JPS60221882A - Collating system of person himself - Google Patents

Abstract

PURPOSE:To identify a person himself by collating feature points, etc., of a fingerprint due to microcomputer processing through use of only picture data readout by straight line scanning with respect to a fingerprint pattern on the picture. CONSTITUTION:A fingerprint input device 11 generates a photoelectric transfer picture of a fingerprint pattern by means of a light source and an image pickup device such as an ITV by utilizing optical boundary conditions of glass from the reverse surface with respect to a finger placed on a glass plate. The picture signal is A/D converted 12 and stored in a memory 13, while information regarding feature points and the relation between them on a minutiae card readout by a card reader 15 is held in a working memory 103 of a collating device 10, which properly matches a fingerprint picture stored in the memory 13 to the feature point group and the relation between feature points readout from the minutiae card. The successful matching and unsuccessful matching are decided to be coincident and dissident, respectively.

Description

[Detailed Description of the Invention] K1jυL The present invention relates to a person verification method, and particularly to a person verification method for identifying an individual using a fingerprint.

LL1 Personal verification systems that identify individuals are broadly divided into two types: access control and transaction control. In the former, when you want to enter a specific area, instead of using a key, you enter the personal identification number (ID number) given to each person using a keyboard or card, and check it against the ID number registered in advance at the center. This is a method to check the identity of an individual. The latter type of transaction management is used for commercial transactions by banks and so-called credit cards, and is a method in which the identity of an individual is confirmed, file checks, updates, etc. are performed using the same method as access management.

But long. However, none of the above systems is perfect, and accidents such as unauthorized use have been occurring frequently with the recent increase in credit transactions. Therefore, a more reliable and low-cost identity verification method is required.

1 shellfish! An object of the present invention is to provide an accurate, highly reliable, and low-cost identity verification method that is suitable for access control and transaction management systems.

11 animals 1 za The person verification method according to the present invention includes a conversion means that collects a fingerprint from a finger tip and converts it into image data, and a storage card that stores information indicating the minutiae of the fingerprint and the relationship between the minutiae points in advance. a card reading means for reading the stored information of the memory card when the card is read, and a collation means for checking and collating the information from the card reading means from the image data by the converting means, and the converting means and the card reading means are plural. The present invention is characterized in that the verification means is provided in one center device, and these terminal devices and the center device are connected by a communication line.

Song Xl [ Hereinafter, the present invention will be explained in detail using the drawings.

FIG. 1 is a block diagram of an embodiment of the present invention, which has a plurality of terminal devices 20.1 to 20-n and one center device 10, and a communication line 16 is connected between each shaft and the center. They are each connected.

Each terminal includes a card reader 15 that reads stored information when a memory card is input, a fingerprint input device 11 that inserts the tip of a finger to collect a fingerprint, an A/D (analog/digital) converter 12, and an image It is composed of a memory 13, and each part is connected by signal lines for exchanging various data as shown in the figure. - Fingerprint input device @11 utilizes the optical boundary conditions of the glass from the back side to connect the finger placed on the glass plate to the light source and the interface.
A photoelectric conversion image of a fingerprint pattern is generated using an imaging means such as a TV. For example, it is disclosed in Japanese Unexamined Patent Publication No. 54-69300 and Japanese Unexamined Patent Publication No. 54-85600, and these configurations are utilized. (I) A/D converter 12 and image memory 1 that convert light and conversion signals
As for 3, a commonly known one can be used, so the details thereof will be omitted.

The digital two-dimensional quantized image data obtained in this way is stored in the memory 13, and information regarding the feature points (end points, branch points) and relations between feature points on the so-called minutiae card read by the card reading device 15. is the working memory 10 in the matching processor 10
3. The fingerprint features are the position (X, Y) of the minutiae point M and the direction of the pattern, as shown in FIGS. 2(a) and (b).

1 in units of relation R1 to R4 with neighboring feature points
This is a list representation of a plurality of minutiae points M in a fingerprint, and is recorded in advance on the minutiae card by an electromagnetic method or the like.

FIG. 4(a) shows an example of a fingerprint, and FIG. 2(b) shows the concept of a list. Detailed definitions and extraction methods are described in Japanese Patent Application Laid-open No. 138174/1983.

