KR20050014406A - System And Method For Identifying Fingerprint In Mobile Communication Terminal - Google Patents

Abstract

PURPOSE: A system and a method for identifying a fingerprint on a mobile communication terminal using a minutia comparing method are provided to implement a fingerprint identification system on the mobile communication terminal having a small storage space and a low processing speed at low cost without improving a specification of the terminal. CONSTITUTION: A fingerprint inputting module(10) receives/preprocesses/outputs fingerprint image data after sensing the fingerprint of a user. A fingerprint identifying module(20) registers the fingerprint of the user by extracting fingerprint minutia from the preprocessed fingerprint image data and classifying/storing a pattern to a fingerprint database(24), certifies the fingerprint by comparing the inputted minutia with the stored minutia, and outputs a fingerprint identification result for the user. A managing module(30) displays the fingerprint identification result, and transfers a fingerprint identification command to the fingerprint inputting module and the fingerprint identifying module.

Description

System and Method for Identifying Fingerprint In Mobile Communication Terminal}

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to fingerprint recognition in a mobile communication terminal, and in particular, a fingerprint recognition system and method in a mobile communication terminal to provide security to a mobile communication terminal by registering and authenticating a fingerprint of a user to be queried using a feature point collation method. It is about.

With the recent information age, there is an increasing interest in the security of personal information. In the past, most people use physical tools such as passwords, keys, cards, and ID cards to protect personal information. It had the disadvantage of being easy to spill or lose.

In order to secure these shortcomings, with respect to authentication technologies that take into account the biological characteristics (physical, morphological, and behavioral characteristics) of each individual, various biometric authentication methods such as fingerprint, face, iris & retina, hand vein, palm, and voice ( Biometrics techniques have been proposed, of which two characteristics have been verified: fingerprints as biometric means of immutability (which does not change until birth and death) and univariate (different from person to person). As a means of identification, a fingerprint identification system is widely used.

Hereinafter, a fingerprint (Fingerprint) will be described with reference to the accompanying drawings, FIGS. 1, 2, and 3, FIG. 1 is a view showing feature points of a fingerprint, and FIG. 2 is a view showing singular points of a fingerprint, and FIG. It is a figure which shows the form of a fingerprint.

The fingerprint is the palm print from the tip of a person to the first node. That is, the fingerprint is made by swelling the pores with stripes formed by lines (ridges) formed by the epidermis being raised on the palm of the fingertip.

In describing the structure of the fingerprint, the raised part of the fingerprint like a mountain range is called a ridge and the valley between the ridge and the ridge is called a valley, and in addition to the ridge and the valley in the fingerprint, There are some that show morphological characteristics. Looking at the ridge portion of the fingerprint in Figure 1 there is a point in which the ridge is broken, that is, the terminal portion of the ridge flows and breaks, which is called the end point or the end point (ending point), the ridge branches, that is, the ridge flows two There is a branch in the direction, which is called a bifurcation point or forked point. These endpoints and branching points are referred to as feature points of the fingerprint. The number of feature points is different for each fingerprint, and the number is generally about 100 to 150 or more.

When the direction of the feature point is described, the end point is set in the direction toward the ridge, and the branch point is set in the center direction of the two ridges branching. In the actual fingerprint recognition program, the direction of the feature points is quantized, and the current system is to quantize 360 degrees in 64 directions.

In addition, in FIG. 2, a portion of the ridge has the greatest curvature, which is referred to as a center point or core of the fingerprint, and includes an upper core and a lower core, where the ridges are divided into triangles. Is called a delta or a delta of a fingerprint, and there are a left delta and a right delta. These center points (cores) and deltas (triangle points) are called singular points of the fingerprint.

In addition, as shown in Figure 3, the shape of the fingerprint according to the flow shape of the ridges (Whorl), right key pattern (Right Loop), left key pattern (Left Loop), arch pattern (Fuzzy) The above five classifications are most convenient for computerized processing. In general, all ten prints of a person do not have the same shape. How to classify the types of fingerprints is an important issue to speed up the fingerprint control. In other words, pointing out the form of the fingerprint in the fingerprint control checks only the fingerprints of the form, so that the matching time is much shorter.

The fingerprint control refers to authenticating or determining a person having a fingerprint in question by comparing a fingerprint whose identity is unknown with a fingerprint that already knows who is a fingerprint.

