JPS6334673A - Finger print collator - Google Patents

Abstract

PURPOSE:To facilitate collation processing by fetching the image of finger print when the superposing of a displayed group on that on an input finger print image is maximum. CONSTITUTION:An ID code inputted from a keyboard 16 goes through a control part 12, and used in accessing a minutiae list 14 to which said ID code is allocated. A minutiae coordinate is fetched from the minutiae list 14, and the minutiae is displayed at a corresponding minutiae coordinate on a monitor 15 via the control part 12. As watching this minutiae on the monitoring picture, an operator inputs a finger print from an image input part 11. The operator moves the position of finger so that the minutiae displayed on the monitor screen is best superposed on that in the input finger print image, a person also moves the minutiae group on the finger print pattern. At the time when the minutiae on the input finger print is almost completely superposed on that on the monitor screen, the finger print image is finally fetched.

Description

Detailed Description of the Invention (Industrial Field of Application) The present invention provides a fingerprint matching device that leaves the user to perform alignment when inputting an object to be inspected having a draft pattern, and automatically collects an image after alignment is completed. Regarding.

(Prior art) Fingerprints have the characteristics of being "unique for everyone" and "unchangeable throughout life."
In order to have extremely high reliability in personal verification,
A method and apparatus for comparing registered fingerprints and input fingerprints using the button recognition technique have been proposed in Japanese Patent Application No. 59-781, Japanese Patent Application No. 59-778, and the like.

In such a fingerprint verification device, a fingerprint image is obtained by applying ink to a finger and stamping it on a paper surface, or an optical image at the interface between the glass and the finger is obtained by applying ink to the finger and stamping it on a transparent surface such as glass. The fingerprint image obtained using boundary changes is I
TV (Industrial Television)
), etc., to perform photoelectric conversion, and extract fingerprint pattern features from this photoelectric conversion image.

Feature points that represent a fingerprint pattern include the coordinate values, directions, relations, etc. of the points where the ridges that make up the fingerprint pattern break, that is, the end points, and the points where the ridges merge or diverge, that is, the branch points (hereinafter referred to as minutiae). is proposed. Here, the relation is defined as the number of ridges existing between the nearest minutiae and the origin in each quadrant of the local coordinate system with each minutiae as the origin. Details of these are described in Japanese Patent Application Laid-Open No. 10525/1983. The following information regarding minutiae is stored in the minutiae list.

(Problems to be Solved by the Invention) Conventionally, in such collation devices, no restrictions were placed on the user when requesting the user to stamp on the placement surface. However, the fingerprint pattern is the same (even if it is from the same finger, it depends on the position and inclination of the finger at the time of stamping), the center position of the pattern stamp is shifted,
Rotation, distortion, etc. may occur, making the alignment between the input pattern and the registered button unstable prior to the verification process.
This had a major impact on the subsequent verification process.

On the other hand, in order to solve the above-mentioned alignment problem, the registered fingerprint image is stored, the registered fingerprint image is presented to the user at the time of fingerprint input, and the user himself/herself operates his or her finger or the device to input the registered fingerprint image. Japanese Patent Application Laid-Open No. 1983-1989 discloses a matching process by collecting the input fingerprint image with the highest degree of matching with the fingerprint image and matching the input image with the registered image.
This is proposed in Japanese Patent No. 24674, but in general, storing images requires a large amount of storage capacity, and the transmission time from the storage medium to the display device f1) also takes a long time in proportion to the size of the image. As expected, the efficiency of the system could not be expected.

(Means for Solving the Problems) In order to solve the above-mentioned problems, the present invention provides means for solving the above-mentioned problems by changing the input fingerprint image and fingerprint feature file obtained when the user moves the fingerprint imprinting position arbitrarily. means for displaying a group of minutiae based on the stored minutiae coordinates on the same screen; means for determining the minutiae coordinates in the fingerprint feature file, their neighboring points, and gray values of arbitrary coordinate points; Means for automatically collecting images when the grayscale value of a displayed feature point group or its neighboring coordinate points exceeds a predetermined threshold value, and the grayscale value of an arbitrary coordinate point that is not a feature point also falls below the threshold value. and a means for comparing the collected input fingerprint image with the minutiae group in the minutiae file, and the comparison process is performed after positioning by the user.

(Function) Fingerprint patterns generally have a spiral shape from the outside to the inside), and a method used for determining the center position and aligning the center prior to the matching process between the input fingerprint and the registered fingerprint has been disclosed in Japanese Patent Application. Publication No. 52-82166, patent application No. 59-181
) No. 06, etc., but these graphic center position determination methods require complex and enormous processing, and are therefore unsuitable for constructing a simple collation processing system.

In order to reduce the huge amount of processing and time required for the above-mentioned alignment, the present invention displays a group of minutiae in a minutiae file, and displays an input fingerprint generated by a user's stamp at an arbitrary position on the same screen. When the user's finger is moved arbitrarily, the image is captured when the number of overlaps between the minutiae group on the input fingerprint image and the displayed minutiae group is the greatest.

(Example) Embodiments of the present invention will be described below with reference to the drawings.

