. Therefore, it is necessary to distinguish whether the contacted fingerprint is a biometric fingerprint of a real person or a counterfeit fingerprint by copying.

General methods for distinguishing between counterfeit fingerprints and biometric fingerprints include, but are not limited to, methods of analyzing the characteristics of fingerprint images, methods of using the capacitance of a living body, methods of measuring oxygen saturation of a finger by comparing infrared and visible light transmittances .

However, the method of analyzing the characteristics of the fingerprint image has an advantage that it can be implemented without an additional device, but it has a problem that it is not distinguished from the biometric fingerprint when the counterfeit fingerprint is precisely manufactured.

Particularly, a method using a capacitance difference between a capacitance of a living body and a counterfeit fingerprint makes it difficult to distinguish the counterfeit fingerprint because the difference in capacitance between the two is reduced if the counterfeit fingerprint is made thin.

In addition, there is a method of using the light reflection characteristic of the subject, but this requires an additional optical sensor and the like, and thus there has been an inconvenience to change the existing fingerprint sensor design.

Therefore, it is required to use a conventional fingerprint sensor as it is, but a counterfeit fingerprint detection method with improved accuracy.

SUMMARY OF THE INVENTION The present invention has been made to solve the above-mentioned problems of the prior art, and it is an object of the present invention to enable not only authentication of a fingerprint pattern but also detection of a forged fingerprint by using an existing fingerprint authentication device without a separate device.

Another object of the present invention is to enable fine fingerprint pattern authentication and counterfeit fingerprint detection by specifying minutiae in the fingerprint image as points where the change in the density value is above the threshold value.

It is still another object of the present invention to provide a fingerprint authentication system capable of detecting a fake fingerprint based on a gain value applied at the time of fingerprint registration and authentication and / or a result obtained by applying a predetermined gain value to an output value from the fingerprint sensor, .

According to an embodiment of the present invention, there is provided a method of processing a fingerprint image, the method comprising: (a) acquiring feature points based on a change in a grayscale value of a specific pixel and a neighboring pixel on a fingerprint image acquired through a fingerprint sensor; (b) comparing the feature points with previously registered feature points; (c) comparing the authentication gain value applied to the output value from the fingerprint sensor with the registered gain value applied for fingerprint image registration when the similarity degree is determined to be equal to or greater than the threshold value, ; And (d) determining that the fingerprint authentication fails if the difference between the authentication gain value and the registration gain value is out of the reference range.

The fingerprint authentication method of the fingerprint authentication device may further include: (e) when a difference between the authentication gain value and the registered gain value is within a reference range, a result obtained by applying a predetermined gain value to the output value from the fingerprint sensor, Comparing the result of applying the predetermined gain value to an output value from the sensor; And (f) determining that the fingerprint authentication is successful only when the difference between the fingerprint registration and the result of applying the predetermined gain value to the output value from the fingerprint sensor during authentication is within the reference range.

The predetermined gain value may be different from the registered gain value.

In the step (a), an average of pixel density values for a specific area centered on each pixel of the fingerprint image is calculated, and the pixel density value averages Calculating a sum and assigning the sum to each pixel; And a step of shifting a window having a specific width on the fingerprint image, and selecting pixels having change in density value among the pixels in each window equal to or greater than a threshold value as feature points.

An average calculation of the pixel grayscale values may be performed based on a summation array generated for the fingerprint image.

According to another embodiment of the present invention, there is provided an image processing method comprising the steps of: acquiring feature points based on a change in a shade value of a specific pixel and a surrounding pixel on a fingerprint image acquired through a fingerprint sensor; Comparing the feature points with previously registered feature points; Comparing a result obtained by applying a predetermined gain value to an output value from the fingerprint sensor and a result of applying the predetermined gain value to an output value from the fingerprint sensor when fingerprint registration is performed, when the similarity degree is determined to be equal to or greater than a threshold value; And determining that the fingerprint authentication is successful only when the difference between the fingerprint registration and the result of applying the predetermined gain value to the output value from the fingerprint sensor during authentication is within the reference range, / RTI >

The predetermined gain value may be a value different from the registered gain value applied to acquire the fingerprint image upon fingerprint registration.

According to another embodiment of the present invention, there is provided a fingerprint feature extraction apparatus comprising: a fingerprint feature point acquiring unit that acquires feature points based on a change in a shade value of a specific pixel and a surrounding pixel on a fingerprint image acquired through a fingerprint sensor; A fingerprint feature point comparing unit comparing the feature points with previously registered feature points; And when the similarity degree is determined to be equal to or greater than the threshold value, the authentication gain value applied to the output value from the fingerprint sensor for fingerprint image acquisition is determined as a fingerprint image acquisition And a gain value comparing unit which determines that the fingerprint authentication is successful only when the difference is within the reference range.

