KR20010096440A - Ten fingerprint system for the identification - Google Patents

Abstract

PURPOSE: A fingerprint detecting apparatus and method are provided to efficiently detect fingerprints by comparing a limited registering fingerprint data with an input fingerprint data using twelve fingerprint code and data. CONSTITUTION: A fingerprint input section(610) receivers a fingerprint. A key input section(620) executes or controls a fingerprint detector. A processor(630) performs a pattern recognition operation with respect to the input fingerprint and makes a fingerprint code and a fingerprint data. The processor(630) transmits the fingerprint code and data to the server(700) through a communication network(800). A screen display device(640) displays a processing state from a fingerprint input to an information output. A server(700) judges whether the fingerprint code and data are identical with registered fingerprint code and data. A permission device(710) stores a certified signal in a database(730).

Description

Fingerprint recognition device and method using fingerprint fingerprint and fingerprint data {Ten fingerprint system for the identification}

The present invention relates to a fingerprint recognition device and method using a fingerprint fingerprint data and fingerprint data, particularly when comparing the fingerprint input and the fingerprint registered in advance a large number of fingerprints using the fingerprint code assigned to ten fingers first Doubt fingerprint code that automatically identifies the same fingerprint by comparing the fingerprint data of the input fingerprint with the registered fingerprint data using the fingerprint data created by extracting the feature points of the fingerprint after limiting the fingerprint to be compared. And a fingerprint recognition device and method using fingerprint data.

Everyone has different shapes, and even ten fingers of the same person all have different shapes. In addition, the shape of the fingerprint formed at birth is maintained until death, the epidermis is peeled off the fingerprint shape, but the new epidermis is formed, the fingerprint shape is revived, it is also reproduced in the same shape as before. Because of these characteristics, fingerprints are most widely used in personal identification and criminal identification worldwide.In recent years, the use of fingerprint recognition systems is becoming more common due to the development of related technologies.In the US, crime prevention as well as various threats are prevented. As an effective and inexpensive tool, the method using fingerprint recognition is widely used.

The fingerprint recognition system can be broadly divided into an identification system and a verification system for determining whether to enter or leave a place requiring security. Inquiry system is based on a large database, and the rejection rate should be almost 0%. It is to inquire who the party of the fingerprint is. The verification system is a system that judges the identity within a limited number of people. It is used everywhere. Therefore, in the former case, a more efficient fingerprint recognition method is required due to the enormous number of registered fingerprints. To this end, it is essential to minimize the number of registered fingerprints to be compared by classifying the input fingerprint into a specific category. For this purpose, the prior art fingerprint recognition system generally uses a classification method based on the feature points of the fingerprint.

Classification by fingerprint feature points is to use the feature points present in the fingerprint. The ridge is a linear shape of the inner surface of the gelatinous branch with the correct directionality, and the feature points are the shortcomings of the ridge progressing and breaking, the branching point where the ridge progresses, the short ridge, the isolated point of the ridge, and the triangle. About 100 to 150 are distributed per finger, and the type, location, and direction of the finger are different for each person and each finger. Therefore, by using the information on the feature point as the fingerprint recognition means, if the positional relationship and the direction of the feature point match 50 to 60% or more, it is usually determined to be the same fingerprint.

When comparing the input fingerprint and the registration fingerprint by the method of classification by the feature point, classify the feature point into the normal area, the special area, the discontinuous area, etc., and then compare the feature point area of the input fingerprint with the feature point area of the registered fingerprint. The comparison is made by calling fingerprints with feature points. However, this method has a problem in that the area of the feature point must be compared until a fingerprint matching a large number of registered fingerprints is determined once. In addition, the prior art fingerprint inquiry system first selects a preprocessing process for extracting a feature point of a fingerprint, a direction pattern used for combination and discrimination, a feature point extraction process, and a recognition process using the same. There is a problem in that it takes a very long time to perform a search and retrieval without a specific category classification of a fingerprint to be processed at the time, and a large amount of calculation to be processed. In addition, for this reason, real-time processing has not been performed, and the performance of the recognition system is not ideal yet, and most of them rely heavily on visual discrimination by experts.

In fact, the Korean National Police Agency operates the automatic fingerprint recognition system for the purpose of inquiry, but due to the reasons mentioned above, only the special convict extracts the characteristic points of the wormwood and operates the system that recognizes the fingerprints within a limited range. This situation relies on visual discrimination by the human eye, which is why it takes a long time until the fingerprint is judged and the case processing is delayed.

