JPH0518152B2 - - Google Patents

Description

[Detailed Description of the Invention] (Industrial Application Field) The present invention relates to a fingerprint verification device, and in particular, identifies an individual by determining the identity of an input fingerprint and a registered fingerprint based on two fingerprint pattern characteristics. The present invention relates to a fingerprint verification device.

(Prior art) In systems that use fingerprints to identify individuals, pattern recognition technology is used to identify collected fingerprints and to ensure extremely high reliability due to the characteristics of fingerprints, which are ``unique for all people'' and ``unchangeable throughout life.'' Many proposals have already been made for devices that automatically verify identity with previously registered fingerprints, including Japanese Patent Applications No. 54-39648 and No. 57-111114.

On the other hand, with the progress of the information society in recent years, there has been an increasing need for personal identification when accessing computer terminals and databases and when managing key entrances and exits, and these are done using PIN numbers and ID cards. However, it is easy to lose its effectiveness due to theft or loss, and it must be said that its reliability is low, so a fingerprint verification device has been considered.

(Problems to be Solved by the Invention) The conventional verification devices using personal identification numbers or ID cards described above have a higher risk of theft and loss than fingerprint verification, and are extremely unreliable.

The purpose of the present invention is to eliminate the drawbacks of conventional verification devices, and to provide a simple and high-performance fingerprint verification device for personal identification, which has a function to distinguish between fingerprints forged using rubber or the like and human fingerprints. It is about providing.

(Means for Solving the Problems) The fingerprint matching device of the present invention determines whether an input fingerprint image and a registered fingerprint image are the same fingerprint based on pattern characteristics of the two fingerprint images. The fingerprint matching device for determination includes an image input unit that collects a fingerprint image of a fingerprint placed on a transparent body and photoelectrically converts the image, and an A/D converter that converts the photoelectrically converted signal into two-dimensional quantized image data. , detect and store the gray scale values of pixels within a predetermined area in order to extract image features in the image, observe changes in gray scale values of the fingerprint image within the area for a certain period of time, and determine whether the fingerprint is A placement inspection unit that confirms that the placement on the transparent body is completed and outputs a signal indicating completion of placement, and the time after receiving the inspection output and the information that has been taken within a certain period of time after this time. detecting a gray value of a pixel in the region in the image and comparing it with the stored gray value;
a forged input determination unit that determines whether a normal fingerprint has been placed; and a gray level value of a pixel at the predetermined position or a new position after the determination unit determines that the input is normal; an input image collection control unit that collects an input image when
an image storage unit that stores dimensional quantized image data;
a verification processing section consisting of a central processing unit, a program memory, a working memory and an external interface;
A file that stores a feature list that registers the positions of minutiae extracted from registered fingerprints, pattern directions, and the number of ridges existing between neighboring minutiae points, and an input for inputting a file identification number to access the feature list. The pattern direction of each feature point and the number of ridges between neighboring feature points in the feature file accessed from the identification number from the input section are inputted from the image input section. A straight line in a predetermined direction passing through the feature points in the input fingerprint image determined to be a normal fingerprint by the forgery input determination unit is scanned, a cutting pattern of the image cut by the straight line is extracted, and a matching process is performed. It is characterized by doing the following.

(Example) Next, an example of the present invention will be described with reference to the drawings.

