JPH05135160A - Fingerprint collation device - Google Patents

Abstract

PURPOSE:To provide a fingerprint collation device which speedily calculates a local threshold value close to the density of a valley line. CONSTITUTION:A fingerprint collation device consists of a fingerprint sensor 1 which inputs fingerprint information, a gradation image storage part 10 which converts an output of the fingerprint sensor 1 into a gradation image and stores image, a local threshold value calculating circuit 11 which divides the fingerprint image read out from the gradation image storage part 10 into a lattice shape (measure), further divides the respective measures into small measures, and compares the mean value (measure mean) of density of each measure with the mean value (small measure mean) of each small measure to calculate the local threshold value for the binarization of the measures, a binarization circuit 12 which converts the gradation image into binary data by using the threshold value determined by the local threshold value calculating circuit 11, a fingerprint binary image storage part 13 which is stored with feature points of the binary image and a registering and collating circuit 14 which extracts the feature points from the binary image and registers them as a fingerprint dictionary in a fingerprint binary image storage part 13, and collates the fingerprint dictionary with the fingerprint image for collation.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a fingerprint collation apparatus for comparing an input fingerprint image with a dictionary pattern stored in a dictionary to identify the person.

In recent years, as computers have been introduced into a wide range of social systems, system security has attracted a lot of people concerned. Many questions have been raised in terms of ensuring security for the ID cards and passwords that have been used up to now as a means of confirming the identity when entering a computer room or using a terminal. On the other hand, fingerprints are considered to be the most effective means for personal identification because they have two major characteristics, that is, they are the same for all people and are constant throughout life, and a lot of research and development has been conducted on a simple personal identification system using fingerprints. There is.

[0003]

2. Description of the Related Art FIG. 7 is a conceptual diagram showing the structure of a conventional fingerprint collation device. A fingerprint collation device normally handles a fingerprint as an image. First, the operation at the time of registration will be described. The fingerprint pattern is detected by pressing a finger against the fingerprint sensor 1 and converted into digital data by an A / D converter (not shown) in the fingerprint sensor 1. The converted digital data (fingerprint data) is converted into binary data of “0” and “1” by the subsequent binarizing circuit 2 and stored in the binarizing memory (frame memory) 3.

The fingerprint data stored in the binarized memory 3 is sequentially read out, and then enters the characteristic information extracting section 4 to extract characteristic information. Here, the characteristic information is, for example, as shown in FIG.
It refers to a branch point as shown in (a) or an end point as shown in (b). The fingerprint can be specified by the number of such branch points and end points at which positions. The extracted feature information is stored in the fingerprint dictionary storage unit 5. The above operation is repeated for a plurality of individuals, and the characteristic information of the individuals is stored in the fingerprint dictionary storage unit 5.

When the registration of the feature information amount in the fingerprint dictionary storage unit 5 is completed in this way, the collation operation of the individual fingerprint is started this time. In the case of matching, after inputting his / her own ID number with a ten-key pad (not shown), a finger (a finger used for registration in advance, for example, an index finger) is placed on the fingerprint sensor 1. As a result, the collation unit 6 can narrow the search range by determining the search range of the fingerprint dictionary storage unit 5 based on the ID number and reading it at the time of collation.

A fingerprint pattern is detected in the same manner as at the time of registration, and is converted into digital data by an A / D converter (not shown) in the fingerprint sensor 1. The converted digital data is converted into binary data of “0” and “1” by the subsequent binarization circuit 2 and stored in the binarization memory 3.

The collation unit 6 reads out the fingerprint image for collation stored in the binarization memory 3 and the characteristic information for each individual stored in the fingerprint dictionary storage unit 5, and collates both (pattern matching). I do. When the number of matching feature patterns is greater than or equal to a predetermined number, it is determined that the fingerprints match.

[0008]

In the conventional apparatus as described above, as a method of binarizing the fingerprint image, the input image is divided into a grid pattern, and the average density (cells) in each grid of the divided grids is divided. Eye mean) was used as the local threshold. The binarized image thus obtained often has discontinuous steps at the boundary line of the cells. When registering the feature information as a fingerprint dictionary, thinning and feature extraction are performed on the binarized image, but if there is a step, pseudo feature points such as cracks and whiskers are likely to occur.

In order to avoid the occurrence of such cracks and sweat glands, it is necessary to select a local threshold value close to the density of the valley line and turn the cracks, sweat glands, etc. having a higher density than the valley line to the ridge side. ..
That is, if a density close to the valley line is used as the binarization threshold,
Cracks, sweat glands, etc. will be binarized to "1".

For this reason, as an improvement method, it was necessary to calculate the average of the cells and then collect the pixels (valley line parts) having a value lower than the average to calculate the threshold value of the second stage. In this method, the number of pixels in the valley line portion (<256:
Since each pixel value (1 byte) in the valley portion (2 ⁶ to 2 ⁷ ) has to be compared with the local threshold value by a number close to the case where the grid width is 16 pixels, there is a problem that the processing takes that much time. there were.