However, there are several advantages to using a thumbprint rather than a single fingerprint. First of all, the error rate becomes small with exponential power. For example, if the probability of discriminating one fingerprint as another person is 10%, the probability of discriminating two fingerprints as another person is reduced to 10% × 10% = 1%, and when using ten fingerprints, there is almost no error rate. By using these advantages, even in extracting feature points of a fingerprint, a smaller amount is extracted than when using a single fingerprint. Extracting a small amount simplifies operation and speeds up the error, but increases the error rate. However, if you use all the dug, the error rate is reduced by the exponential power, so there is no problem with fingerprint recognition.

The reason for assigning a code to each of the ten fingers by dividing it into 10 patterns is to reduce the number of fingerprints to be compared in advance when comparing the input fingerprint image with a large number of pre-registered fingerprints. In this case, the number of fingerprint codes that can come out with ten fingers is 10,000,000,000 (10 powers of 10) assuming that the same code does not come out. That is, 10-digit code can manage 10 billion fingerprints. Using such a fingerprint code has a systematic classification system, so even if a large fingerprint exists, the data can be easily found. In other words, if a code is assigned in the order of left hand finger, left hand finger, left hand finger, left hand finger, left hand finger, right hand finger, right hand finger, right hand finger, right hand finger, right hand finger, The probability that all codes are the same is significantly lowered, reducing the number of fingerprints to be compared to a limited range. For example, if the fingerprint code of the input fingerprint is 64275-64456, the first digit is 6, the second digit 4, the third digit 2, the fourth digit 7, the fifth digit 5, the sixth digit 6, the seventh digit 4, the eighth digit is 4, the ninth digit is 5, the tenth digit is registered 6 only calls.

In the conventional fingerprint recognition device, the fingerprint must be identified by one fingerprint, and thus, the extraction of the fingerprint feature point requires a fairly complicated calculation process and a lot of data. In the present invention, since the data of all ten fingers are used, the amount of data of the feature points extracted from each finger is smaller than that of a single fingerprint, and the ten fingers are compared, so that more accurate recognition can be performed.

Representative methods for evaluating the performance of the device in fingerprint matching are the acceptance rate of the other person and the rejection rate. Acceptance of another person means that when a fingerprint has been compared with a fingerprint that has already been registered, it is wrongly judged as the person even though it is not the same person. However, fingerprints are flexible as part of the body, so they can look different and damaged every time they are collected. Low-quality fingerprints frequently occur, and even when taking fingerprints of the same person, they may look different depending on conditions such as temperature and humidity. The acceptance rate or rejection rate of others cannot be ideal. However, in the present invention, even if the rejection rate reaches 10% per fingerprint, the rejection rate of the entire ten fingerprints becomes 10 million of 10%, which is 0.00000001%. In other words, there is little rejection rate. Also, the acceptance rate of others is likewise almost zero. By using a plurality of fingerprints as described above, the accuracy of fingerprint recognition is very high. In addition, the data of the singularity extracted from the fingerprint can use a smaller amount of data than other devices performing a single fingerprint.

Existing fingerprint identification systems have the trouble of registering the fingerprint separately and verifying the identity, and the method of recognizing the fingerprint without the line classification that limits the fingerprint to a specific range when comparing the input fingerprint with the registered fingerprint. It is useful when the data is small, but it is not suitable for processing very large data for the whole country, and there is a problem that cannot be confirmed in real time.

As such, identification systems using existing fingerprints are cumbersome to register the fingerprint separately and verify the identity, and recognize the fingerprint without preclassification limiting the fingerprint to a specific range when comparing the input fingerprint with the registered fingerprint. This method is useful when data is small, but it is not suitable for processing very large data for the whole country, and there is a problem that cannot be confirmed in real time.

Accordingly, an object of the present invention is to use the data of the stamped seal when making the first national ID card to solve the above problems, to recognize the input fingerprint and convert it into a ten-digit personal fingerprint code to register a large number of registration fingerprints After classifying the target fingerprints to be used for comparison by classifying them into a specific range, the fingerprint data of the limited registration fingerprint and the input fingerprint are compared to perform efficient fingerprint recognition, and the same fingerprint is identified in real time by comparing fingerprints at high speed. The present invention provides a fingerprint recognition device and method using a dozen fingerprint code and fingerprint data.

