JP2001126067A - Device and method for collating picture, memory for storing picture collation processing program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To reduce a load concerning collation processing and to reduce collating time in a picture collating device for collating fingerprinting picture data, e.g. SOLUTION: When a collating fingerprint picture is read, a registered fingerprint picture previously stored and registered in a memory is read and the collating fingerprint picture is normalized corresponding to the number of elements and pixel pitch of this registered finger print picture. After then, both of the fingerprint pictures of the registered finger print picture and the collating fingerprint picture are compress- processed into 1/4. Then, by the collating judgment using the calculation of the correlation coefficient between the compressed registered finger print picture and the collation fingerprint picture, judgment display of coincidence and non-coincidence of a registered fingerprint and a collating fingerprint is executed. Thus, even in the case of calculating the correlation coefficient of each pixel on the collating fingerprint picture with the registered fingerprint picture as a reference, the registered finger print picture and the collating fingerprint picture to be the objects of collation judgment are compress- processed to be 1/4, thereby a load with the calculation processing is reduced and time required for the whole fingerprint collation is reduced.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image collating apparatus for collating fingerprint image data, an image collating method, and the like.
And a storage medium storing an image matching program.

[0002]

2. Description of the Related Art Image data of a subject's fingerprint (registered fingerprint image) registered in advance in a database or the like is compared with image data of a subject's fingerprint acquired this time by an image sensor (collation fingerprint image). 2. Description of the Related Art There is a fingerprint collation device that performs individual confirmation of a subject.

There are various types of specific procedures for collating a registered fingerprint image and a collated fingerprint image by this fingerprint collating apparatus. One of the procedures is one of a plurality of rectangular areas arranged on the registered fingerprint image. A plurality of rectangular areas on the reference fingerprint image where the maximum correlation coefficient is calculated with the reference image in each range are detected, and the positional relationship between the plurality of rectangular areas on the registered fingerprint image and the position on the collation fingerprint image are detected. By comparing and comparing the difference between the plurality of rectangular areas and the positional relationship detected in the above, a fingerprint is determined to be a match if the difference is smaller than a preset threshold, and is determined to be a mismatch if the difference is larger than the threshold. A collation apparatus has been filed by the same applicant (Japanese Patent Application No. 10-372205).

In calculating the correlation coefficient of the registered fingerprint on the reference fingerprint image with respect to the reference image in the fingerprint verification apparatus, a rectangular area having the same size as that of the reference image is formed by one pixel in the horizontal and vertical directions on the reference fingerprint image. This is calculated each time raster scanning is performed, and a rectangular area where the maximum correlation coefficient is obtained is detected as an area corresponding to the reference image of the registered fingerprint.

[0005]

However, the calculation of the correlation coefficient on the reference fingerprint image with respect to the reference image of the registered fingerprint is not only complicated and time-consuming, but also requires the same calculation as the reference image. Since a large area is raster-scanned one pixel at a time on the collation fingerprint image and the calculation is performed each time the scanning is performed, it takes a lot of time to complete the calculation of the correlation coefficient for the entire area of the collation fingerprint image. However, there is a problem that the time included from when the image of the collation fingerprint image is read to when the determination result is obtained cannot be practically ignored.

SUMMARY OF THE INVENTION The present invention has been made in view of the above-described problems, and an image collating apparatus, an image collating method, and an image collating method capable of reducing a load relating to a collating process and greatly reducing a collating time. An object of the present invention is to provide a storage medium storing a collation processing program.

[0007]

According to a first aspect of the present invention, there is provided an image collating apparatus for acquiring an image serving as a reference for collation, and a reference image acquired by the reference image acquiring means. Collation image acquisition means for acquiring an image to be collated with, image compression means for compressing both the reference image acquired by the reference image acquisition means and the collation image acquired by the collation image acquisition means, The image processing apparatus further includes an image comparison unit that performs image comparison based on the reference image and the comparison image compressed by the compression unit.

In such an image collating apparatus, a reference image which is a reference for collation acquired by the reference image acquiring means, and
The collation image to be collated with the reference image acquired by the collation image acquisition unit is both compressed by the image compression unit, and image collation is performed based on the compressed reference image and the collation image. The load associated with pixel-by-pixel matching between images is reduced.

According to a second aspect of the present invention, there is provided the image collating apparatus according to the first aspect, wherein the image collating means is provided on the compressed reference image of the compressed reference image. And an image matching means for calculating the correlation coefficient in step (1) to perform image matching.

