JPS62212781A - Personal identification system - Google Patents

Abstract

PURPOSE:To prevent the generation of illegality due to a third person by collating a personal feature image picked up by a sensor with the detailed data of personal feature images stored in a different file. CONSTITUTION:A face image inherent in a subject 1 is inputted by a sensor 2 and stored in a memory and the feature part pattern of the image, feature data indicating a position and a personal identification code are stored in a memory card 4 to be a personal memory medium possessed by the subject. The feature data are collates with the image of the subject which is stored in the memory 3 to decide whether the feature data indicate the same person or not. When the same person is decided as the result of the collation, a processor 6 decides whether detailed decision is necessary or not on the basis of an access condition. In the detailed decision, data corresponding to the personal identification code of the subject which is read out from the memory card 4 are read out from the detailed face feature data with the personal identification code of the user which are previously stored in the file 8 and compared/ collated with the image of the memory 3. Consequently, illegality due to forgery of a card can be suppressed and danger of misdecision can be reduced.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Application of the Invention] The present invention relates to a maintenance method for a system that targets a plurality of users, such as a room entry/exit control system.

When identifying users using personal characteristic data such as faces and fingerprints in systems such as bank terminals.

The present invention relates to a personal authentication method suitable for reducing system load and preventing abuse by third parties.

[Background of the invention]

Traditionally, the purpose of checking whether a user is a legitimate user is to prevent illegal intrusion into important facilities such as nuclear power plants and computer rooms, or to prevent unauthorized use of files in systems that are intended for multiple users, such as banking systems. 6 The main means of personal authentication are using a secret password or PIN number, using an ID card, and using personal authentication data unique to the individual such as finger, voice, and facial data. There were things to use, etc.

However, methods that use passwords, pin numbers, or ID cards do not have sufficient measures against theft, and
In the case of cards, there is a risk of unauthorized intrusion and unauthorized use due to forgery.

On the other hand, when using personal authentication data such as fingerprints and faces,
In order to reduce the data transmission load between the central file generated when the data is stored in the central file and the authentication input/output terminal, each user is provided with a portable information storage medium such as an IC card, and the medium is individually There is also a method of storing authentication data. An example of this is ``Smart Receipt Cards'' from IEE, Spectrum, February 1984 issue.

The answer to cashless shopping”, pp. 43-49
Page (IREE Spectrum February
19R4゜”5Llart credit cards
: the answer to cashless
If images such as fingerprints and faces are stored as they are using this method, a large amount of memory capacity is required.

Therefore, data indicating characteristics unique to an individual is extracted from the image. Alternatively, the above-mentioned literature considers a method of reducing the amount of data and storing it on a medium by compressing the data while preserving its inherent characteristics.

However, with this method, there is a risk of fraud due to falsification of the data stored on the medium, and the feature data created from the original image generally lacks the details of individual characteristics that are present in the original image, so it is difficult to compare with the original image. Due to the poor ability to identify individuals,
The problem was that the possibility of misjudgment increased.

[Purpose of the invention]

SUMMARY OF THE INVENTION In order to solve the above-mentioned problems, it is an object of the present invention to reduce the load on the required memory capacity of a personal storage medium and to
The objective is to provide a personal authentication method that prevents fraud by individuals.

[Summary of the invention]

In order to achieve the above object, the present invention provides a system for authenticating an individual using an image of an individual's unique characteristics such as a face or a fingerprint taken by a sensor. The present invention is characterized by the provision of a means for comparing the individual characteristic image captured by the sensor with data indicating a portion of the characteristic image stored in the sensor, and a means for comparing the personal characteristic image imaged by the sensor with data indicating details of the characteristic image of the consignee individual stored in a separate file. There is.

In normal personal authentication, partial feature data in a personal storage medium is compared with a sensor input image on a terminal.

Communication load associated with file inquiries can be avoided. Furthermore, the amount of data recorded on a personal storage medium can be reduced.

Access to systems for particularly important transactions, etc.
When necessary, the details of the characteristic image stored in the file are called up to the terminal and verified as the second stage of personal authentication. This can prevent misuse due to forgery of personal storage media. Furthermore, the reliability of personal authentication can be increased.

[Embodiments of the invention]

Hereinafter, one embodiment of the present invention will be described using FIGS. 1 to 5g.

FIG. 1 is an overall configuration diagram of a personal authentication system according to the present invention. The image of the face unique to the subject 1 who is going to use the system 9 to be protected is captured by the sensor 2 and the memory 3
Store in. On the other hand, it is assumed that the memory card 4, which is a personal storage medium held by the subject, has previously stored characteristic data indicating the pattern and position of the characteristic portion of the image and a personal identification code. The card reader 5 reads the above-mentioned characteristic data and personal identification code from the memory card 4. The processing device 6 processes according to the flow shown in FIG.