The fingerprint control method includes a minutiae matching method and a phase correlation method. The feature matching method extracts the feature points of a fingerprint and calculates the degree of matching between two fingerprints based on a network of features that reflects the mutual relationship therebetween. In the phase correlation method, when the fingerprint image is viewed, the ridges may have a feeling of being arranged at a certain distance from the flow direction in the local region. From this, the frequency characteristics of the fingerprint image are analyzed to determine the degree of agreement. How to evaluate.

On the other hand, users of mobile communication terminals including mobile phones are rapidly increasing, and accordingly, e-business services using mobile communication terminals are being activated. However, there has been a problem that various e-business services that require security are not available through a mobile communication terminal having a weak security. This is because it is difficult to authenticate whether the user using the e-business service through the mobile terminal is a real user of the mobile communication terminal in the corresponding mobile communication terminal, which is characterized by a small storage space and a low processing speed. Accordingly, a user of a mobile communication terminal cannot use various e-businesses, and an e-business service provider also cannot provide various e-business services to a user of the mobile communication terminal.

SUMMARY OF THE INVENTION The present invention has been made in view of the above-mentioned problems, and an object thereof is to implement a fingerprint recognition system using a feature point matching method in a mobile communication terminal having a small storage space and low-speed processing without an improvement in the specification of a device.

Another object of the present invention is to implement a fingerprint recognition system using a feature point matching method in a mobile communication terminal to implement an authentication function suitable for a wireless environment such as a mobile communication terminal at a low cost, and at the same time provide user authentication and security of the mobile communication terminal It is.

1 and 2 is a view showing the feature and singularity of a typical fingerprint.

3 is a view showing a pattern type of a typical fingerprint.

4 is a view showing a fingerprint recognition system in a mobile communication terminal according to an embodiment of the present invention.

5 is a flowchart illustrating a fingerprint recognition method in a mobile communication terminal according to an embodiment of the present invention.

FIG. 6 is a flowchart illustrating a fingerprint registration process of a user to be queried in FIG. 5.

FIG. 7 is a flowchart illustrating a fingerprint authentication process of a user to be queried in FIG. 5.

Explanation of symbols on the main parts of the drawings

10: fingerprint input module 11: fingerprint input unit

12: preprocessing unit 20: fingerprint recognition module

21 memory unit 22 fingerprint feature extraction unit

23: fingerprint registration / authentication unit 24: fingerprint DB

30: management module 31: display unit

32: user interface unit

A feature of the present invention for solving the above object is a fingerprint input module for receiving the fingerprint image data after sensing the fingerprint of the user to be queried; After receiving the preprocessed fingerprint image data from the fingerprint input module, the fingerprint feature points are extracted, the pattern type is classified, and stored in a database to register the fingerprint of the user to be queried, or the user fingerprint feature and the registered fingerprint feature point to be queried. A fingerprint recognition module for authenticating a fingerprint of the corresponding query target user and outputting a fingerprint recognition result for the corresponding query target user; A fingerprint in the mobile communication terminal including a management module for displaying a fingerprint recognition result for the query target user output from the fingerprint recognition module and transferring a fingerprint recognition command for the query target user to the fingerprint input module and the fingerprint recognition module. To provide a recognition system.

The fingerprint input module may include: a fingerprint input unit configured to sense fingerprints of a corresponding query target user and receive fingerprint image data according to a fingerprint recognition command for the query target user; Receiving the fingerprint image data from the fingerprint input unit characterized in that it comprises a pre-processing unit for removing the noise of the fingerprint and performing the fingerprint pre-processing using the histogram smoothing.

The fingerprint recognition module may include: a memory unit configured to temporarily store fingerprint image data preprocessed by the preprocessor; A fingerprint feature extracting unit for extracting fingerprint feature points and classifying a pattern type by binarizing and thinning fingerprint image data stored in the memory unit; A fingerprint DB for dividing the fingerprint feature points extracted by the fingerprint feature extractor into a database by dividing the fingerprint feature points into corresponding database pattern types; According to the fingerprint recognition command for the query target user, the fingerprint feature points extracted by the fingerprint feature extracting unit are stored in the fingerprint DB for each pattern type to register a fingerprint of the query target user or the query target extracted by the fingerprint feature extracting unit. By comparing the fingerprint feature points of the user with the feature points of the registered fingerprints, by measuring the similarity of the types, locations, and directions of the feature points, it is determined whether the fingerprints are the same and the fingerprints of the user to be queried are authenticated. Characterized in that it comprises a fingerprint registration / authentication unit for outputting.