According to another embodiment of the present invention, there is provided a fingerprint feature extraction apparatus comprising: a fingerprint feature point acquiring unit that acquires feature points based on a change in a shade value of a specific pixel and a surrounding pixel on a fingerprint image acquired through a fingerprint sensor; A fingerprint feature point comparing unit comparing the feature points with previously registered feature points; And comparing the result of applying the predetermined gain value to the output value from the fingerprint sensor and the result of applying the predetermined gain value to the output value from the fingerprint sensor in fingerprint registration when the similarity degree is determined to be equal to or greater than the threshold, A fingerprint authentication device including a comparison / determination unit for determining that the fingerprint authentication is successful only when the difference is within the reference range.

According to the embodiment of the present invention, it is possible to detect not only the fingerprint pattern authentication but also the counterfeit fingerprint by using only the existing device for fingerprint authentication without a separate device.

According to the embodiment of the present invention, minutiae points in the fingerprint image are specified as points where the change in the grayscale value is equal to or more than the threshold value, thereby enabling detailed fingerprint pattern authentication and forgery fingerprint detection.

According to the embodiment of the present invention, even if the fingerprint pattern authentication is successful, by comparing the gain value applied at the time of fingerprint registration and the authentication value at the time of authentication and / or the result obtained by applying a predetermined gain value to the output value from the fingerprint sensor, It is possible to judge whether or not it is an attempt.

1 is a diagram showing a configuration of an electronic apparatus according to an embodiment of the present invention.
2 is a view showing an example of a fingerprint image generated based on a signal obtained by the fingerprint sensor of FIG.
FIG. 3 is a view for explaining a method for finding feature points in a fingerprint image in which a typical negative is not present according to an embodiment of the present invention.
4 is a diagram illustrating a method of performing first processing on a fingerprint image in a minutiae point extraction process according to an embodiment of the present invention.
5 is a diagram illustrating a method of performing second processing on a fingerprint image in a feature point extraction process according to an exemplary embodiment of the present invention.
6 is a view for explaining a method of selecting feature points for performing fingerprint recognition on the second processing image according to an embodiment.
7 is a view showing a fingerprint image obtained through a biometric fingerprint and a counterfeit fingerprint according to an exemplary embodiment of the present invention.
8 is a diagram showing the result of extracting feature points from the image shown in FIG.
9 is a flowchart illustrating a fingerprint registration and authentication process according to an embodiment of the present invention.
10 is a block diagram showing the configuration of a fingerprint registration and authentication apparatus according to an embodiment of the present invention.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, the present invention will be described with reference to the accompanying drawings. The present invention may, however, be embodied in many different forms and should not be construed as limited to the embodiments set forth herein. In order to clearly illustrate the present invention, parts not related to the description are omitted, and similar parts are denoted by like reference characters throughout the specification.

Throughout the specification, when a part is referred to as being "connected" to another part, it includes not only "directly connected" but also "indirectly connected" . Also, when an element is referred to as "comprising ", it means that it can include other elements, not excluding other elements unless specifically stated otherwise.

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

1 is a view showing an electronic apparatus equipped with a counterfeit fingerprint sensor according to an embodiment of the present invention.

Referring to FIG. 1, an electronic device 100 according to an exemplary embodiment includes a display unit 110 and a fingerprint sensor 120.

The display unit 110 displays information on the operation state of the electronic device 100. When the display unit 110 is implemented as a touch screen, the function of receiving a user's input may also be performed.

The fingerprint sensor 120 is formed in a partial area of the electronic device 100 and outputs a signal generated in relation to a user finger touching the surface in the form of an electrical signal. For example, the fingerprint sensor 120 can be driven by capacitive, optical, ultrasonic, or the like, and can output a signal generated in relation to the user's finger, and based on the signal, . ≪ / RTI > The fingerprint sensor 120 acquires a fingerprint image by an area method in which a fingerprint image is obtained in a state where a finger is in contact with a predetermined position or a sliding method in which a signal necessary for generating a fingerprint image is acquired in a process of moving the finger in a predetermined direction .

2 is a view showing an example of a fingerprint image generated based on a signal obtained by the fingerprint sensor of FIG.

2 (a), the fingerprint image is composed of ridges and valleys of the fingerprint. In order to recognize such fingerprint image, a fingerprint image such as a fingerprint bifurcation, a ridge end, A minutiae method that compares a previously registered fingerprint image with a new fingerprint image through a singular point of the fingerprint image matching method.

However, since the size of the fingerprint sensor is gradually reduced in electronic devices, the fingerprint image acquired by the fingerprint sensor is also reduced in size. Accordingly, fingerprint recognition must be possible even through a minimal fingerprint image as shown in FIG. 2 (b).

Referring to FIG. 2 (b), a minischer may not be included in a minimum fingerprint image. In the embodiment of the present invention, even in such a minimum fingerprint image, Lt; / RTI >

FIG. 3 is a view for explaining a method for finding feature points in a fingerprint image in which a typical negative is not present according to an embodiment of the present invention.

Referring to FIG. 3, a ridge is simply arranged in a specific direction in a fingerprint image, so that a minutiae such as a branch point or an end point may not exist. However, if we enlarge it, we can confirm that the ridges have a characteristic shape, not a uniform thickness. For example, a specific point of a ridge may protrude outward, a specific point of a ridge may be recessed inward, and a hole may be formed in the middle of the ridge by a pore.

