JPH0821082B2 - Personal fingerprint matching method - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION [Outline] The present invention relates to a personal fingerprint collation method for collating input fingerprint information with a registered personal fingerprint data file to obtain a collation result in a short time, and a personal identification number with as few digits as possible. In order to provide a method of calling the fingerprint characteristic information of the person and obtaining the fingerprint collation result in accordance with Files and
A personal fingerprint having a collating circuit and applying the input information from the personal identification number input means together with the detected fingerprint information from the fingerprint sensor to the collating circuit to collate with the information of the personal fingerprint data file to obtain the collation result. In the collation method, in addition to the personal fingerprint data file, a search order table capable of searching the storage address in the file by the abbreviated number is created and placed through the abbreviated number grouping / grouping means based on the personal identification number. It is configured by obtaining a matching result by a matching circuit using the abbreviated number input from the identification number input means and the search order table.

TECHNICAL FIELD The present invention relates to a personal fingerprint matching method for matching input fingerprint information with a registered personal fingerprint data file and obtaining a matching result in a short time.

In the conventional fingerprint collation method of this type, as the number of registered fingerprints increases, the number of digits of the identification number for identifying each individual increases, so it is necessary to input a large number of digits for each collation. It took a long time to search and match. Therefore, it was requested to develop a matching method that can obtain the result in a short time.

[Prior Art] With the widespread introduction of computers into social systems in recent years, the interest of system security is gathering interest from related parties. It has been considered and put into practical use to use an ID card or a password or a fingerprint as a means of confirming the identity when entering a computer room or using a terminal connected to a computer. The use of fingerprints is considered to be the most powerful means of personal identification, because fingerprints have two major characteristics: they are "universal" and "invariant for life", and collation systems for personal identification are being studied. .

The verification system includes a method of registering characteristic information of an individual's fingerprint on a card that can be carried by an individual, such as a card, and a method of registering it in a verification device. The former is relatively easy to accept because it does not leave fingerprints to others, but there is a running cost problem because an IC card or memory card is used as the card. The latter is to call the characteristic information of the fingerprint of the person himself / herself from the data file by the personal identification number and to collate with the input fingerprint. FIG. 10 is a block diagram showing a configuration of a conventional method. In FIG. 10, 1 is a personal characteristic information data file, 2 is a matching circuit, 3 is a fingerprint sensor,
Reference numeral 4 is a personal identification number input means, and 5 is an output terminal of the collation result. In the personal characteristic information data file 1, the personal characteristic information of the prospective collation person is detected by the fingerprint sensor 3 in advance, and a large amount of the information is stored together with the identification number. At the time of matching, the fingerprint sensor 3 detects the characteristic of the personal fingerprint and applies it to the matching circuit 2. At the same time, the identification number is input from the personal identification number input means 4 such as a ten-key pad. The matching circuit 2 searches the data file 1 using the inputted identification number as a key, and retrieves the data having the same number. Information from the fingerprint sensor 3 and the data file 1
The collating circuit 2 collates the information from the above and outputs the result of the pass / fail from the terminal 5.

Although it is desirable to shorten the personal identification number within an identifiable range in terms of system configuration, it is often shared with other personal numbers, such as employee numbers in a company, in order to make it easier to remember for each individual.

[Problems to be Solved by the Invention] When it is necessary to use an identification number having a large number of digits for individual identification, sharing with an employee number of 6 to 7 digits is easy to memorize, but each time collation is performed, Entering a large number of digits each time was a hassle, and in the case of a hurry, there was a case where an input error was made. In addition, since the identification number is long, it takes a long time to input, and it also takes a long time to search the data file 1 to find a match of the identification numbers. Therefore, it took an extremely long time to obtain the matching result. .

SUMMARY OF THE INVENTION It is an object of the present invention to provide a method for improving the above-mentioned drawbacks and for calling the fingerprint characteristic information of the person with a personal identification number having the smallest possible number of digits to obtain a fingerprint collation result in a short time.

[Means for Solving the Problems] FIG. 1 is a diagram showing a principle configuration of the present invention. In FIG. 1, 1 is a personal fingerprint data file, 2 is a matching circuit,
Reference numeral 3 is a fingerprint sensor, 4 is a personal identification number input means, 5 is an output terminal of a collation result, and 6 is a search order table.