In such an image collating apparatus, since the correlation coefficient of the compressed reference image on the compressed collation image is calculated and collation of the image is performed, one pixel of the reference image is collated on the collation image. Even when calculating the correlation coefficient for each,
By compressing the reference image and the collation image to be collated, the load involved in the calculation process can be reduced.

[0011]

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

FIG. 1 is a block diagram showing a configuration of an electronic circuit of a fingerprint matching device according to an embodiment of the image matching device of the present invention.

This fingerprint collation device includes a control unit (CPU) 1
1, storage device 12, RAM 13, image reading device 14, display unit 15, input unit 16, storage medium reading unit 1
8. It is composed of a transmission control unit 19, and is mutually connected via a bus 17.

The control unit (CPU) 11 receives a control program stored in advance in the storage device 12 or an external storage medium 18a such as a floppy disk FD from a floppy disk drive F in response to an input signal from the input unit 12.
The storage device 1 via a storage medium reading unit 18 such as a DD
2 or the fingerprint matching program read from the storage device 22 of the external computer terminal (program server 21) to the storage device 12 via the transmission control unit 19 via the communication network 20. It starts up and controls the operation of each part of the circuit using the RAM 13 as a work memory.

The storage device 12 is constituted by a ROM of a semiconductor memory, a hard disk device HD, or the like, and in addition to previously storing the above-described control program read out by the control unit (CPU) 11 immediately after power-on of the present verification device. And a registered fingerprint image of the subject read in advance by the image reading device 14.

The RAM 13 is a work memory used by the control unit (CPU) 11 when executing the above-described control program.

The image reading device 14 is a device for acquiring a registered fingerprint image or a collation fingerprint image of a subject, and is, for example, an image scanner device or an image sensor such as a CCD (Charge Coupled Device).

The display unit 15 is a display device, such as a CRT or a liquid crystal display, for displaying a result of collation between the registered fingerprint image and the collation fingerprint image.

The input unit 16 is an input device such as a keyboard device for the user of the present verification device to instruct the control unit (CPU) 11 to obtain a fingerprint image or to start a fingerprint verification process described later. .

FIG. 2 is a diagram showing a data memory secured in the RAM 13 of the fingerprint collating apparatus.

The RAM 13 has a read image memory 13
a, a registered fingerprint compressed image memory 13b, a collated fingerprint image memory 13c, a collated fingerprint rotated image memory 13d, a collated fingerprint compressed image memory 13e, a registered fingerprint position memory 13f, a collated fingerprint position memory 13g, a rotation angle memory 13h, etc. .

In the read image memory 13a, the image data of the fingerprint of the subject read by the image reading device 14 is temporarily stored as a multi-tone image.

In the registered fingerprint compressed image memory 13b, the image data of the fingerprint of the subject to be registered read by the image reading device 14 is registered in the storage device 12 while being registered in the storage device 12. The image data of the registered fingerprint is compressed to a quarter size and stored.

The collation fingerprint image memory 13c stores the fingerprint image data of the subject to be collated read by the image reader 14 as a multi-tone image. The collation fingerprint image data is stored in the storage device. 12 is normalized and stored according to the number of pixels and the pixel pitch of the registered fingerprint image data stored and registered.

The rotated fingerprint image memory 13d stores image data that has been subjected to rotation processing when the matched fingerprint image data stored in the matched fingerprint image memory 13c is rotated by Δθ.

The collated fingerprint compressed image memory 13e stores collated fingerprint image data at the beginning of image reading stored in the collated fingerprint image memory 13c or collated fingerprint image data after rotation processing stored in the collated fingerprint rotated image memory 13d. Is compressed to a quarter size and stored.

In the registered fingerprint position memory 13f, coordinate data indicating the defined position of each of a plurality of rectangular areas defined for the quarter-compressed registered fingerprint image data stored in the registered fingerprint compressed image memory 13b. It is memorized.

The collation fingerprint position memory 13g stores the partial image of the registered fingerprint in the range of each of a plurality of rectangular areas defined for the quarterly-compressed registered fingerprint image data in the registered fingerprint compressed image memory 13b. 4 in the collation fingerprint compressed image memory 13e in which the correlation coefficient of
A plurality of rectangular regions are detected on the collated fingerprint image data of the one-half compression, and coordinate data indicating the positions of the plurality of detected rectangular regions on the collated fingerprint image data of the quarter compression are stored. .

The rotation angle memory 13h stores the current rotation angle θ of the verification fingerprint image data stored in the verification fingerprint rotation image memory 13d.

FIG. 3 is a diagram for explaining a quarter compression process of fingerprint image data by the fingerprint collating apparatus.