The present invention has been made in view of the above problems, and an object of the present invention is to provide a fingerprint collation device capable of rapidly calculating a local threshold value close to the density of a valley line.

[0012]

FIG. 1 is a block diagram showing the principle of the present invention. The same parts as those in FIG. 7 are designated by the same reference numerals. In the figure, 1 is a fingerprint sensor for inputting fingerprint information, 10 is a grayscale image storage unit 10 for converting the output of the fingerprint sensor 1 into a grayscale image and storing the grayscale image, and 11 is a fingerprint image read from the grayscale image storage unit 10. Divide into grids (squares),
Furthermore, each square is divided into small squares, and the average of the density of each square (the average of the squares) is compared with the average value of each small square (the average of the small squares), and the squares are binary. A local threshold value calculating circuit for calculating a local threshold value for converting the local threshold value into a local threshold value calculating circuit 11;
Binarize the grayscale image using the threshold value determined in 2
A binarization circuit, 13 is a fingerprint binarized image storage unit for storing the feature points and the binarized image of the image binarized by the binarization circuit 12,
Reference numeral 14 extracts feature points from the image binarized by the binarization circuit 12 and registers them in the fingerprint binary image storage unit 13 as a fingerprint dictionary, and collates the fingerprint dictionary with the collation fingerprint image. It is a registration / verification circuit.

[0013]

First, the local threshold value calculation circuit 11 operates as shown in FIG.
The fingerprint image 20 is divided into LW × LW squares 21 as shown in FIG. 3, and each square 21 is further divided as shown in FIG.
Divide into small squares 22 of W2 × LW2. When the density is obtained by scanning the small cells in the horizontal direction, the density characteristic shown in FIG. 3 is obtained. The horizontal axis represents distance, and the vertical axis represents concentration.

In the figures, 1 to 4 respectively show the small-grained area 22. The area ~ constitutes one cell. L
1 is the average density of the area, L2 is the average density of the area,
L3 is the average density of the area. Here, it is considered that the density characteristic of the area of is dented near the center is due to the influence of sweat glands. Therefore, if the region is binarized with the threshold value L2, the sweat glands become white and the connection of the ridges becomes worse.

Here, since the local threshold value close to the valley line is set as the threshold value of the mesh, the local threshold calculation circuit 11 determines whether the average of each small mesh in FIG. 3 is smaller than the average L of the meshes. When the specified number of small-cell averages smaller than L are obtained, the average (valley-line part average) of them is output as the local threshold value for the cells. The local threshold value thus obtained has a density close to the valley line. As described above, according to the present invention, it is possible to quickly calculate the local threshold value close to the density of the valley line.

[0016]

Embodiments of the present invention will now be described in detail with reference to the drawings. FIG. 4 is a configuration block diagram showing an embodiment of the present invention. The same parts as those in FIG. 1 are designated by the same reference numerals. The local threshold value calculation circuit 11 includes a grid average calculation circuit 3
0, small square average calculation circuit 31 and local average comparison circuit 32
It is composed of The gray-scale image storage unit 10 is connected to the mesh average calculation circuit 30, and the mesh average calculation circuit 31 further divides the mesh given from the mesh average calculation circuit 30 into meshes. The local average comparison circuit 32 is a grid average calculation circuit 3
0 and the output of the small square average calculation circuit 31 are compared to select a local threshold value. The other circuits are the same as those in FIG. The operation of the circuit thus configured will be described below.

(At the time of registration) The fingerprint image read by the fingerprint sensor 1 is converted into multivalued digital data and stored in the grayscale image storage unit 10. The mesh average calculation circuit 30 reads out the fingerprint image data stored in the grayscale image storage unit 10, divides the fingerprint image data into a grid pattern, and calculates an average density value for each of the divided meshes.

On the other hand, the small-cell averaging circuit 31 further divides the cells given from the cell-average calculation circuit 30 into small cells,
Based on the density data read from the grayscale image storage unit 10 for each small grid, the density average value is obtained.

The local average comparing circuit 32 compares the average of the cells given by the average calculating circuit 30 with the density average of each of the small averages given by the average calculating circuit 31 to obtain the average of the averages. Count the number of small square averages with a density less than the value. When a specified number of small-mesh average values smaller than this one are obtained, the average of these small-mesh averages is calculated (valley line average), and the obtained valley-line average is calculated for that mesh. Output as the threshold value in

The binarization circuit 12 is a local threshold calculation circuit 11
It receives a threshold value for each cell given from the above, and binarizes the image in the cell using the threshold value. The binarized image is stored in the fingerprint binary image storage unit 13. In this way, when the binarized data of all the fingerprint images are stored in the fingerprint binarized image storage unit 13, the registration / collation circuit 14 extracts the feature points and registers them in the fingerprint binarized image storage unit 13. To do. (At the time of collation) The fingerprint image read from the fingerprint sensor 1 is
It is stored in the grayscale image storage unit 10 as multivalued density data. The procedure for binarizing the fingerprint image data stored in the grayscale image storage unit 10 is the same as that at the time of registration.
The binarized collation fingerprint image is the fingerprint binarized image storage unit 13
Stored in.