A fingerprint sensor 3, a personal fingerprint data file 1 for registering the fingerprint detected by the fingerprint sensor 3 together with a personal identification number, and a matching circuit 2 are provided, and a personal identification number input means 4 is provided.
The present invention provides a personal fingerprint matching method in which the input information from the fingerprint sensor 3 is applied to the matching circuit (2) together with the detected fingerprint information from the fingerprint sensor 3 to match the information in the personal fingerprint data file 1 to obtain the matching result. It has the following structure. That is, in addition to the personal fingerprint data file 1, a search order table 6 that allows the storage addresses in the file to be searched by the abbreviated number through the abbreviated / grouped identification number based on the personal identification number is created and placed. The collation circuit 2 obtains the collation result using the abbreviated number input from the identification number input means 4 and the search order table 6.

[Operation] The search order table 6 in FIG. 1 is a file in which a plurality of persons who have become common due to the shortened data regarding the personal identification number are collected into one group and filed. Since the abbreviated number is input from the personal identification number input means 4, the matching circuit 2 searches the search order table 6 based on the number to obtain the order table (arrow (a)). Among the plurality of persons having the same abbreviated number for the table, the address in the data file of the first person is known (arrow (b)). The matching circuit 2 searches the data file 1 based on the address and reads the data at the corresponding address (arrows (c) and (d)).
The collation circuit 2 collates the data from the fingerprint sensor 3, and if yes, repeats from the search for the second column of the search order table 6 by the arrow (a). That is, the data of the person in the order following the previous time is known by the arrow (b), and the data file 1 is read by the arrows (c) and (d). The above method is repeated until a match is obtained by the matching circuit 2.

Since the collation is repeated in this manner, it is convenient to advance the order of the search order table for the data of the matched person so that the next collation can be performed in a shorter time.

[Embodiment] As an embodiment of the present invention, FIG. 2 shows concrete zeros of the search order table in FIG. 1, FIG. 2A shows the initial state, and FIG. 2B shows the table updated after the search and collation. Is shown. Fig. 2A
Shows the case where a 7-digit employee number is used as a personal identification barbaric and shortening is common to the last 3 digits "141". The employee numbers are given in the modified year, educational background, alphabetical order of the family name, etc., but when examining the number of groups that share the last 2 digits, the last 3 digits, the last 4 digits, it is shown in Figure 3. Can be analyzed as
That is, looking at employees whose last three digits are common, there are about 350 groups in which only one person has a certain number set, for example, 001, and there are two persons who have other numbers in common. Is about 150, and the number of common groups of three is about 40. Therefore, when shortened to only the last three digits, there are only a few people who are shared with the same number, and it can be seen that there are many cases where the person can be identified immediately because there is only one person.
Therefore, as shown in FIG.
There are six persons who share the digit 141, and the formal personal identification number and the address (indicated by hexadecimal digits) in the data file in which the data of the person is stored are used as a comparison table. This table is created from the personal fingerprint data file by the shortening and grouping means before starting the collation work. Since the arrangement order in the search order table may be appropriate at the beginning, it is a shortening and grouping means to arrange the personal identification numbers with commonality of only the last three digits, for example.

When collating fingerprint information using the search order table,
The matching circuit 2 in FIG. 1 operates and executes as indicated by arrows (a) to (d).

Next, the search order of the search order table will be described. When the search order is fixed according to the order table shown in FIG. 2A, for example, when matching is performed for the person with the identification number 750141 of rank 4, the useless processing is performed before that and the matching result is gradually obtained. It will be. In order to eliminate the inconvenience, it is desirable to provide a means for updating the search order table. That is, when the result of the search / matching indicates that a matching match is found, the matching circuit judges and raises the order column of the search sequence table to the first column, and the others are 1
Go down column by column to get the table in Figure 2B. If you do this,
Since the person who is frequently used is used so as to be able to be collated quickly, the same person can be searched very quickly in the next search.

FIG. 4 shows an operation flowchart for fingerprint registration of the system of the present invention based on the operation of the matching circuit, and FIG. 5 also shows the matching procedure. At the time of registration, all digits such as employee numbers are used for registration as before, and are written in the search order table in the order of registration. At the time of matching, matching is performed in order from the first rank in the search order table. When the identity is confirmed, the employee's number is re-arranged to the first rank. Further, the search order table is updated by sequentially moving down the existing order table.