That is, the quarter compression processing of the image data in the fingerprint collation apparatus is performed by 3 pixels × 3 pixels of the original image.
The process of compressing nine pixels into one pixel is repeated every other pixel in both the vertical and horizontal directions to reduce the total data amount to four.
As shown in FIG. 3, the data compression method is a method of data compression in a region of 9 pixels (i−1, j−1) to (i + 1, j + 1) to be subjected to unit compression. The grayscale data (signal intensity) of each pixel is multiplied by the corresponding weight and added, and this is used as the grayscale data (signal intensity) of the central pixel (i-j). Corner pixel (i-1, j-1) (i + 1, j-1) (i-1, j)
+1) (i + 1, j + 1) has a weight of 16
The pixel at the center of the top, bottom, left and right (i, j-1) (i-1,
j) (i + 1, j) (i, j + 1) is set to 1/8, and the center pixel (i, j) is set to 1/4.

That is, in this compression, four regions a, b, c, and d of four pixels each including upper, lower, left, and right always including the center pixel (i, j) are set for nine pixels of three pixels vertically and three pixels horizontally. This is based on the idea that the average of the average gradation data (signal intensity) in the a, b, c, and d regions is taken as one pixel data. In this case, the four corner pixels (i-1, j-1) (i + 1, j-1) in the entire process of compressed pixel data calculation
The data of (i-1, j + 1) (i + 1, j + 1) is added only once in each of the four pixel areas a, b, c, and d, and the pixel (i, j-1) (i- 1,
j) (i + 1, j) (i, j + 1) data is common to the adjacent four pixel regions ab, cd, ac, bd.
Since the data of the center pixel (i, j) is added to the four pixel regions a, b, c, and d four times in common in the upper, lower, left, and right directions, the total number of added pixel data is 16 Is 1/16, 1/8 (= 2/16), 1/4 (= 4
/ 16).

FIG. 4 is a diagram showing an example of compressing a fingerprint image in the fingerprint collating apparatus.

FIGS. 4 (A1) and (A2) show 25
FIG. 4 shows read fingerprint images of 6 gradations and 128 gradations.
(B1) and (B2) are corresponding fingerprint images after quarter compression.

That is, the registered fingerprint image stored and registered in the storage device 12 or the collated fingerprint image stored in the collated fingerprint image memory 13c in the RAM 13 is as shown in FIG.
When the fingerprint image is as shown in (A2) and (A2), the RA
The fingerprint image stored in the registered fingerprint compressed image memory 13b or the collation fingerprint compressed image memory 13e in M13 is a quarter-compressed fingerprint image as shown in FIGS. 4 (B1) and 4 (B2).

Next, the operation of the fingerprint collating apparatus having the above configuration will be described.

FIG. 5 is a flowchart showing the fingerprint collation processing by the fingerprint collation apparatus.

The fingerprint image (uncompressed image) of the subject to be registered read by the image reading device 14 is stored and registered in the storage device 12 in advance.
The fingerprint image of the subject to be collated is image reading device 1
4 and written into the read image memory 13a in the RAM 13, and transferred to and stored in the collation fingerprint image memory 13c (step S1).

Then, the registered fingerprint image stored in the storage device 12 is read out, and the matching fingerprint image memory 13 is stored in accordance with the number of pixels and the pixel pitch of the registered fingerprint image.
The collation fingerprint image stored in c is normalized (steps S2 and S3).

The registered fingerprint image read from the storage device 12 is compressed to a quarter by the above-described quarter image compression processing, and stored in the registered fingerprint compressed image memory 13b in the RAM 13. (Step S4).

Then, the verification fingerprint image stored in the verification fingerprint image memory 13c is also compressed to a quarter by the same quarter image compression processing as described above, and stored in the verification fingerprint compressed image memory 13e (step). S5).

Then, the registered fingerprint compressed image memory 13
The collation processing of the quarter-compressed registered fingerprint image stored in the memory b and the quarter-compressed fingerprint image stored in the collated fingerprint compressed image memory 13e is performed according to a flowchart shown in FIG. (Step S6).

In this collation processing, if the coincidence between the registered fingerprint compressed image and the collated fingerprint compressed image is not determined (step S7), the initial collated fingerprint image read from the collation fingerprint image stored in the collation fingerprint image memory 13c (step S7). After the uncompressed image is rotated by a fixed angle Δθ and stored in the collation fingerprint rotated image memory 13d (step S8 → S9),
The data is compressed to one-half and stored in the collation fingerprint compressed image memory 13e (step S5).