The registration / collation circuit 14 performs pattern matching between the collation fingerprint image stored in the fingerprint binary image storage unit 13 and the fingerprint dictionary (feature point window image). If the matching error at the time of the pattern matching is smaller than a predetermined allowable value, it is determined that they match, and if it is larger than the allowable value, it is determined that they do not match.

In the above-described embodiment, when the number of small-cell averages smaller than the cell average as the threshold value is larger than the predetermined number, the average value thereof is used as the local threshold value in the cell. I took it as an example. However, the present invention is not limited to this.

FIG. 5 is an explanatory diagram of another method of calculating the local threshold. Let TH1 be the average of the squares and TH2 be the average of the valley lines, and T
The value of the point that internally divides between H1 and TH2 at a predetermined ratio can also be used as the local threshold. For example, as shown in the figure, the value TH at the point where TH1 and TH2 are internally divided into M: N is used as the local threshold. An optimum value can be selected for the internal division ratio, if necessary.

In the above embodiment, the case where the rectangle is used as the lattice shape is taken as an example, but the present invention is not limited to this. Triangles can be used as the grid, as shown in FIG. In this case, the present invention can be applied by using the minimum unit of triangle as a small lattice.

[0025]

As described above in detail, according to the present invention, the grid pattern of the fingerprint image is further divided into small grids, and the average value of the small grids is smaller than the average value of the grid grids. When the number of small ones is counted and the number is larger than the reference value, by using the average value of those small squares as the local threshold value, the local threshold value close to the density of the valley line can be quickly calculated. A fingerprint collation device can be provided.

According to the present invention, by bringing the threshold value near the valley line, it is possible to perform binarization in which ridge cracks, sweat glands, etc. are regarded as the ridge side. According to the present invention, the comparison of the valley line portion is
Since the average of small squares can be compared to the maximum, the number of small squares (64 squares when the width of the squares is 2 pixels and the number of small squares is 64) can be compared with the average of squares, so cracks more quickly than the conventional method. ，
It is possible to obtain an appropriate binarized image that does not cause sweat glands.

[Brief description of drawings]

FIG. 1 is a principle block diagram of the present invention.

FIG. 2 is a diagram showing division of squares and small squares.

FIG. 3 is a diagram showing generation of a threshold value close to a valley line.

FIG. 4 is a configuration block diagram showing an embodiment of the present invention.

FIG. 5 is an explanatory diagram of another calculation method of a local threshold.

FIG. 6 is a diagram showing another example of the shape of a square.

FIG. 7 is a conceptual diagram showing the configuration of a conventional fingerprint matching device.

FIG. 8 is a diagram showing an example of fingerprint characteristic information.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 Fingerprint sensor 10 Grayscale image storage unit 11 Local threshold calculation circuit 12 Binarization circuit 13 Fingerprint binary image storage unit 14 Registration / collation circuit

Claims (3)

[Claims] 1. A fingerprint sensor (1) for inputting fingerprint information.
And a grayscale image storage section (10) for converting the output of the fingerprint sensor (1) into a grayscale image and storing the grayscale image, and dividing the fingerprint image read from the grayscale image storage section (10) into a grid pattern (squares). Further, each square is divided into small squares, and the average of the density of each square (the average of the squares) is compared with the average value of each small square (the average of the small squares) to determine the number of squares. Local threshold calculation circuit (11) for calculating a local threshold for binarization
A binarization circuit (12) for binarizing the grayscale image using the threshold value determined by the local threshold value calculation circuit (11), and an image binarized by the binarization circuit (12) The fingerprint binarized image storage unit (13) for storing the feature points and the binarized image, and the fingerprint 2 as a fingerprint dictionary by extracting the feature points from the image binarized by the binarization circuit (12). Value image storage unit (13)
A fingerprint collation device configured by a registration / collation circuit (14) that registers the image in the fingerprint dictionary and collates the fingerprint dictionary with the fingerprint image for collation. 2. The local threshold calculation circuit (11) divides a grayscale image into a grid pattern, calculates a grid average for each grid, and divides each grid into finer small grids. Calculate the small-mesh average for each eye and check if they are smaller than the average for each mesh, and when the specified number of small-mesh averages are obtained, those averages (valley line part average) ) Is set as a local threshold value in the mesh, and the fingerprint collation apparatus according to claim 1, wherein. 3. The local threshold calculation circuit (11) internally divides the grid average and the valley line average at a predetermined ratio, and calculates the density of the internally divided point as a local threshold. The fingerprint collation device according to claim 1, wherein