FIG. 1 shows an embodiment of the invention. In FIG. 1, one embodiment of the present invention is an image input unit 1 that collects a fingerprint image of a fingerprint placed on a transparent body and converts it photoelectrically.
0 and an A/D converter 11 that converts the photoelectric conversion signal from the image input unit 10 into two-dimensional quantized image data.
, a placement inspection unit 12 that detects and stores the density value of a pixel at a predetermined position in the collected image, and inspects and outputs that the fingerprint has been placed on the transparent body based on the change in density; Detecting the grayscale value of the pixel at the predetermined position in the image taken at a time after receiving the inspection output of the unit 12 and within a certain period of time after this time, and comparing it with the stored grayscale value, A forged input determination unit 13 determines whether a normal fingerprint has been placed, and a gray value of a pixel at a predetermined position or a new position after the determination unit 13 determines that the input is normal is equal to or higher than a predetermined threshold value. When , an input image acquisition control unit 14 that acquires an input image, an image storage unit 15 that stores two-dimensional quantized image data,
A file 23 that stores a maneuver list in which the positions of minutiae points (hereinafter referred to as "maneuvers") extracted from registered fingerprints, pattern directions, and the number of ridges existing between neighboring minutiae points (hereinafter referred to as relations) are registered. , an input section 21 for inputting the identification number assigned to access the manual list in the file 23, an output section 22 for outputting the verification results and various messages, and an input section for inputting image data in the image storage section 13. A verification processing unit 2 that performs verification with the manual list from the file 23 obtained by the identification number 21 and sends out the result.
Including 0.

The matching processing section 20 includes an image section storage 15 and an input section 2.
1, an external interface 16 connected to an output section 22 and a file 23, a central processing unit 17, a program memory 18, and a working memory 19 for storing a manual list from a file 21.

Regarding the fingerprint input unit 21, a photoelectric conversion image of a fingerprint pattern is inputted using an imaging device such as an ITV (industrial television) using an optical boundary condition caused by light from a light source on a finger placed on a transparent body. For example, devices described in Japanese Patent Application Laid-open No. 54-69300 and Japanese Patent Application Laid-open No. 54-85600 are available and used.

Next, the image input unit 10 determines that the fingerprint placement is complete by checking the temporal change in density of pixels at a predetermined position in the image that is captured at any time after the identification number is input from the input unit 10 for manualist access. At the same time as detection, when the gray value of the pixel exceeds a threshold value, a fingerprint image is automatically collected.

The A/D conversion section 11 converts the photoelectric conversion signal from the image input section 10 into two-dimensional quantized image data.
The fingerprint placement detection unit 12 detects and stores the shading value of a pixel at a predetermined position in the collected image, and checks whether the fingerprint has been placed on the transparent body based on the change in shading,
It is output to the forgery input determination section 13. The forgery input determination unit 13 detects the time after receiving the test output and the gray value of the pixel at a predetermined position in the image captured within a certain period of time after this time, and stores it in the fingerprint placement detection unit 12. It is then compared with the gradation value to determine whether a normal fingerprint has been placed. This method utilizes the fact that when a human fingerprint is input, sweating occurs within a certain period of time and the shading value of the fingerprint image changes. This change does not occur in the case of forged fingerprints. The input image acquisition control unit 14 is the determination unit 1
When the grayscale value of a pixel at a predetermined position or a new position after being determined to be a normal input in step 3 exceeds a predetermined threshold value, an input image is sampled to obtain two-dimensional quantized image data.

After the multivalued two-dimensional quantized image data obtained in this manner is written in the image storage section 15, the collation processor 20 accesses it from the file 23 using the manipulator identification number inputted from the input section 21. Save the manual list to working memory 1.
Store in 9.

Figures 2a and 2b show the fingerprint pattern and manual list in this embodiment. In FIG. 2, the fingerprint features are located at the position
This is a list representation of a plurality of manicures M existing in one fingerprint image, with Y, fingerprint direction D, and relations R ₁ , R ₂ , R ₃ , and R ₄ with neighboring manicures as units. The details of these definitions and extraction methods are detailed in JP-A No. 55-138174.

As shown in FIG. 3, the collation processor 20 uses the grayscale fingerprint image 30 stored in the image storage unit 15 and the manifold (for example, 31) and relation (for example, 32) in the manual list accessed from the file 19. Processing is performed to accurately match the If a correspondence is established, it is considered a match, and if not, it is considered a mismatch, and the individual is identified for various purposes.