In this case, the fingerprint feature extracting unit calculates a fingerprint direction by applying a gradient method and 8 quantization directions to the fingerprint image data, corrects the fingerprint direction by using a weight-based method, and calculates a unit of 9 * 9 pixel units. It is characterized by detecting the ridges using.

In addition, the management module receives a fingerprint recognition result for the user to be queried output from the fingerprint registration / authentication unit of the fingerprint recognition module to check whether the fingerprint is registered and authenticated by the user to be queried and the registered and authenticated fingerprint image. A display unit for displaying; And a user interface unit configured to receive a fingerprint recognition request from a user and transfer a fingerprint recognition command to a query target user to a fingerprint input unit and a fingerprint registration / authentication unit of the fingerprint input module.

According to another aspect of the present invention, there is provided a method of performing preprocessing by receiving fingerprint image data after sensing a fingerprint of a user to be queried according to a predetermined fingerprint recognition command; Extracting fingerprint feature points by binarizing and thinning the preprocessed fingerprint image data; If the fingerprint recognition command is a fingerprint authentication command for the query target user, performing fingerprint authentication of the query target user using the extracted fingerprint feature points; The present invention provides a fingerprint recognition method in a mobile communication terminal including a process of displaying a fingerprint authentication result and an authentication fingerprint image as a fingerprint authentication result for the query target user.

Further, in the above-described fingerprint recognition method in the mobile communication terminal, when the fingerprint recognition command is a fingerprint registration command for the inquiry target user, the fingerprint feature point is stored in a database to perform fingerprint registration of the inquiry target user. Process of doing; The method may further include displaying a fingerprint registration and a fingerprint image as a fingerprint registration result for the query target user.

The extracting of the fingerprint feature point may include performing image binarization with different threshold values for each fingerprint image block after dividing the preprocessed fingerprint image data into a predetermined number of fingerprint image blocks; And extracting fingerprint feature points by performing thinning to produce a line image having the width of the ridges in the pixel unit from the binarized fingerprint image.

The binarization of the preprocessed fingerprint image data may include determining an initial ridge direction by a projection method paying attention to the maximum and minimum values of concentration distribution variance of the preprocessed fingerprint image data, Determining the ridge direction of the fingerprint by calculating a high accuracy direction; Extracting a ridge of a fingerprint through a direct extraction method from a density gradation image and a discrimination process on a region having a large curvature value; Comprising the step of performing the fingerprint image binarization by calculating the ridge interval distribution of the extracted fingerprint.

In the extracting of the fingerprint feature point, the fingerprint direction is calculated in the 8 quantization direction using the gradient method, the fingerprint direction is corrected using the center of gravity point, and the ridge is detected in 9 * 9 pixel units. Characteristic points are extracted.

Further, the process of performing fingerprint authentication of the query target user may include: reading a matching fingerprint feature point that is most similar to the extracted query target fingerprint feature point among pre-registered fingerprint feature points; Analyzing the fingerprint feature point to be queried to determine whether it is satisfied as a feature point matching candidate; Rejecting fingerprint authentication for the corresponding query target user when the verification result is not satisfied as the feature verification candidate, and generating the son point number of the central relationship when the verification is satisfied as the feature verification candidate; Obtaining a collation score indicative of fingerprint similarity by performing local collation by calculating local information of the querying fingerprint feature point and the collation target fingerprint feature point, and calculating an entire correspondence relationship to perform collation for all fingerprint feature points; And comparing the obtained control score with a predetermined authentication criterion score, and authenticating whether the inquiry target fingerprint and the verification target fingerprint are identical.

In this case, the process of performing fingerprint authentication of the user to be queried includes: determining whether the center is analyzed by analyzing the fingerprint feature to be queried; As a result of the checking, if there is a center, the method may further include performing preliminary coordinate system matching with respect to the center of the inquiry target fingerprint feature point and the collation target fingerprint feature point.

On the other hand, the process of performing a fingerprint registration of the query target user, the step of extracting the initial feature point from the binarized and thinned fingerprint image data; Determining an indeterminate region after extracting fingerprint feature points through feature neighborhood relation calculation by a correction and labeling method based on arbitrary region reliability and spacing information; And registering the fingerprint of the user to be queried by classifying the extracted fingerprint feature points according to the pattern types of the corresponding fingerprints and storing them in a database.

Hereinafter, an embodiment according to the present invention will be described in detail with reference to the accompanying drawings.

In the present invention, a fingerprint recognition system using a feature point matching method of a fingerprint is implemented in the terminal without improving the device specification of the mobile communication terminal. The fingerprint recognition system in the mobile communication terminal according to the embodiment of the present invention is shown in the accompanying drawings. As shown in the drawing, the fingerprint input module 10, the fingerprint recognition module 20 and the management module 30 is provided.