Next, by utilizing the fingerprint type classification, the matching time can be further shortened. For that purpose, a registration column for pattern classification mathematics is provided in the index sequence table. FIG. 6 is a diagram showing an example of fingerprint pattern classification. In FIG. 6, the classification of the detection output of the fingerprint sensor is checked to determine which one is classified, and as shown in the example, the numbers are stored in the search order table. FIG. 7A shows the original search order table based on the embodiment, and FIG. 7B shows the case where the table is updated. In FIG. 7A, the shortened identification number is 100, the personal identification number before the shortening is 6 digits, and the number in FIG. 6 is used 1 digit in the pattern number column. An example of shortening is shown below.
Rather than applying the commonality of girders, for example, those who are working on the same floor of the research building and have "patterns" that have characteristics that are significantly different from each other are selected and created To do.

FIG. 8 shows an operation flowchart for creating the fingerprint data file and the search order table in this case.
That is, the fingerprint sensor extracts features, stores them in a data file, and then a person who is accustomed to classify "patterns" registers them while assigning them to the search order table.

When matching is performed using the search order table of FIG. 7A, a shortened number is input, the fingerprint sensor 3 detects a rough classification of patterns, and the detected pattern is input to the matching circuit. As pattern classifications, for example, 1 and 5, 2, 3 and 6, 7 are patterns similar to each other. Therefore, the pattern classifications in the first column of the search order table are compared with each other, and when there is a pattern classification having a significantly different numerical value, , Go to the next column without getting the next data file address. In this way, by utilizing the similarity of pattern classification, the search of the search order table can be performed faster.

FIG. 9 is a flowchart showing a verification procedure including processing by the verification circuit. When the identity is confirmed by collation, it is desirable to update the search order table in the same manner as in the above embodiment.

The feature classification of fingerprints stored in the search order table detects the thickness of ridges and valleys in addition to patterns and places them as digitized numerical values so that people with similar numerical values do not belong to the same group. You can also register.

[Effects of the Invention] According to the present invention as described above, in the method of confirming the person by inputting the personal identification number and collating with the data read from the file, the personal identification number can be shortened without inputting all digits. It is possible to collate accurately in time. It's easy to remember at that time, but since you don't use numbers with many digits,
The effect of avoiding input errors and shortening the matching time is remarkable.

[Brief description of drawings]

FIG. 1 is a diagram showing the basic configuration of the present invention, FIG. 2 is a diagram showing a search order table as an embodiment of the present invention, and FIG. 3 is an example of analyzing employee numbers for application to FIG. FIG. 4, FIG. 4 and FIG. 5 are flowcharts showing the operating procedure, and FIG.
FIG. 7 is a diagram showing an example of fingerprint pattern classification, FIG. 7 is a diagram showing a search order table to which FIG. 6 is applied, FIGS. 8 and 9 are flowcharts showing the operation procedure according to FIG. 7, and FIG. It is a figure which shows the conventional fingerprint collation method. 1 ... Individual fingerprint data file 2 ... Collation circuit 3 ... Fingerprint sensor 4 ... Personal identification number input means 5 ... Collation result output terminal 6 ... Search sequence table

 ─────────────────────────────────────────────────── ─── Continuation of front page (72) Hiroyuki Ikeda Inventor Hiroyuki Ikeda 1015 Kamiodanaka, Nakahara-ku, Kawasaki-shi, Kanagawa Fujitsu Limited (56) References JP-A-60-159980 (JP, A)

Claims (3)

[Claims] 1. A fingerprint sensor (3) and the fingerprint sensor (3)
It comprises a personal fingerprint data file (1) for registering the fingerprint detected by the personal identification number together with a personal identification number, and a matching circuit (2). In addition to the detected fingerprint information from (1), it is applied to a matching circuit (2) and matched with the information of the personal fingerprint data file (1) to obtain a matching result. , A personal identification number input means (4) is prepared by preparing a retrieval order table (6) capable of retrieving the storage address in the file by the abbreviated number through the abbreviated and grouped identification number based on the personal identification number. A personal fingerprint matching method, characterized in that a matching result is obtained by a matching circuit (2) using the abbreviated number inputted from the above and the search order table (6). 2. When the fingerprint collation result of a person is obtained by the method according to claim 1, the collation circuit updates the search order table through the search order table updating means. Personal fingerprint matching method. 3. The personal fingerprint collation method according to claim 1, wherein the search order table is created so as to include personal fingerprint feature classification data in addition to the storage address.