As shown in FIG. 3, the matching processor 10 accurately matches the fingerprint image 30 stored in the image memory 13 with the feature point group (for example, 31) and relation (for example, 32) read out from the minutiae card. It all comes down to this. If the response is successful, it is a match, otherwise it is a mismatch, and management decisions can be made for various purposes.

The Begi matching process performed by the matching processor 10 is as follows.

Referring to FIG. 4, a plurality of scans 411 perpendicular to the direction of the feature point with the feature point position as the origin are performed on the local pattern 41 on the image memory corresponding to the position of the feature point 40 read from the minutiae card. ...412.413.
414, 415, 416, ..., 417, and inspect the image data obtained according to this scanning, for example, 412.413, ..., 416 in Figure (a) are shown in Figure (b).
422, 423, ..., 426. The feature points (endpoints in the example) 43 are discovered by matching the pattern edges associated with the dotted lines of each image data.

Furthermore, the feature direction is determined by averaging the matching angles of pattern edges. The feature points 43 and 40 thus discovered are held as candidate corresponding points based on their positional displacement and direction difference. Candidate corresponding points are obtained in the same manner for neighboring feature points that are related to this feature point by relations in the feature point list. Next, the fifth
As shown in the figure, the fingerprint pattern line 53 on the connection line between two neighboring minutiae points 51 and 52 is counted as Yl-3 as shown in figure (b), and checked against the list.

As described above, one fingerprint is verified by sequentially inspecting the arrangement of minutiae and the relations between neighboring minutiae points.

The matching processor 10 can be configured exclusively for the above processing, but as described above, this matching processing is realized only by local linear scanning on the image memory, so as shown in FIG. , medium heat processing device 101, program memory 1
02, a working memory 103, and an external interface 104 may also be used.

As described above, using only the image data read out by linear scanning of the fingerprint pattern on the image, the identity of fingerprint minutiae and relations between minutiae points can be verified with high accuracy through simple microcomputer processing. As a result, the identity of the individual can be confirmed.

The image data sent from each terminal to the center has approximately 160 bytes of information in the case of flat fingerprint data, so if this continues, the communication line 16 will require extremely advanced characteristics. It turns out. Therefore, if the A/D-converted fingerprint image data is compressed using a well-known data compression method such as a predictive encoding method and sent to the center, the amount of data to be sent can be approximately 80 bytes. It's very convenient. In this case, it goes without saying that it is necessary to add a data decompression function on the center side.

As described above, according to the present invention, since a fingerprint is used for verifying the person, it becomes an extremely reliable verification method, and a general-purpose microprocessor such as a microcomputer can be used. Therefore, it is a small and low-cost system, and in particular, it is a person verification system that replaces the current cash card system for deposit and savings transactions.

[Brief Description of the Drawings] Figure 1 is a block diagram of an embodiment of the present invention, Figure 2 is a diagram defining minutiae information of a fingerprint, Figure 3 is a diagram showing a fingerprint verification state, and Figure 4 is a diagram of a fingerprint FIG. 5 is a diagram illustrating an inspection of feature point arrangement, and FIG. 5 is a diagram illustrating an inspection of relations between feature points. Explanation of symbols of main parts 10... Verification processor 11... Fingerprint input device 12... A/D converter 15... Card reader Applicant Japan Denki Co., Ltd. Representative Patent Attorney Shin Yanagawa Figure 1 Figure 2 ((1) Figure 3 Figure 5 Figure 4 (0) (b)

Claims (1)

[Scope of Claims] - (1) A conversion means for collecting a fingerprint from a finger tip and converting it into image data, and a memory card in which information indicating minutiae and minutiae stiffness of the fingerprint are stored in advance are input. a card reading means for reading the stored information of the memory card; and a collation means for verifying and collating the information from the card reading means from the image data by the converting means, wherein the converting means and the card reading means A person verification method characterized in that the verification means is provided in each of a plurality of terminal devices, the verification means is provided in one center device, and these terminal devices and the center device are connected by a communication line. (2 ) The image data is compressed using a predetermined data compression method and sent from the terminal, and the center device is configured to decompress this compressed data. Matching method.