The fingerprint input module 10 receives a fingerprint image data after sensing the fingerprint of the user to be queried and preprocesses the fingerprint image data. The fingerprint input module 10 includes a fingerprint input unit 11 and a preprocessing unit 12. ) Senses the fingerprint of the corresponding query target user and receives the fingerprint image data according to the fingerprint recognition command of the query target user, and the preprocessor 12 transfers the fingerprint image data from the fingerprint input unit 11. It removes the noise of the fingerprint and outputs it by performing fingerprint preprocessing using histogram equalization. Here, the fingerprint input unit 11 is located on the surface of the mobile communication terminal in order to sense the fingerprint of the user to be queried, which may be implemented using an optical fingerprint input device or a semiconductor fingerprint input device.

The fingerprint recognition module 20 receives the preprocessed fingerprint image data from the fingerprint input module 10, extracts fingerprint feature points, classifies the pattern type, and stores the data as a database to register the fingerprint of the user to be queried or to be queried. By comparing the feature point of the user fingerprint with the feature point of the registered fingerprint, the fingerprint of the corresponding query target user is authenticated and the fingerprint recognition result is output.

To this end, the fingerprint recognition module 20 includes a memory unit 21, a fingerprint feature extraction unit 22, a fingerprint registration / authentication unit 23 and a fingerprint DB (DataBase) 24, the memory unit 21 ) Temporarily stores the fingerprint image data preprocessed by the preprocessor 12, and the fingerprint feature extractor 22 binarizes and thins the fingerprint image data stored in the memory 21. Extract fingerprint features and classify pattern types. The fingerprint DB 24 stores the fingerprint feature points extracted by the fingerprint feature extractor 22 as a database by dividing the fingerprint feature points according to the pattern type of the corresponding fingerprint, and the fingerprint registration / authentication unit 23 prints the fingerprint for the user to be queried. The fingerprint feature point extracted by the fingerprint feature extractor 22 is stored in the fingerprint DB 24 according to a pattern type according to a recognition command to register a fingerprint of the user to be queried, or the query object extracted by the fingerprint feature extractor 22. By comparing the fingerprint feature points of the user with the feature points of the registered fingerprints, by measuring the similarity of the types, locations, and directions of the feature points, it is determined whether the fingerprints are the same and the fingerprints of the user to be queried are authenticated. Outputs

In addition, the fingerprint feature extractor 22 uses the gradient method instead of the existing projection method extension, 8 quantization direction instead of 16 quantization, and corrects the fingerprint direction in calculating the fingerprint direction with respect to the fingerprint image data. In this case, the weight focus method is used instead of the conventional stochastic relaxation method, and the convergence detection operation of 9 * 9 pixel unit instead of 3 * 3 pixel unit improves the processing speed in the mobile terminal operating environment. have.

The management module 30 displays a fingerprint recognition result for the query target user output from the fingerprint recognition module 20 and sends a fingerprint recognition command for the query target user to the fingerprint input module 10 and the fingerprint recognition module 20. To pass. Hereinafter, "user" refers to a person who makes a fingerprint recognition request as an administrator of the mobile communication terminal, and "lookup target user" refers to a person who is subject to fingerprint registration and authentication according to the fingerprint recognition request of the "user".

The management module 30 includes a display unit 31 and a user interface unit 32, which displays the fingerprint recognition result for the user to be queried output from the fingerprint registration / authentication unit 24. Receives and displays the fingerprint registration and authentication of the user and the corresponding fingerprint image and the registered and authenticated, and the user interface 32 receives the fingerprint recognition request of the user to remind the fingerprint recognition command for the user to be queried The fingerprint input unit 11 and the fingerprint registration / authentication unit 23 are transferred. Here, the fingerprint recognition request of the user means a fingerprint registration request or a fingerprint authentication request, and correspondingly, the fingerprint recognition command for the query target user means a fingerprint registration command or a fingerprint authentication command. The display unit 31 may be an LCD screen of the mobile communication terminal, and the user interface 32 may be an input key of the mobile communication terminal.

The fingerprint recognition method in the mobile communication terminal according to the embodiment of the present invention as described above will be described with reference to FIG.

First, when a user of a mobile communication terminal makes a fingerprint recognition request to a mobile communication terminal for fingerprint registration or fingerprint authentication for a user to be queried, the user interface unit 32 of the management module 30 provided in the corresponding mobile communication terminal Receives a user's fingerprint recognition request (step S501), and the fingerprint input command 11 of the fingerprint input module 10 and the fingerprint registration / authentication module 20 of the fingerprint recognition module 20 23).