According to an embodiment of the present invention, fingerprint recognition using a fingerprint code and fingerprint data to identify the same fingerprint in real time compared to a plurality of fingerprints registered in advance by using the fingerprint code and fingerprint data of the ten fingers to be input A device for dividing the input fingerprint image into units of a predetermined pixel area, and then determining a shape of a central portion of the fingerprint and whether or not a triangle is present; and 10 patterns for each shape determined by the pattern determination unit. And a code granting unit for assigning a code to each of them and making it into a ten-digit fingerprint code, a fingerprint data generating unit for extracting a feature point of each fingerprint after attaching the code, and creating fingerprint data, and a fingerprint after generating the fingerprint data. Sends code and fingerprint data to server, compares and discriminates fingerprint codes from pre-registered database The fingerprint code and the fingerprint data extracting unit for extracting the fingerprint code and the corresponding fingerprint data, and the fingerprint data of the matching fingerprint codes registered with the fingerprint data of the fingerprint code input by extracting the fingerprint code and fingerprint data Comparing the same and includes the same fingerprint determination unit to determine whether or not the final fingerprint.

According to another embodiment of the present invention, fingerprint recognition using a fingerprint code and fingerprint data to identify the same fingerprint in real time compared to a plurality of fingerprints registered in advance using the fingerprint code and fingerprint data of the ten fingers to be input As a method, after dividing the input fingerprint image into units of a predetermined pixel area, a pattern determination step of determining the shape of the center of the fingerprint and the presence of a triangular figure, and 10 patterns for each shape determined in the pattern determination step A code granting step of assigning a code to each code and making it into a ten-digit fingerprint code, a fingerprint data generating step of extracting a feature point of each fingerprint after applying the code to create fingerprint data, and a fingerprint after generating the fingerprint data Sends code and fingerprint data to server and compares fingerprint codes in pre-registered database. A fingerprint code and a fingerprint data extracting step of discriminating and matching a fingerprint code and corresponding fingerprint data; and extracting the fingerprint code and the fingerprint data and matching fingerprint codes previously registered with the fingerprint data of the input fingerprint code. Comparing the fingerprint data of the two and comprises the same fingerprint determination step to determine whether the final fingerprint.

1 to 10 is a method of applying a code for each fingerprint type used in the present invention

11 is a block diagram of a fingerprint recognition method using a fingerprint fingerprint and fingerprint data according to an embodiment of the present invention

12 is a block diagram of a fingerprint recognition device using a fingerprint fingerprint code and fingerprint data according to an embodiment of the present invention