Such features on the fingerprint image can be detected by detecting the difference between the change value of the tint in the fingerprint image and the average change value. That is, in FIG. 3, it can be confirmed whether or not the characteristic point exists based on how rapidly the density change value in the area where the features appear or in the vicinity thereof is changed compared to the average change value.

For example, the value of the change in density in the image of ridges arranged in a uniform thickness does not deviate from a certain range based on the average change value. However, when there is a protruding point or depression at the ridge, or when the pore is formed, The change value of the density change becomes large.

In the present invention, the feature point of the fingerprint refers to a point detected based on the change value of the tint in the region where the features appear or in the vicinity thereof.

4 is a diagram illustrating a method of performing first processing on a fingerprint image in a minutiae point extraction process according to an embodiment of the present invention.

4, when the fingerprint sensor 120 includes a plurality of fingerprint sensor elements arranged in a matrix form of NxM, the density value of each pixel constituting the fingerprint image is also temporarily stored in the memory in the form of a matrix of NxM type. Lt; / RTI > In FIG. 4, a matrix form in which a density value of each pixel is stored in an 8 × 8 size fingerprint image is illustrated, but it is needless to say that the size of the fingerprint image is not limited thereto.

The first processing is performed by matching the density value summing result for the pixels in the predetermined area including the corresponding pixel for each pixel. Specifically, a sum of all the pixels arranged in a rectangular area having a straight line extending from the pixel disposed as a reference pixel, preferably the pixel arranged in the first row and the first column to the pixel concerned is given to the pixel. Hereinafter, for convenience of explanation, the arrangement of numbers assigned to each pixel by performing the first processing will be referred to as a "summation array ".

For example, referring to FIG. 4, the first processing will be described in detail. The values of 5 rows and 3 columns of the summation array can be calculated by adding all the values from 1 row 1 column to 5 column 3 column in FIG. According to this method, the sum of all the values in the matrix shown in FIG. 4 becomes a value of 8 rows and 8 columns on the sum array generated through this.

In the method of generating such a sum array, the values of the previously calculated sum array can be used. For example, in the specific sum array generated through the matrix of FIG. 4, the values of the 3 rows and 6 columns are obtained by subtracting the values of the 2 rows and 5 columns from the values obtained by adding the values of the 3 rows and 5 columns and the values of the 2 rows and 6 columns, By summing the values in the third row and sixth column in Fig. According to this method, it is possible to calculate the values constituting the summation array without performing a large amount of addition operation as described above.

Also, as will be described later, a second processing is performed in order to acquire feature points in the fingerprint image. At this time, a sum value of density values of pixels existing in a region having a constant width is required.

In embodiments of the present invention, in this case, the speed of the process can be improved by using the previously calculated sum array.

For example, when the total sum of the density values included in the D area is required in the data of the matrix form of FIG. 4, the sum of the density values from the first row to the first row to the point d, The sum of the density values of the pixels existing in the D region can be calculated by subtracting the sum of the density values up to the point c and summing the sum of the density values up to the point a. That is, if the value of 6 rows and 3 columns and the value of 3 rows and 6 columns are subtracted from the values of 6 rows and 6 columns in the above calculated sum array, and the values of 3 rows and 3 columns are added to the values, To obtain the total sum of the pixel density values.

As described above, by using the sum array generated in advance for a specific image, the sum of the density values of a specific area on the fingerprint image can be calculated by only two subtractions and one addition. In this way, if the sum array is used, the sum of the shading values in a specific area can be calculated only by adding and subtracting without complicated calculation such as multiplication or division. Therefore, speed improvement without waste of hardware resources is performed in performing fingerprint recognition .

5 is a diagram illustrating a method of performing second processing on a fingerprint image in a feature point extraction process according to an exemplary embodiment of the present invention.

As described above, the density value of each pixel constituting the fingerprint image can be temporarily stored in a matrix format of NxM type.

When the second processing is performed on the obtained fingerprint image, the values after the second processing is performed on each pixel may be calculated and stored as matrix format data of the same size as the fingerprint image.

A method of performing second processing with respect to a density value of each pixel is as follows. First, a density value average of pixels existing in a specific surrounding area is calculated based on one pixel, and then the width of the specific area is gradually expanded The second processing application value of the corresponding pixel can be calculated using the final value obtained by calculating the average value of the density value and summing the average values calculated at each step.

Referring to FIG. 5, a process of calculating a second processing application value for a specific pixel will be described. A specific pixel is a pixel existing in the center of the C region 230 including the A region 210 including the corresponding pixel, the B region 220 including the A region 210, and the B region 220. [ As the expansion from the A region 210 to the B region 220 and from the B region 220 to the C region 230 is performed, the number of pixels existing in each region increases.

According to one embodiment, the average value of the density values of the pixels included in the A area 210, the average value of the density values of the pixels included in the B area 220, and the average value of the C area 230 ) May be summed to calculate a second processing performance value for the particular pixel.

As described above, in calculating the average value of the density values of the pixels included in each region in the second processing, a pre-generated summation array through the first processing can be utilized. As described with reference to FIG. 4, when the sum array is used in the process of calculating the sum of the pixel density values existing in a specific area, a desired value can be calculated quickly using only a few hardware resources by merely adding and subtracting .