Accordingly, when the fingerprint input unit 11 receives the fingerprint recognition command for the query target user from the user interface unit 32 (step S502), the fingerprint of the query target user according to the fingerprint recognition command for the corresponding query target user. Is sensed to receive fingerprint image data (step S503).

Then, the preprocessing unit 12 receives the fingerprint image data from the fingerprint input unit 11 and preprocesses the corresponding fingerprint image data and outputs it (step S504). Here, the preprocessing of the fingerprint image data refers to a processing operation for extracting fingerprint feature points from the fingerprint image data. In this case, a small dot-like noise after redistributing the contrast of the fingerprint image data using a median filter is used. It removes sweat, moisture, and other oils that may always be present in a person's hands, dust such as dust that may be present due to the cleanliness of the hand, and noise that may be caused by the sensitivity of the fingerprint input unit. Image preprocessing is performed to unfold the intensity values concentrated in one place to have a uniform distribution.

Thereafter, the fingerprint image data output after being preprocessed by the preprocessor 12 of the fingerprint input module 10 is received by the fingerprint recognition module 20 and stored in the memory unit 21 (step S505). In this case, the size of the preprocessed fingerprint image data stored in the memory unit 21 may be 128 * 128.

The fingerprint feature extracting unit 22 of the fingerprint recognition module 20 extracts fingerprint feature points by binarizing and thinning the fingerprint image data stored in the memory unit 21 (step S506). That is, the fingerprint feature extracting unit 22 divides the 128 * 128 preprocessed fingerprint image data into several fingerprint image blocks and binarizes the images with different threshold values for each fingerprint image block. The fingerprint feature points are extracted more easily from the fingerprint image data by performing a thinning process to create a line image having a ridge width of 1 pixel from the fingerprint image. In this case, the fingerprint feature extracting unit 22 uses the gradient method. The fingerprint direction is calculated in 8 quantization directions, the fingerprint direction is corrected using the center of gravity, and the ridge is detected in 9 * 9 pixel units. Can be applied as

In addition, the fingerprint registration / authentication unit 23 of the fingerprint recognition module 20 determines whether the fingerprint recognition command for the query target user received from the user interface unit 32 is a fingerprint registration command or a fingerprint authentication command. (Step S507).

At this time, when the determination result (step S507), the received fingerprint recognition command for the inquiry target user is a fingerprint registration command, the fingerprint registration / authentication unit 23 is extracted by the fingerprint feature extraction unit 22 After the fingerprint feature is stored in the fingerprint DB 24 to perform fingerprint registration of the user to be queried (step S508), the management module 30 checks whether the fingerprint registration result and the fingerprint image are the fingerprint recognition result for the corresponding user to be queried. The display unit 31 is transferred to the display unit 31, and the display unit 31 receives the fingerprint registration result and the fingerprint image, which is the fingerprint registration result for the user to be queried from the fingerprint registration / authentication unit 23, and the liquid crystal. By displaying on the screen or the like (step S509), the user of the mobile communication terminal can recognize whether the fingerprint of the user to be queried is recognized and the registered fingerprint image.

On the other hand, if the fingerprint recognition command for the received inquiry target user is a fingerprint authentication command (step S507), the fingerprint registration / authentication unit 23 is extracted by the fingerprint feature extraction unit 22. After performing the fingerprint authentication of the user to be queried using the fingerprint feature point (step S510), the fingerprint authentication result and the fingerprint image of the fingerprint recognition result of the corresponding user are displayed on the display unit 31 of the management module 30. As a result, the display unit 31 receives the fingerprint authentication result and the fingerprint image from the fingerprint registration / authentication unit 23 as a result of fingerprint authentication for the user to be queried and displays it on the LCD screen ( Step S511), the user of the mobile communication terminal can recognize whether the user to be queried is authenticated and whether the authenticated fingerprint image is recognized.

On the other hand, the feature recognition method is adopted in the fingerprint recognition system and method in the mobile communication terminal according to the embodiment of the present invention, in general, about 100 to 150 feature points are distributed on one finger, all the feature points for each individual Since the type, position, and direction of the appear different, it is determined by the same fingerprint when 50-60% or more of the feature points are matched.

In the above-described fingerprint recognition method in the mobile communication terminal, the fingerprint registration process of the user to be queried is performed according to the fingerprint registration command from the mobile communication terminal user, which will be described with reference to FIG. .