FIG. 4 shows the maneuver detection method according to this embodiment. In FIG. 4, the verification processor 20 performs verification processing on a local gradation pattern 41 on the image storage unit 15 corresponding to the position of the manifold 40 in the manipulative list read from the file 23, using the manipulative position as the origin, A plurality of scans 411,..., 412, 413, perpendicular to the pattern direction
..., 416, ..., 417 and associate the binary rasters obtained according to this scanning, for example, 412, 413, ..., 41 in Fig. 4a.
6 becomes like 422, 423, . . . , 426 in FIG. 4b. Magnitudes (for example, end points) 43 are detected by associating the binary rasters. Manufacturers 43 and 40 detected in this manner are saved as candidate points based on their positional displacements.
Candidate points can be obtained in the same manner for neighboring maneuvers that are related to this maneuver by relations in the maneuver list. Next, as shown in FIGS. 5a and 5b, the number of ridges on the connecting straight line, that is, the relation 53, for the two neighboring manuals 51 and 52 is as shown in FIG.
Yi=3 is counted as shown in FIG. In this way, the matching process is performed by sequentially detecting the arrangement and relation of the maniture on the fingerprint image. Finally, the matching results are displayed on the output section 20. As described above, this embodiment provides a simple fingerprint verification device that detects the arrangement of manicures and the relationship between neighboring manicures using image data obtained by performing linear scanning on a gray-scale fingerprint image.

In this embodiment, the fingerprint image stored in the image storage unit 15 is a multivalued image, but a binary image that has been binarized by a known method is stored and that image is processed. However, the essence of the process remains unchanged.

[Brief explanation of the drawing]

FIG. 1 is a block diagram showing one embodiment of the present invention;
FIG. 2 is a diagram showing the concept of the manipulatives and manipulative list in this embodiment, FIG. 3 is a diagram showing the verification state in this embodiment, FIG. 4 is a diagram showing manipulative detection in this embodiment, and FIG. The figure is a diagram showing relation detection in this embodiment. 10... Image input section, 11... A/D conversion section,
12... Fingerprint placement inspection unit, 13... Fake fingerprint input determination unit, 14... Input fingerprint image collection control unit, 15
...Image storage unit, 16...External interface,
17...Central processing unit, 18...Program memory, 19...Work memory, 20...Verification processor, 21...Input section, 22...Output section, 23...
file, 30... shading fingerprint image, 31... manicure, 32... crayon, 40... list storage manipulative, 41... local shading pattern, 43... detection manipulative, 411-417
... Scanning line, 422-426 ... Associated raster, 51, 52 ... Neighborhood manifold, 5
3... Relation.

Claims (1)

[Claims] 1. In a fingerprint matching device that determines whether an input fingerprint image and a registered fingerprint image are the same fingerprint based on the pattern characteristics of the two fingerprint images, an image input unit that collects a fingerprint image and performs photoelectric conversion; an A/D conversion unit that converts the photoelectric conversion signal into two-dimensional quantized image data; Detects and stores the gray value of the pixel,
a placement inspection unit that observes changes in the density value of the fingerprint image within the area for a certain period of time, confirms that placement of the fingerprint on the transparent body is completed, and outputs a signal indicating completion of placement; The time after receiving the test output and the gray value of the pixels in the area in the image captured within a certain period of time after this time are detected and compared with the stored gray value to determine whether the fingerprint is normal. a forgery input determination unit that determines whether or not the input has been placed; and a forgery input determination unit that determines whether or not the input is normal; At the time, an input image acquisition control unit that acquires an input image, an image storage unit that stores the two-dimensional quantized image data, a matching processing unit that includes a central processing unit, a program memory, a working memory, and an external interface; A file that stores a feature list that registers the positions of feature points extracted from fingerprints, pattern directions, and the number of ridges existing between neighboring feature points, and an input section for inputting a file identification number for accessing the feature list. and an output section that outputs the matching results, and the pattern direction of each feature point and the number of ridges between neighboring feature points in the feature file accessed from the identification number from the input section are inputted from the image input section. Scanning a straight line in a predetermined direction passing through the feature points in the input fingerprint image determined to be a normal fingerprint by the forgery input determination unit, and extracting a cutting pattern of the image cut by the straight line;
A fingerprint verification device characterized by performing verification processing.