<Explanation of symbols for the main parts of the drawings>

510: terminal unit 511: pattern recognition determination unit

512: code granting unit 513: fingerprint data generation unit

520: server unit 521: fingerprint code and fingerprint data extracting unit

522: identical fingerprint determination unit

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

1 to 10 relates to a method for assigning a fingerprint for each type of fingerprint used in a fingerprint recognition device using a fingerprint code and fingerprint data of the present invention. When a pattern of a fingerprint as shown in FIG. 1 is recognized, a triangular dot does not exist and the fingerprint is assigned a code number of 1 if the center of the fingerprint has a bow-like appearance. Both the shape of Fig. 1 (a), which is a smoothly flowing bow shape, and (b) of Fig. 1 (b), a shape of a protruding bow, have no triangular points, and these fingerprints are all unusual fingerprints. In addition, when the pattern of the fingerprint is recognized, one triangulation exists, and the central ridge is formed in the shape of a horse hoof, and the fingerprint as shown in FIG. When the triangular diagram 201 is formed on the left side and the fingerprint of the rain is taken, the triangular diagram 202 is formed on the right side, and the fingerprint of the first grade and the left-handed fingerprint such as (a) of FIG. When triangular diagram 203 is formed on the right side and the fingerprint is taken on the right side, triangular diagram 204 is formed on the left side of FIG. Code number 2 is assigned to the first grade of the first grade because the left and right sides can be distinguished. This is because a small number of people with the fingerprint of the first grade are given code number 2 to all the first grade. As in FIGS. 3 to 6, the Euljong Sangmun is given a code according to the fusion between the inner ends 311, 321, 331, 341 and the outer ends 312, 322, 332, 342. That is, in the case of Euljong Seongmun, many people have a virtual line between the inner and outer ends (311) and the outer end (312) (313) (323) (333) by subdividing the code into the number of ridges between the inner and outer ends. Draw (343) and give the code to the number of ridges in between. 3, if the number of ridges between the inner end 311 and the outer end 312 is seven or less, the code number 3 is assigned, and the ridge between the inner end 321 and the outer end 322 as shown in FIG. If the number of is between 8 and 11, the code number 4 is given, and as in the case of FIG. 5, if the number of ridges of the inner end 331 and the outer end 332 is between 12 and 14, the code number 5 If the number of ridges between the inner end 341 and the outer end 342 is 15 or more, as shown in FIG. 6, the code number 6 is assigned. Also, the patterns of FIGS. 7 to 10 are called vortices in the classical classification, which are fingerprints having two or more triangular diagrams 411, 412, 421, 422, 431, 432, 441, and 442. Say. Since this phenomenon is a pattern of many people, it is further divided into codes. Recognize the pattern of the vortex sentence and classify as shown in FIGS. 7 to 10 to give a code. Draw a virtual straight line 414, 424, 434, 444 between the end points 413, 423, 433, 443 and the standard points 414, 424, 434, 444 of the tracking line. The code number is given by counting the number of ridges in contact with the straight line. More specifically, the tracking lines 415 and 425 enter the inside of the right triangle 412 and 422 as shown in FIGS. 7 and 8 and the tracking lines 435 and 445 as shown in FIGS. 9 and 10. The right triangles 432 and 442 flow outwards. When four or more ridges exist between the end point 413 and the standard point 412 of the tracking line as shown in FIG. When four or more ridges exist between the end point 443 and the standard point 442 of the tracking line as shown in FIG. 10, the code number 9 is assigned and the end points 423 and 433 of the tracking line as shown in FIGS. Code number 8 when three or less ridges exist between the reference point and the standard points 422 and 432. And if the finger is not cut due to cutting the ground or other reasons, and if the classification reference point is damaged and normal classification is impossible, the code number is assigned as 0. Thus, the code number is assigned to the left, left, stop, left ring, When a fingerprint code of 00000-00000 form is completed in the order of left handed, left handed, right handed, right handed, right handed, right handed, and handed out, the number of registered fingerprints is limited using this code for fingerprint identification. Secondly, the feature points of the fingerprint are compared and examined.

FIG. 11 is a block diagram of a fingerprint recognition method using a fingerprint code and fingerprint data according to an embodiment of the present invention. The apparatus includes a pattern determination step 511, a code granting step 512, and a fingerprint data generation step 513. ), The fingerprint code and fingerprint data retrieval step 521 and the same fingerprint determination step 522.

In addition, the pattern determination step 511 is to determine the shape of the center of the fingerprint and the presence of the triangle and the position and shape of the feature point to receive the digital image signal of the input fingerprint through the camera, scanner or fingerprint sensor, etc. By the pattern recognition operation, the shape of the center of the fingerprint and the number of triangles are determined, and the position and shape of the feature point to be used in the fingerprint data generator are determined. The code assigning step 512 assigns a code to each of the ten fingerprints by using a result calculated in the pattern determination step, and generates a ten-digit fingerprint code. Fingerprint data generation step 513 is a place to generate fingerprint data using the position and shape of the pattern recognition feature pattern in the pattern determination step 511 and thus the fingerprint code and fingerprint data generated in the terminal unit 510 is shown in FIG. The fingerprint code of the server unit 520 and the fingerprint data retrieval step 521 are transmitted through the communication network 800 of FIG. In the fingerprint code and fingerprint data retrieval step 521, the fingerprint code of the input fingerprint is compared with the fingerprint code stored in the database 523 of the server unit 520. Extract the data. In the same fingerprint determination step 522, the extracted fingerprint data and the sent fingerprint data are precisely compared to determine whether or not the final fingerprint is identical.

In this way, a large number of registered fingerprints is limited to a specific range and compared with input fingerprints, thus eliminating unnecessary comparison processes, thereby enabling real-time processing and reducing a fingerprint recognition error rate.

12 is a block diagram of a fingerprint recognition device using a fingerprint fingerprint and fingerprint data according to an embodiment of the present invention, such a device includes a terminal 600, a server 700, a communication network 800.