If the pixels constituting the fingerprint image are positioned at the edges of the fingerprint image, pixels may not exist in some areas within the area formed around the pixel. In this case, the average density value of each area can be calculated in such a manner that the average of the density values is calculated in consideration of only the number of pixels existing in the area.

A process for calculating the second processing performance value for a specific pixel on the fingerprint image according to another embodiment may be performed by applying different weights to the average values of the regions shown in FIG. For example, a weighting factor of 5 times is applied to the average value of the A region 210, a weighting factor of 3 times is applied to the average value of the B region 220 that is extended by one step, and an average value of the C region 230 is less than 3 times Weight can be added. The reason why a larger weight is added to the average value of the inner area is as follows. That is, as the area for calculating the average value is enlarged, the difference in the relationship with the pixel density value at the center of the average value of the density value is prevented . As described above, when a weight is added, the result of the second processing can be obtained by dividing the weighted value by adding different weights to the average values of the density values of the areas A, B, and C (210 to 230) have. For example, when the A region 210 is weighted five times and the B region 220 is weighted three times, the average value of the density values of each region is multiplied by a weight, , And dividing the total area by 8, which is the sum of the weights applied to the entire area, to the second processing performance value of the specific pixel. For example, the method of calculating the second processing performance value for each area is shown in FIG. 5 as follows.

In Equation (1), PC represents a second processing performance value for a specific pixel, and Mean (A), Mean (B) and Mean (C) Represents an average. W1, w2, and w3 represent weights to be applied to the respective regions. According to an exemplary embodiment, as described above, the weights for the regions located inside are set to be high, Can be formed. The second processing performance value calculated in this manner can be used for the minutiae point calculation by comparison with a predetermined threshold value in the future.

Although the shape of the region arbitrarily determined in the second processing is circular, the shape of the region may be embodied in various forms such as a deformed circle or polygon such as an ellipse.

According to an embodiment of the present invention, in the step of calculating a second processing calculation value for a density value of a particular pixel, a predetermined region centering on the pixel is stepwise expanded, And the average value of the regions.

5, an average of the pixel density values included in the A and B regions 210 and 220 is preferentially calculated, and the average of the calculated pixel density values is calculated. The average values may be compared to determine whether to average the C region 230. [

Specifically, when the average value of the specific region is equal to or greater than the average value of the previous region, or when the decrease value when compared with the average value of the previous region is equal to or less than the predetermined reference value, the average value for the next region can be calculated by further expanding the region. Alternatively, when the average value of the specific region is compared with the average value of the previous region, if the decrease value is not less than the predetermined reference value, the average value is not calculated by expanding the region, Value of the second processing execution value.

5, the average value of the density values of the pixels included in the A area 210 and the average value of the density values of the pixels included in the B area 220 are compared with each other, If the average value of the density values of the pixels included in the B area 220 is smaller or smaller than the predetermined reference value, the average value of the density values of the pixels included in the C area 230 extended from the B area 220 And the calculated value can be used for calculating the second processing performance value. Thereafter, the average value of the density values of the pixels included in the B area 220 is compared with the average value of the density values of the pixels included in the C area 230, and the D area (not shown) It is possible to determine whether to calculate the grayscale value average of the pixels.

The average human fingerprint image contains ridges and ridges that rotate in an elliptical shape. If the area in which the average of the pixel density values is calculated is expanded and a large number of pixels corresponding to the goal are included, the average value of the corresponding area is decreased and the decreased value exceeds the reference value. The second processing performance value is calculated through an average of the pixel density values calculated to the region where the current expansion is completed or the region immediately before the expansion, and through this, a second processing performance value suitable for analysis of the human fingerprint type can be calculated.

In the present invention, the above-described methods are applied to the respective pixels constituting the fingerprint image to calculate the second processing performance value, and thereby, a matrix having the same size as the matrix format data storing each pixel density value of the fingerprint image The format data can be calculated and stored on the memory of the electronic device 100. [ Hereinafter, a distorted image formed through the set of second processing execution values calculated for one fingerprint image will be referred to as a "second processing image ".

6 is a view for explaining a method of selecting feature points for performing fingerprint recognition on the second processing image according to an embodiment.

Referring to FIG. 6, a window of a predetermined size including a plurality of pixels is set on the second processing image, and the density values of the pixels existing in the window region are analyzed, so that the minutiae in the corresponding region can be selected.

The process of extracting feature points in a specific window area is as follows. The second processing performance value of each pixel in the fingerprint image for which the second processing has been completed, that is, the density value of each pixel for which the second processing is completed, is compared with the density value of surrounding pixels, It is selected as a feature point. In this process, the number of feature points selected in an area within one window can be limited. When a larger number of pixels than the limited number of pixels meet the feature point selection criterion, only a limited number of pixels collide with each other, . That is, in a region within one window, only a limited number of pixels or less can be selected as the minutiae. Such a contention process can be performed by scoring each pixel located in the window region, and the selection is performed in the order of higher scores, so that only a limited number of pixels can be selected as the minutiae.