FIG. 1 shows a block diagram of an embodiment of the present invention. An ID code input from a keyboard 16 is passed through a control unit 12, and an O mechanicalist accesses a mechanicalist 14 to which the ID code is assigned. The coordinates of the minutiae inside are passed through the control section 12, and the minutiae is displayed on the corresponding coordinates of the minutiae on the monitor 15. While viewing the minutiae on this monitor, the user
A fingerprint image is input from an image input unit 1) that includes a CCD camera, a light source, etc. Here, the user moves the finger position so that the minutiae displayed on the monitor and the minutiae in the input fingerprint image overlap most, and moves the minutiae group on the fingerprint pattern.

Here, an example of the minutiae group displayed on the monitor screen is shown in FIG. 2(a). Here, each minutiae does not display its direction.

This is because if the positions and directions of all the minutiae are known, it will be possible to reproduce the fingerprint pattern to some extent, but there is a risk that a fake fingerprint will be created that is reproduced from the minutiae information on the monitor.

As a second example of minutiae display, an example in which only a part of the minutiae group is displayed is shown in FIG. 2(b). here,
Displaying only a portion of the minutiae group contributes to reducing the possibility of creating forged fingerprints as described above. Furthermore, by not using the displayed minutiae in the later verification process, even if some of the displayed minutiae produce forged fingerprints, fraudulent verification is prevented.

Third Example of Minutiae Display FIG. 2(c) shows an example in which, in addition to a group of minutiae, dummy minutiae, which are arranged on randomly generated coordinates and have a cryptographic meaning, are displayed. In order to facilitate understanding by comparison with other figures, in FIG. Among the minutiae shown in the figure, only the user knows which minutiae is a dummy minutiae, and this is to prevent unauthorized input by a third party. Methods for arranging the dummy minutiae include arranging them randomly, arranging them regularly with respect to the coordinates of the displayed minutiae, and the like.

Fourth Minutiae Display Example FIG. 2(d) shows an example in which the aforementioned dummy minutiae is added and displayed in addition to a part of the minutiae group. In the figure, a stable verification result is obtained by verification processing using a minutiae that is not displayed and prevention of unauthorized input using a dummy minutiae.

The image acquisition timing of the input fingerprint is determined as follows: 0 Detects the density change of the minutiae coordinates on the repeatedly captured image, and determines the density value of the minutiae pixels including all the minutiae pixels or their neighboring pixels. When the number of pixels whose density exceeds a certain level (that is, becomes dark) exceeds a certain number, image acquisition is performed. That is, the final fingerprint image is captured when the minutiae on the input fingerprint almost overlaps with the minutiae on the monitor.

Alternatively, in images that are sequentially captured, when there is almost no change from dark to light or from bright to dark in pixels at a fixed position, the finger position is assumed to be almost fixed and the final image is There is one that collects images into the memory 13.

Details of the subsequent collation process and feature extrusion process when creating a mechanical list are omitted here because they are described in, for example, Japanese Patent Application Laid-Open No. 55-138174 and Japanese Patent Application Laid-Open No. 56-24675. In this verification process, there is no need to align the input fingerprint and the registered fingerprint.
This reduces the processing time and also affects the matching accuracy.〇 (Effects of the Invention) According to the present invention, a matching system for an object to be inspected exhibiting a pattern using a button recognition technology is provided. It is displayed on the monitor screen of the manualist, and after the user's fingerprint image is stamped, it is displayed on the same monitor screen, and the user moves the finger position to capture the most matching image and perform position matching. It is possible to improve the accuracy of matching and reduce the complexity of the processing, and as a result, it is possible to improve the matching accuracy.

[Brief explanation of the drawing]

FIG. 1 is a block diagram showing one embodiment of the present invention, and FIG. 2 (
a) to (d) are diagrams respectively showing examples of minutiae display according to the present invention. 1)...image input section, 12...control section, 13...
Image menu 7, 14...Manufacturer list, 15...
・Monitor, 16...Keyboard, 17.18,19
．． 21...-v2:s--shi'r, 20. 22...
・Tammy Minutiae 0 Agent Patent Attorney Shinsuke Honjo Law to Figure 1

Claims (4)

[Claims] (1) means for displaying the input fingerprint image obtained and a feature point group based on the feature point coordinates stored in the fingerprint feature file on the same screen when the user moves the fingerprint imprinting position arbitrarily; ; means for determining the gray scale value of the minutiae points in the fingerprint file, their neighboring coordinate points, and arbitrary coordinate points; and; means for automatically collecting an image when the shading value of an arbitrary coordinate point that is not a feature point is smaller than a threshold; means for comparing the collected input fingerprint image with a group of feature points in a fingerprint feature file; A fingerprint verification device comprising: (2) The fingerprint verification device according to claim 1, wherein only a part of the feature point group based on the feature point coordinates stored in the fingerprint feature file is displayed on the display means. (3) Displaying a pseudo minutiae point group on the display means on the minutiae point group based on the minutiae coordinates stored in the fingerprint feature file and the coordinate values randomly generated to give cryptographic meaning. A fingerprint verification device according to claim 1. (4) Display means for displaying a pseudo minutiae point group on a part of the minutiae point group based on the minutiae coordinates stored in the fingerprint feature file and randomly generated coordinate values to give cryptographic meaning. 1. A fingerprint verification device according to claim 1.