First, the fingerprint input unit 11 of the fingerprint input module 10 included in the mobile communication terminal senses a fingerprint of a user to be queried and receives fingerprint image data (step S601). In this case, the received image data is a normal fingerprint image. It is confirmed whether or not it is data (step S602).

If the image data input by the fingerprint input unit 11 is not normal fingerprint image data, the fingerprint of the corresponding query target user is sensed again to receive the fingerprint image data.

When the image data input from the fingerprint input unit 11 is normal fingerprint image data, the image data is transmitted to the preprocessor 12 to perform preprocessing to extract the feature points of the corresponding fingerprint. In this case, the preprocessor 12 ) Performs noise reduction and histogram smoothing on the fingerprint image data received from the fingerprint input unit 11 and performs preprocessing to emphasize the corresponding fingerprint image (step S603). The fingerprint recognition module receives the preprocessed fingerprint image data. It is stored in the memory unit 21 of (20).

Then, the fingerprint feature extraction unit 22 of the fingerprint recognition module 20 analyzes fingerprint image data stored in the memory unit 21 to extract fingerprint feature points. That is, the fingerprint feature extracting unit 22 determines the initial ridge direction by the projection method that pays attention to the maximum and minimum values of concentration distribution variance with respect to the fingerprint image data, and the ridge of the fingerprint through the calculation of the high accuracy direction by the candidate direction handling. After determining the direction (step S604), the ridge of the fingerprint is extracted through the direct extraction method from the gradation gradation image and the discrimination processing for the region having the large curvature value (step S605).

Then, the fingerprint image binarization is performed by calculating the ridge spacing distribution of the fingerprint extracted above. After performing the fast ridge restoration on the binarized fingerprint image data, the line image having the width of the ridge is 1 pixel by referring to the affix table. The fingerprint image thinning produced is performed (step S606).

Then, the fingerprint feature extracting unit 22 extracts the initial feature points from the binarized and thinned fingerprint image data, and calculates the fingerprint feature points by calculating the feature neighborhood relation by correction and labeling based on arbitrary region reliability and interval information. In step S607, the indeterminate area is determined, and the fingerprint feature point extracted earlier is classified according to the pattern type of the corresponding fingerprint and transferred to the fingerprint registration / authentication unit 23. FIG.

Then, the fingerprint registration / authentication unit 23 stores the fingerprint feature points extracted by the fingerprint feature extraction unit 22 in the fingerprint DB 24 according to the pattern type according to the fingerprint registration command for the user to be queried as a database. The fingerprint of the target user is registered (step S608), and the fingerprint registration result and the fingerprint image, which are the fingerprint registration result of the corresponding query target user, are output through the display unit 31 (step S609).

In addition, in the above-described fingerprint recognition method in the mobile communication terminal, the fingerprint authentication process of the user to be queried is performed according to the fingerprint authentication command from the mobile communication terminal user, which will be described with reference to FIG. Same as

First, the fingerprint input module 10 included in the mobile communication terminal senses the fingerprint of the user to be queried, receives the fingerprint image data, and then preprocesses the fingerprint image data to pass it to the fingerprint recognition module 20 (step S701). The fingerprint feature extraction unit 22 of the fingerprint recognition module 20 analyzes fingerprint image data received from the fingerprint input module 10 to extract fingerprint feature points and transfers the fingerprint feature points to the fingerprint registration / authentication unit 23 (step). S702).

Then, the fingerprint registration / authentication unit 23 selects the matching target fingerprint feature point that is most similar to the inquiry target fingerprint feature point extracted by the fingerprint feature extraction unit 22 according to the fingerprint authentication command for the user to be queried. ) Is read (step S703). Here, the fingerprint feature point to be referred refers to the fingerprint feature point of the user to be queried, and the fingerprint feature to be matched is a fingerprint that is registered in the fingerprint DB 24 as being a collation object to the fingerprint feature to be queried at the time of fingerprint authentication. The registered fingerprint feature point having the feature point most similar to the query target fingerprint feature point among the feature point information.

Then, the fingerprint registration / authentication unit 23 analyzes the fingerprint feature point to be queried to check whether there is a center (step S704). In the case where there is a center, after preliminary coordinate system matching about the center is performed for the inquiry target fingerprint feature point and the collation target fingerprint feature point (step S705), it is checked whether it is satisfied as the feature point matching candidate (step S706).