According to another embodiment, the minimum number of pixels to be selected as feature points, along with the maximum number of pixels that can be selected as feature points within a window region, can also be set. At this time, the maximum number and the minimum number of pixels may be variously set according to the security level of the fingerprint recognition or the characteristic of the fingerprint recognition device. As described above, when only a fewer number of pixels than the minimum number of pixels to be selected as the minutiae in one window region satisfy the minutia point selection criterion, the minutiae can be selected among the pixels that do not meet the minutiae point selection criteria. Such a process can be performed by scoring each pixel in the window region as described above, and selecting a pixel having a small number of pixels as feature points in descending order of the score.

According to one embodiment, the positional relationship of each feature point can be limited even within one window area. For example, when the positions of the pixels selected as the minutiae are concentrated among the pixels existing in one window region, it may be difficult to perform the fingerprint authentication later. Therefore, the pixels selected as the minutiae in one window region The minimum separation distance between the two can be set.

Hereinafter, a method of moving a window that is a reference region of minutia point setting within one second processing image will be described. Hereinafter, the movement of the window on the second processing image will be referred to as "shift ".

First, referring to FIG. 6 (a), the windows may be shifted so that there is no overlapping area. In this case, each pixel in the image in which the second processing is completed is included once in the specific window.

Also, referring to FIG. 6B, the window may be shifted to overlap the neighboring window and a part of the area. In this case, feature points specified in the first window can be selected as feature points even in the second window. In order to prevent duplication of the feature point extraction process, in the region where the analysis for feature point extraction is performed in the first window, It is not included in the analysis area in the window.

The feature points selected through the above process can be used for the fingerprint authentication by comparing with the feature points extracted from the fingerprint image obtained later.

According to one embodiment, when feature points are selected on the fingerprint image, information on the feature points is stored in the memory. The original fingerprint image temporarily stored on the memory, the summation array generated through the first processing, It can be deleted after minutiae extraction.

7 (a) and 7 (b) show a fingerprint image obtained through a fingerprint sensor according to an embodiment, respectively, a fingerprint image obtained through a finger of an actual person and a counterfeit fingerprint image obtained through a counterfeit fingerprint.

8A and 8B show the result of extracting feature points through the process according to the embodiment of the present invention after normalizing the images of FIGS. 7A and 7B, respectively.

7, an absolute gray value difference between the image obtained through the actual finger (FIG. 7A) and the image obtained through the counterfeit fingerprint (FIG. 7B) You can see that it is flying.

8, since the gray value distribution according to the ridge flow of the fingerprint is also different, the geographical position of the minutiae extracted from the image obtained through the actual finger and the minutiae extracted from the image obtained through the fake fingerprint It can be seen that there is a mutual difference.

When the fingerprint image is acquired through the forged fingerprint, the distribution of the gray value according to the ridge flow is different from the difference between the actual finger and the difference .

In addition, the manufacturing process of the forged fingerprint may include a process of duplicating the shape of the biometric fingerprint on the forged fingerprint frame, and fabricating a counterfeit fingerprint using the duplicated fingerprint frame. In this process, . ≪ / RTI >

Accordingly, in the forged fingerprint than in the biometric fingerprint, the ridge is further spread on the basis of the center of the fingerprint. Therefore, the geographical location of the feature points extracted from the images obtained through the actual fingerprints and the geographical positions of the feature points extracted from the images obtained through the forged fingerprints differ.

According to the embodiment of the present invention, since the feature points are found through the change value of the shade change in each pixel of the fingerprint image and the fingerprint authentication is performed through the characteristic points, a delicate joke according to the ridge direction between the previously registered fingerprint and the obtained fingerprint It is possible to judge whether or not the fingerprint is authenticated by reflecting the difference in the value change or the minute difference in the ridge direction. Therefore, according to the embodiment of the present invention, it is possible to accurately distinguish the actual fingerprint and the counterfeit fingerprint.

9 is a flowchart illustrating a fingerprint registration and authentication process according to an embodiment of the present invention.

Referring to FIG. 9, the fingerprint registration process (S910) will be described first. When the user's finger touches the fingerprint sensor at the time of fingerprint registration, an output value is obtained from the fingerprint sensor (S911).

When an output value is obtained from the fingerprint sensor, an optimal fingerprint image for extracting a fingerprint feature point according to an embodiment of the present invention is obtained based on the output value.

In order to acquire a fingerprint image, a predetermined gain value may be applied to a signal output through the fingerprint sensor. The applied gain value may be determined by hardware or software depending on the magnitude of the signal output through the fingerprint sensor. The gain value applied to obtain an optimal fingerprint image for enabling fingerprint feature point extraction and authentication through the fingerprint feature point will be referred to as 'registered gain value' hereinafter.

The gain value depends on the magnitude of the signal obtained in the relationship between the fingerprint sensor and the user fingerprint, and thus relates to the conductivity of the user finger. In general, a dry finger requires a high gain because a signal obtained from the fingerprint sensor is small, and a wet finger requires a low gain because the signal obtained from the fingerprint sensor is large. Therefore, the basic property of the finger contacted to the fingerprint sensor can be grasped through the 'registered gain value'.