In addition, as illustrated, the terminal unit 600 includes a fingerprint image input unit 610, a key input unit 620, a processor unit 630, and a screen display unit 640. The fingerprint image input unit 610 is a part for inputting a fingerprint. The fingerprint is input through the scanner 611, the camera 612, the fingerprint sensor 613, and the like, and the image data of the input fingerprint is combined into the processor unit 630. do. The processor unit 630 extracts the center point, the triangle angle, and the feature point by performing pattern recognition on the input fingerprint, and makes a fingerprint code by using the relationship between the shape of the center point, the number of triangles, and the number of ridges. Code is generated and fingerprint data is created by the location and size of feature points. The fingerprint code and the fingerprint data thus made are sent to the server unit 700 through the communication network 800. The screen display unit 640 is a part for outputting the situation and the result that proceeds from the input of the thumbprint to the output of the personal information and the key input unit ( 620 is a part for executing or controlling a fingerprint recognition device using a fingerprint fingerprint code and fingerprint data. The server 700 receives the fingerprint code and the fingerprint data sent from the terminal unit and determines the same fingerprint of the fingerprint sent using the registered fingerprint code and the fingerprint data. 710, server 720, and database 730. Permit device 710 is a part that allows access to only the authenticated signal to the database 730 and the server 720 compares the fingerprint code sent from the terminal unit 600 and the fingerprint code registered in the database 730 Thus, when there is a matching fingerprint code, the fingerprint data of the matching fingerprint code is extracted from the database 730, and the final identical fingerprint is determined through a precise comparison with the fingerprint data sent from the terminal unit.

As described above, the present invention, in the process of extracting the same fingerprint as the input fingerprint when there is a large amount of fingerprint, using only the fingerprint code and the fingerprint data to extract only a specific range of fingerprints of the enrolled fingerprint and precisely using the fingerprint data By comparing fingerprints, it is possible to quickly and accurately identify personal information and receive and display personal information through a communication network. Therefore, it is advantageous to provide an apparatus and method for quickly and accurately identifying identification in police or prosecution services. have.

Claims (4)

A fingerprint recognition device using a fingerprint fingerprint and fingerprint data to identify the same fingerprint in real time compared to a plurality of fingerprints registered in advance by using the fingerprint code and fingerprint data of the ten fingers, the fingerprint of the input fingerprint After dividing the image signal into units of a predetermined pixel region, the pattern determination unit determines the shape of the center of the fingerprint and the presence of a triangular figure, and divides the image signal into 10 patterns according to the shapes determined by the pattern determination unit. And a fingerprint data generation unit for creating a fingerprint data by extracting a feature point of each fingerprint after granting the code and assigning it to a ten-digit fingerprint code, and after generating the fingerprint data, the fingerprint code and the fingerprint data to the server. Compares and distinguishes fingerprint codes from registered database and sends matching fingerprint codes The fingerprint code and the fingerprint data extracting unit for extracting the corresponding fingerprint data are extracted, and the fingerprint data of the input fingerprint code and the registered fingerprint data of the registered fingerprint code are compared with the fingerprint data of the fingerprint code and the fingerprint data. Fingerprint recognition device using a fingerprint fingerprint code and fingerprint data including the same fingerprint determination unit for determining whether the fingerprint The fingerprint fingerprint and fingerprint of claim 1, wherein the image signal of the fingerprint of the fingerprint is determined by dividing the image signal of the fingerprint of the fingerprint into a region having a size of 9 × 9 unit area. Fingerprint reader using data The fingerprint recognition apparatus according to claim 1, wherein the pattern classification unit classifies the pattern by the shape determined by the pattern determining unit and adds the pattern to the N code unit to add N patterns to the fingerprint unit. Fingerprint recognition method using a fingerprint and fingerprint data to identify the same fingerprint in real time compared to a plurality of pre-registered fingerprints using the fingerprint code and fingerprint data of the ten fingers to input the predetermined fingerprint image After dividing into pixel unit of, the pattern determination step of determining the shape of the center of the fingerprint and the existence of the triangular shape, and classified into 10 patterns for each shape determined in the pattern determination step, and give a code to each A code granting step of making the ten-digit fingerprint code, and a fingerprint data generation step of extracting a feature point of each fingerprint after the code assignment, and creating fingerprint data; and sending the fingerprint code and the fingerprint data to the server after the fingerprint data is generated. Compares and discriminates fingerprint codes from pre-registered database and matches fingerprint codes. A fingerprint code and a fingerprint data extracting step of extracting the fingerprint data corresponding to the fingerprint data are extracted, and the fingerprint data of the input fingerprint code is compared with the fingerprint data of the matching fingerprint codes previously registered by extracting the fingerprint code and the fingerprint data. Fingerprint recognition method using a fingerprint fingerprint data and a fingerprint fingerprint comprising the step of determining the same fingerprint to determine whether the same fingerprint.