If, as a result of checking whether it is satisfied as the feature point matching candidate, if it is not satisfied as the feature point matching candidate, the fingerprint registration / authentication unit 23 rejects fingerprint authentication for the corresponding inquiry target user (step S707).

However, as a result of checking whether the feature point matching candidate is satisfied, when the content point matching candidate is satisfied, the fingerprint registration / authentication unit 23 generates the son point number of the central relationship (step S708), Local matching of the fingerprint feature to be collated is calculated to perform local collation (step S709), and the entire correspondence relationship is calculated to collate the entire fingerprint feature points (step S710).

Subsequently, the fingerprint registration / authentication unit 23 obtains a matching score indicating the fingerprint similarity through local matching and total matching of the fingerprint feature point to be compared with the fingerprint feature to be compared (step S711). It is checked whether it is larger than the reference score (step S712). Here, the authentication criterion score refers to a score that is a criterion for authenticating the identity of the fingerprint to be compared with the fingerprint to be checked, and the user of the mobile communication terminal may determine the authentication level of the fingerprint recognition system by adjusting the authentication criterion score. have.

In this case, as a result of confirming that the control score is greater than the authentication reference score, when the control score is larger than the authentication reference score, the fingerprint registration / authentication unit 23 authenticates that the fingerprint to be queried is the same as the verification target fingerprint. In operation S713, this means that the user of the fingerprint to be queried is the same person as the user of the fingerprint to be checked.

On the other hand, as a result of confirming that the control score is greater than the authentication reference score, when the control score is not greater than the authentication reference score, the fingerprint registration / authentication unit 23 does not have the same fingerprint as the verification target fingerprint. It is authenticated (S714). This means that the user of the fingerprint to be queried is not the same person as the user of the fingerprint to be collated.

In addition, the fingerprint registration / authentication unit 23 outputs the fingerprint authentication whether the fingerprint authentication result for the user to be queried and the authentication fingerprint image through the display unit 31 (S715). Here, the authentication fingerprint image means the fingerprint image to be compared with the fingerprint image to be compared.

On the other hand, the embodiment according to the present invention is not limited to the above, it can be carried out by various alternatives, modifications and changes within the scope apparent to those skilled in the art.

As described above, the present invention implements a fingerprint recognition system using a feature point collation method without improving the device specification of a mobile communication terminal having a small storage space and low-speed processing due to the characteristics of the product. It can be implemented as, it is possible to provide user authentication and security of the mobile communication terminal.

Claims (13)