According to an embodiment of the present invention, after step S911, a process of acquiring a fingerprint image by applying a registration gain value to an output value from the fingerprint sensor and a process of storing the registered gain value in a memory may be performed (S912) .

According to an embodiment, the result of applying the predetermined gain value to the output value from the fingerprint sensor obtained in step S911 may be calculated and the result thereof may be stored (S913). The predetermined gain value may be a value different from the registered gain value, but is not limited thereto. If the preset gain value and the registered gain value are equal, step S913 may be omitted. Meanwhile, the result of applying the predetermined gain value may be stored as an average value of the result of applying the predetermined gain value to the output value from each sensing pixel included in the fingerprint sensor, but the present invention is not limited thereto.

For the fingerprint image obtained in step S912, the first processing and the second processing described above are performed and the feature points of the fingerprint are obtained (S914).

Although the steps S912 to S914 are illustrated as being sequentially performed in FIG. 9, the three steps may be performed at the same time, in reverse order, or in a different order. If three steps are performed during the fingerprint registration process, Should be understood.

Next, the fingerprint authentication process (S920) will be described.

In fingerprint authentication, when a user's finger touches the fingerprint sensor, an output value is obtained from the fingerprint sensor (S921).

As a gain value to be applied to the output value from the fingerprint sensor acquired in step S921, the gain value for obtaining the fingerprint image optimized for minutia extraction and authentication is determined either hardware or software at the time of fingerprint authentication. The gain value thus determined will be referred to as an " authentication gain value ".

After step S921, the fingerprint image is obtained by applying the authentication gain value to the output value obtained from the fingerprint sensor, and temporarily stores the authentication gain value (S922).

Then, the first processing and the second processing described above are performed based on the obtained fingerprint image to extract the feature points of the fingerprint (S923).

When the fingerprint feature point is obtained, a comparison with the feature point of the registered fingerprint image obtained in step S914 is performed (S924). If the comparison result of the minutiae points is less than the reference value, it is determined that the authentication is failed (S929).

On the other hand, if the degree of similarity between the minutiae points is equal to or greater than the reference value, since the authentication target fingerprint pattern coincides with the previously registered fingerprint pattern, a process for determining whether or not the fingerprint to be authenticated is a counterfeit fingerprint is performed.

Specifically, according to an embodiment of the present invention, the authentication gain value applied for fingerprint image acquisition in step S922 is compared with the registered gain value stored in step S912 in fingerprint registration (S925). If the difference between the authentication gain value and the registration gain value is out of the reference range, it is determined that the authentication fails even if it is determined that the fingerprint minutiae coincides in step S924 (S929).

In the case of attempting fingerprint authentication through forged fingerprints, even if the fingerprint feature points are exactly the same as the feature points of the user fingerprint, the physical properties such as the humidity of the surface may be different from those of the user. For example, if fingerprint registration is performed with a dry finger but the surface of the subject to be fingerprint authentication is humid, the gain value applied for fingerprint registration and authentication at the time of authentication will be different. Thus, by performing step S925, Can be verified.

According to another embodiment of the present invention, it may be determined whether the authentication target fingerprint is a fake fingerprint based on the output value from the fingerprint sensor acquired in step S921.

Specifically, the result obtained by applying the predetermined gain value to the output value from the fingerprint sensor acquired in step S921 is obtained (S926). The predetermined gain value may be the same value as the gain value applied in step S913 in the fingerprint registration process.

The result obtained in step S926 is compared with the result stored in step S913 (S927).

If the result of applying the predetermined gain value to the output value from the fingerprint sensor matches the fingerprint registration time and the authentication or the difference is within the reference range, it is determined that the fingerprint to be authenticated is not a counterfeit fingerprint. That is, it is determined that the fingerprint authentication is successful (S928). On the contrary, if the difference exceeds the reference range, it is determined that fingerprint authentication has failed (S929).

For example, when the surface of the registered fingerprint is high in humidity and the predetermined gain value is higher than the registered gain value, the result obtained by applying the predetermined gain value to the output value from the fingerprint sensor during fingerprint registration becomes close to the saturated state. If the surface of the fingerprint to be authenticated is dry, the result of applying the predetermined gain value to the output value from the fingerprint sensor is not close to the saturation state. In this case, it is judged that the fingerprint authentication has failed. That is, by using the result of applying the predetermined gain value different from the registered gain value to the output value from the fingerprint sensor, the forgery fingerprint detection based on the difference in physical properties between the fingerprint registration and the fingerprint surface at the time of authentication can be performed more precisely.

On the other hand, the range of the output value from the fingerprint sensor can be subdivided into the range of the gain value to be compared in step S925. This means that the range of output values from the fingerprint sensor can be freely subdivided according to the sensing sensitivity of the fingerprint sensor, while the range of gain values can be increased if the subdivision is too fine to determine an optimized registered gain value or an authenticated gain value It is because. Therefore, in step S927, the falsification fingerprint detection process based on the output value from the fingerprint sensor enables a more accurate determination than the falsification fingerprint detection process by comparing the gain value in step S925.