The feature data is compared with the image of the subject stored in the memory 3 to determine whether they are the same person. (Step 20)
If it is determined in step 21 that they are not the same person, the subject 1 is not allowed to access the protected system 9, that is, enter the facility or use the system. On the other hand, if it is determined that they are the same person as a result of the verification, the processing device 6, it is determined in step 22 whether or not precise determination is necessary. Input bag for this! ! ! 11 to judge the access conditions to the system entered by subject 10 For example, in the case of a banking system, when the transaction amount is greater than a certain value, precise judgment is considered essential.In the precise judgment in step 23, the system stored in file 8 in advance Read out the data corresponding to the subject's personal identification code read from the memory card 4 from among the detailed facial feature data with personal identification codes of multiple users in step 9, and compare and match this with the image in memory 3. do.

At this time, not only the specific characteristic portion registered in the memory card 4 but also the entire image including details is verified.

This enables more accurate judgment than matching using only characteristic parts. As a result, if it is determined in step 24 that the subject is the person in question, the subject 1 is allowed to access the system 9, and in other cases it is not allowed.

Next, the details of the feature pattern matching and determination (step 20) will be described. The image 30 in FIG. 3 is an example of the entire face image input by the sensor. On the other hand, FIG. 4 is an example of an image of facial features read out from a memory card. A personal identification code 50 is stored in the memory card in the configuration shown in FIG.
Also, the X coordinate 51 . of the center point according to the coordinate system 31 of the image for each feature part. Pattern data of Y4!152 and feature points, that is, partial image data 53 are stored. If the image 32 of each characteristic part is displayed at a predetermined position according to these coordinates, it will look like the one shown in FIG. Here, it is sufficient to select nine eyes, mouths, etc. as the characteristic parts, and the comparison judgment is made between the pattern data read out from the memory card for each characteristic part and the image data of the corresponding part of the image 30 by the sensor centering on the X+Y coordinates that were also read out. Calculate the degree of correlation or degree of coincidence between the two and see if it exceeds the threshold. If a certain number or more of the registered characteristic parts match, it is determined that the person is the same person. Note that when inputting images by sensor 2, lighting conditions, position of subject 1, etc. shall be kept as constant as possible;
Fluctuations occur in shooting conditions. Therefore, when making a comparison judgment, image 30
In addition, preprocessing such as removing noise through smoothing processing and emphasizing edges is performed as necessary, and in order to deal with positional fluctuations, the pattern data 32 is moved or rotated within a certain range in the x and y directions to find the peak of correlation. Bye.

In the precise judgment (step 23), detailed data of the entire face image is stored in file 8, and image 3 input by the sensor is used.
0 and calculate the degree of direct correlation.

Alternatively, instead of the entire face image, store a larger number of feature data in the format shown in Figure 5 than the memory card, that is, a number equivalent to the entire face, in file 8, and perform the same processing as in step 20 for these feature parts. Highly accurate determination may also be performed.

The basic configuration and operation of the present invention have been described above, but the following embodiments are also possible.

First, a fingerprint may be used instead of a face as the authentication data. In this case, a branch point or an end point of a ridge may be selected as the characteristic part.

Next, human judgment may be used instead of the automatic comparison judgment by the processing device 6. In that case, the display device 7 displays the sensor input image and the pattern input from the memory card, or the sensor input image and the image read from the file. The images may be overlaid or displayed in parallel on the same screen of the display device 7, and a person may judge whether or not there is a match and input the result from the input device 5ilo.

Further, the characteristic data to be stored in the memory card is encrypted using DES or other method, and a password is inputted from the input device 11 when inputting the memory card, and the password is used to encrypt the data to the processing device 6.
The feature data may be decrypted with , which can reduce the risk of misuse of the memory card.

In addition, in the configuration shown in FIG. 1, it is possible to configure a part other than the protected system 9 to include the file 8 as a terminal device, but the file 8. Of course, parts other than the protected system 9 can be used as terminals, and the file 8 can be shared by a plurality of terminals.

〔Effect of the invention〕

According to the present invention, in personal authentication using faces, fingerprints, etc., a two-step method is adopted: judgment is made by comparing the characteristic parts stored in the memory card, and judgment is made by comparing the overall image stored in the center file. This has the effect of increasing the reliability of personal authentication because it prevents fraud due to forgery and reduces the risk of misjudgment. Furthermore, since the frequency of inquiries to the center file can be reduced, there is an effect of lowering communication costs.

[Brief explanation of drawings]

Figure 1 is an overall configuration diagram of the personal authentication method according to the present invention, Figure 2
The figure shows the processing flow of personal authentication by the processing device, Fig. 3 shows an example of an image of an individual, Fig. 4 shows an example of a personal characteristic pattern, and Fig. 5 shows personal characteristics in a memory card. The structure of the data is shown for 10, 13 and 2.

Claims (1)

[Claims] In a system that authenticates individuals using images of individual characteristics such as faces and fingerprints taken by sensors, data indicating personal characteristic image data taken by sensors and portions of characteristic images stored in portable information storage media. and, if necessary, a process of comparing the personal characteristic image captured by the sensor with detailed data of the individual's characteristic image stored in a separate file.