A fingerprint input module that senses a fingerprint of a user to be queried, receives fingerprint image data, and preprocesses the fingerprint image data to output the fingerprint image data; After receiving the preprocessed fingerprint image data from the fingerprint input module, the fingerprint feature points are extracted, the pattern type is classified, and stored in a database to register the fingerprint of the user to be queried, or the user fingerprint feature and the registered fingerprint feature point to be queried. A fingerprint recognition module for authenticating a fingerprint of the corresponding query target user and outputting a fingerprint recognition result for the corresponding query target user; And a management module for displaying a fingerprint recognition result for the query target user output from the fingerprint recognition module and transferring a fingerprint recognition command for the query target user to the fingerprint input module and the fingerprint recognition module. Fingerprint Identification System at Terminal. The method of claim 1, The fingerprint input module, A fingerprint input unit configured to receive fingerprint image data by sensing a fingerprint of a corresponding query target user according to a fingerprint recognition command of the query target user; And a pre-processing unit receiving the fingerprint image data from the fingerprint input unit, removing the noise of the fingerprint, and performing a fingerprint preprocessing using histogram smoothing. The method of claim 1, The fingerprint recognition module, A memory unit for temporarily storing the fingerprint image data preprocessed by the preprocessor; A fingerprint feature extracting unit for extracting fingerprint feature points and classifying a pattern type by binarizing and thinning fingerprint image data stored in the memory unit; A fingerprint DB for dividing the fingerprint feature points extracted by the fingerprint feature extractor into a database by dividing the fingerprint feature points into corresponding database pattern types; According to the fingerprint recognition command for the query target user, the fingerprint feature points extracted by the fingerprint feature extracting unit are stored in the fingerprint DB for each pattern type to register a fingerprint of the query target user or the query target extracted by the fingerprint feature extracting unit. By comparing the fingerprint feature points of the user with the feature points of the registered fingerprints, by measuring the similarity of the types, locations, and directions of the feature points, it is determined whether the fingerprints are the same and the fingerprints of the user to be queried are authenticated. Fingerprint recognition system in a mobile communication terminal, characterized in that comprises a fingerprint registration / authentication unit for outputting. The method of claim 3, wherein The fingerprint feature extraction unit, Calculating a fingerprint direction by applying a gradient method and 8 quantization directions to the fingerprint image data, correcting the fingerprint direction using a weight-based method, and detecting a ridge using a calculation of 9 * 9 pixel units. Fingerprint recognition system in a mobile communication terminal characterized in that. The method of claim 1, The management module, A display unit configured to receive a fingerprint recognition result of the query target user output from the fingerprint registration / authentication unit of the fingerprint recognition module to display whether the fingerprint target user is registered and authenticated and the fingerprint image registered and authenticated; A fingerprint recognition system in a mobile communication terminal, comprising a user interface for receiving a fingerprint recognition request of a user and transferring a fingerprint recognition command to a user to be queried to a fingerprint input unit and a fingerprint registration / authentication unit of the fingerprint input module. . Sensing a fingerprint of a user to be queried according to a predetermined fingerprint recognition command and receiving fingerprint image data to perform preprocessing; Extracting fingerprint feature points by binarizing and thinning the preprocessed fingerprint image data; If the fingerprint recognition command is a fingerprint authentication command for the query target user, performing fingerprint authentication of the query target user using the extracted fingerprint feature points; And displaying a fingerprint authentication result and an authentication fingerprint image as a fingerprint authentication result for the user to be queried. The method of claim 6, If the fingerprint recognition command is a fingerprint registration command for the query target user, storing the extracted fingerprint feature points in a database to perform fingerprint registration of the query target user; The fingerprint recognition method of the mobile communication terminal, characterized in that further comprising the step of displaying the fingerprint registration and the registered fingerprint image as a fingerprint registration result for the user to be queried. The method of claim 6, The process of extracting the fingerprint feature point, Dividing the preprocessed fingerprint image data into a predetermined number of fingerprint image blocks and performing image binarization with different threshold values for each fingerprint image block; And extracting a fingerprint feature point by performing a thinning process to produce a line image having a width of a ridge in a pixel unit from the binarized fingerprint image. The method of claim 8, Performing binarization on the preprocessed fingerprint image data, Determining an initial ridge direction by a projection method paying attention to the maximum and minimum values of concentration distribution variance of the preprocessed fingerprint image data, and determining the ridge direction of the fingerprint by calculating a high accuracy direction by handling candidate directions; Extracting a ridge of a fingerprint through a direct extraction method from a density gradation image and a discrimination process on a region having a large curvature value; The fingerprint recognition method of the mobile communication terminal, characterized in that it comprises the step of performing the fingerprint image binarization by calculating the ridge interval distribution of the extracted fingerprint. The method of claim 8, In the extracting of the fingerprint feature point, the fingerprint direction is calculated in the 8 quantization direction by using the gradient method, the fingerprint direction is corrected using the center of gravity point, and the ridge is detected by 9 * 9 pixel units. Fingerprint recognition method in a mobile communication terminal, characterized in that the extraction. The method of claim 6, The fingerprint authentication of the user to be queried is performed by: Reading a matching fingerprint feature point that is most similar to the extracted query target fingerprint feature point among pre-registered fingerprint feature points; Analyzing the fingerprint feature point to be queried to determine whether it is satisfied as a feature point matching candidate; Rejecting fingerprint authentication for the corresponding query target user when the verification result is not satisfied as the feature verification candidate, and generating the son point number of the central relationship when the verification is satisfied as the feature verification candidate; Obtaining a collation score indicative of fingerprint similarity by performing local collation by calculating local information of the querying fingerprint feature point and the collation target fingerprint feature point, and calculating an entire correspondence relationship to perform collation for all fingerprint feature points; And comparing the obtained control score with a predetermined authentication criterion score and authenticating whether the inquiry target fingerprint and the verification target fingerprint are identical. The method of claim 11, Analyzing the fingerprint feature point to be queried to determine whether there is a center; And, if there is a center, performing a preliminary coordinate system registration with respect to the center of the inquiry target fingerprint feature point and the collation target fingerprint feature point. The method of claim 7, wherein The process of performing fingerprint registration of the inquiry target user, Extracting initial feature points from the binarized and thinned fingerprint image data; Determining an indeterminate region after extracting fingerprint feature points through feature neighborhood relation calculation by a correction and labeling method based on arbitrary region reliability and spacing information; And registering the fingerprint of the user to be queried by classifying the extracted fingerprint feature points according to the pattern types of the corresponding fingerprints and storing them in a database.