According to the embodiment of the present invention, at least one of step S925 and step S927 is selectively performed so that forgery fingerprint detection can be performed. That is, determination of whether or not the authentication gain value coincides with the registration gain value, and determination through mutual comparison of the result of applying the predetermined gain value to the output value from the fingerprint sensor may be performed sequentially, but only one determination .

10 is a block diagram showing the configuration of a fingerprint registration and authentication apparatus according to an embodiment of the present invention.

The fingerprint authentication device 130 shown in FIG. 10 receives an electrical signal from the fingerprint sensor 120 (see FIG. 1) included in the electronic device 100 (see FIG. 1) Authentication is performed.

10, the fingerprint authentication apparatus 130 includes a fingerprint image acquisition unit 131, a gain determination unit 132, a fingerprint feature point acquisition unit 133, a fingerprint feature point comparison unit 134, The gain value comparison unit 135, the default gain value application unit 136, the gain value application comparison / determination unit 137, and the control unit 138. [

The fingerprint image obtaining unit 131 performs a function of obtaining a fingerprint image based on an output value transmitted from the fingerprint sensor 120, that is, an electrical signal.

The gain determination unit 132 determines a gain value to be used when acquiring a fingerprint image by the fingerprint image acquisition unit 131, that is, a gain value for acquiring an optimal fingerprint image enabling feature point extraction and fingerprint authentication, Determines a gain value and an authentication gain value, and stores the determined gain value. As described above, the registration gain value and the authentication gain value are compared with each other and utilized in the detection and determination of a fake fingerprint.

The fingerprint feature point acquiring unit 133 performs the first processing and the second processing described above on the basis of the obtained fingerprint image to acquire the feature points of the fingerprint. Fingerprint feature point acquisition may be performed in such a manner that the feature points are selected within each window while shifting the window on the fingerprint image where the second processing is completed as described above. In each window, it is possible to extract a pixel value of a pixel having a gradient change value greater than or equal to a threshold value, after calculating a gradient change value in each pixel, that is, a gradient difference value with neighboring pixels. At this time, the number of feature points that can be extracted in each window may be limited. In addition, the extracted minutiae points can be stored in a template form.

The fingerprint feature point comparing unit 134 compares the feature points of the registered fingerprint image with the feature points extracted through the fingerprint image to be authenticated. The minutiae points of the registered fingerprint image may be stored in a template form. The minutiae extracted from the fingerprint image to be authenticated may correspond to a part of the templates of the registered fingerprint image. Accordingly, the process of determining which part of the template of the previously registered fingerprint image corresponds to the feature points of the fingerprint image obtained through the positional relationship between the feature points extracted from the fingerprint image to be authenticated is performed before the comparison of the feature points Can be preceded. If the comparison target positions between the minutiae points of the obtained fingerprint image and the minutiae points of the previously registered fingerprint image are selected, the second processing performance values of the minutiae points are compared with each other, and the similarity is calculated. The degree of similarity refers to the degree of similarity between the second processing performance value for the first feature point among the feature points of the previously registered fingerprint image and the second processing performance value for the second feature point of the corresponding authentication target fingerprint image. 2 < / RTI > processing execution value difference. When the degree of similarity between the respective feature points is calculated, it is determined whether or not the degree of similarity is equal to or greater than a predetermined threshold. If the degree of similarity is less than the threshold value, it can be determined that the fingerprint authentication has failed.

The gain value comparison unit 135 compares the registration gain value and the authentication gain value with each other when it is determined through the fingerprint feature point comparison unit 134 that the similarities between the feature points of the fingerprint image previously registered and the feature points of the fingerprint image thus obtained are equal to or more than a threshold value And performs a function of mutual comparison. As described above, if there is a difference in physical properties of the subject surface during fingerprint registration and authentication, the registered gain value and the authentication gain value also differ. If the difference between the registration gain value and the authentication gain value is out of the reference range, the authentication failure can be determined. That is, even if the feature points of the fingerprint image to be authenticated are judged to be highly similar to the feature points of the registered fingerprint image through the fingerprint feature point comparing unit 134, the fingerprint to be authenticated can be calibrated through comparison between the registration gain value and the authentication gain value It can be determined that the fingerprint is not the user's fingerprint, and the forged fingerprint can be detected.

The default gain value application unit 136 obtains a result of applying a predetermined gain value to an output value from the fingerprint sensor. Specifically, the default gain value application unit 136 obtains an average value or a value proportional to a result of applying a predetermined gain value to each of the output values from the sensing pixels included in the fingerprint sensor. The result obtained at the time of fingerprint registration and the result obtained at the time of fingerprint authentication are compared with each other and used for detection of a fake fingerprint based on the difference in physical properties of the fingerprint surface.

When the similarity between the minutiae points of the previously registered fingerprint image and the minutiae points of the obtained fingerprint image is determined to be equal to or greater than the threshold value through the fingerprint minutia matching section 134, the comparison / And compares the values obtained at the time of fingerprint registration and authentication as the result values obtained by the application unit 136 with each other. If the comparison results match or the difference is within the reference range, it is determined that the fingerprint authentication is successful.

The control unit 138 includes a fingerprint image acquisition unit 131, a gain determination unit 132, a fingerprint feature point acquisition unit 133, a fingerprint feature point comparison unit 134, a gain value comparison unit 135, And a gain value application result comparison / determination unit 137 perform respective unique operations and control the transmission / reception of information and interlocking of each other.

According to the embodiment of the present invention, since a point at which the change in the grayscale value is greater than or equal to a threshold value is extracted as a minutia point on the fingerprint image, and fingerprint authentication is performed through the minutiae point, the pressure applied by the subject during fingerprint registration and fingerprint authentication, Accurate fingerprint authentication can be performed even when the characteristics are slightly different from each other, and detection of a counterfeit fingerprint can be facilitated.

According to the embodiment of the present invention, even if the extracted minutiae match the registered minutiae in the fingerprint authentication, the gain value applied to the fingerprint image acquisition in the fingerprint authentication or the preset gain value in the output value from the fingerprint sensor is applied It is judged whether the characteristic of the subject matches with the registered fingerprint, and whether the fingerprint authentication succeeds or not is finally determined. Therefore, the fingerprint authentication attempt through the counterfeit fingerprint can be accurately blocked.

The embodiments of the present invention described above can be implemented in the form of program instructions that can be executed through various computer components and recorded on a computer-readable recording medium. The computer-readable recording medium may include program commands, data files, data structures, and the like, alone or in combination. The program instructions recorded on the computer-readable recording medium may be those specially designed and constructed for the present invention or may be those known and used by those skilled in the computer software arts. Examples of computer-readable recording media include magnetic media such as hard disks, floppy disks and magnetic tape, optical recording media such as CD-ROMs and DVDs, magneto-optical media such as floptical disks, 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 generated 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 device may be configured to operate as one or more software modules for performing the processing according to the present invention, and vice versa.

It will be understood by those skilled in the art that the foregoing description of the present invention is for illustrative purposes only and that those of ordinary skill in the art can readily understand that various changes and modifications may be made without departing from the spirit or essential characteristics of the present invention. will be. It is therefore to be understood that the above-described embodiments are illustrative in all aspects and not restrictive. For example, each component described as a single entity may be distributed and implemented, and components described as being distributed may also be implemented in a combined form.

The scope of the present invention is defined by the appended claims, and all changes or modifications derived from the meaning and scope of the claims and their equivalents should be construed as being included within the scope of the present invention.

Claims (9)

(a) acquiring minutiae points based on a change in the density value of a specific pixel and a neighboring pixel on a fingerprint image acquired through a fingerprint sensor;
(b) comparing the feature points with previously registered feature points;
(c) comparing the authentication gain value applied to the output value from the fingerprint sensor with the registered gain value applied for fingerprint image registration when the similarity degree is determined to be equal to or greater than the threshold value;
(d) determining that fingerprint authentication fails if the difference between the authentication gain value and the registration gain value is out of a reference range;
(e) when a difference between the authentication gain value and the registration gain value is within a reference range, a result obtained by applying a predetermined gain value to an output value from the fingerprint sensor and a result of applying the predetermined gain value Comparing the result of applying And
(f) judging that the fingerprint authentication is successful only when the difference between the result of applying the predetermined gain to the output value from the fingerprint sensor at the time of fingerprint registration and the authentication is within the reference range, and Authentication method. delete The method according to claim 1,
Wherein the predetermined gain value is different from the registered gain value. The method according to claim 1,
The step (a)
Calculating an average of pixel density values for a specific area centered on each pixel of the fingerprint image, calculating a sum of average pixel density values included in the extended area by expanding the specific area step by step, ; And
Shifting a window having a specific width on the fingerprint image, and selecting a pixel having a change in density value among the pixels in each window as a minutiae. 5. The method of claim 4,
Wherein the average calculation of the pixel grayscale values is performed based on a summation array generated for the fingerprint image. delete delete A fingerprint feature point acquiring unit that acquires feature points based on a change in a shade value of a specific pixel and a surrounding pixel on the fingerprint image acquired through the fingerprint sensor;
A fingerprint feature point comparing unit comparing the feature points with previously registered feature points;
Wherein the authentication gain value applied to the output value from the fingerprint sensor for obtaining the fingerprint image is used for obtaining a fingerprint image upon fingerprint registration when the similarity degree is determined to be equal to or greater than a threshold value A gain comparison unit comparing the registration gain value applied to the fingerprint authentication unit and determining that the fingerprint authentication is successful only when the difference is within the reference range; And
When the difference between the authentication gain value and the registered gain value is within the reference range, a result obtained by applying the predetermined gain value to the output value from the fingerprint sensor and applying the predetermined gain value to the output value from the fingerprint sensor during fingerprint registration Judges that the fingerprint authentication is successful only when the difference between the result of fingerprint registration and the result of applying the predetermined gain value to the output value from the fingerprint sensor during authentication is within the reference range, A fingerprint authentication device.

delete

Images