Then, the same collation processing is performed between the rotated and compressed collated fingerprint image stored in the collated fingerprint compressed image memory 13e and the registered fingerprint compressed image stored in the registered fingerprint compressed image memory 13b. Is performed (step S6), and if a match is not determined again in this collation processing, the Δθ-rotated collation fingerprint image stored in the collation fingerprint rotation image memory 13d is further subjected to a constant angle Δθ rotation processing. Rewritten (steps S7-
S9) The image data is compressed to one-fourth and stored in the verification fingerprint compressed image memory 13e (step S5).

That is, when the matching of the registered fingerprint compressed image and the collated fingerprint compressed image cannot be obtained, the collation fingerprint image is re-collated with the registered fingerprint compressed image while being repeatedly rotated by a certain angle Δθ. This is performed (steps S5 to S9), and the rotation angle memory 13 in the RAM 13 is used.
h, when the collation coincidence between the registered fingerprint compressed image and the collated fingerprint rotated and compressed image is determined before the current rotation angle θ of the collated fingerprint image stored in h becomes the maximum rotation angle θmax. In step 15, the result of the fingerprint matching determination is displayed (step S7 → S10), and the current rotation angle θ of the verification fingerprint image becomes the maximum rotation angle θmax.
If the collation match determination is not made between the registered fingerprint compressed image and the collated fingerprint rotation compressed image, the display unit 15 displays a fingerprint collation mismatch determination result (step S8 →
S10).

When rotating the collation fingerprint image, the original collation fingerprint image, which is an uncompressed image, is always rotated and compressed to one-fourth each time. Rotate by Δθ and collate,
The data generation time of the verification fingerprint image may be reduced.

Here, a specific collation process between the registered fingerprint compressed image and the collated fingerprint rotation compressed image in step S6 of the fingerprint collation process shown in FIG. 5 will be described.

FIG. 6 is a flow chart showing the details of the matching process steps involved in the fingerprint matching process of the fingerprint matching device.

FIG. 7 shows an example of a fingerprint image used for explaining the collation processing of the fingerprint collation apparatus, and FIG.
2 is a compressed and registered fingerprint compressed image memory 1
3B shows a compressed image of the registered fingerprint of the subject, which is stored in 3b. FIG. 4B is read by the image reading device 14, compressed, rotated as necessary, and stored in the verification fingerprint compressed image memory 13e. 4 shows a quarter compressed image of the subject's verification fingerprint. Both fingerprint images in FIG. 7 are multi-tone images.

FIG. 8 is a diagram showing position data of a collation area stored in the registered fingerprint position memory 13f and the collation fingerprint position memory 13g in the RAM 13 with the fingerprint collation processing of the fingerprint collation device.

The registered fingerprint image read from the storage device 12 is compressed to one-fourth and the registered fingerprint compressed image memory 13
b (step S4), and the verification fingerprint image read by the image reading device 14 is rotated and compressed to one-fourth as necessary, and stored in the verification fingerprint compressed image memory 13e (step S4). S5), FIG.
When the processing shifts to the collation processing between the registered fingerprint compressed image and the collated fingerprint rotation compressed image in step S6, first, FIG.
As shown in (A), the rectangular image K
Is defined (step S61). The size of the rectangular area K may be any size as long as required fingerprint matching accuracy is obtained.

Next, as shown in FIG. 7 (B), a rectangular area having the same shape and size as the rectangular area K is set on the compressed image of the collation fingerprint, and the rectangular area is two-dimensionally displayed in pixel units. , And the correlation coefficient between the rectangular area and the rectangular area K is calculated each time the area is moved. As a result, the rectangular area at the position where the correlation coefficient becomes the maximum is defined as the rectangular area KR (step S62). The calculation of the correlation coefficient will be described later.

Here, coordinates serving as references for specifying the positions of the rectangular area K and the rectangular area KR on the image, for example, the coordinates of the vertexes at the upper left corner of both rectangles are obtained, and K (X0, Y
0) and KR (XR0, XR0) are stored in the registered fingerprint position memory 13f and the collated fingerprint position memory 13g in the RAM 13 as shown in FIG. 8 (step S63).

Next, as shown in FIG. 7A, rectangular regions P1 to P4 are defined in the compressed image of the registered fingerprint such that each vertex of the rectangular region K is the center (step S64). The arrangement of the rectangular areas defined here is arbitrary, and the number of the rectangular areas is not limited to four, but may be arbitrary as long as required fingerprint matching accuracy can be obtained. Similarly, the size of the rectangular area defined here may be arbitrary.

Then, similarly to the extraction of the rectangular area KR, as shown in FIG. 7B, each of the rectangular areas P1 to P4 (hereinafter, Pi (i = 1, 2, 3, 4) and Shown)
, A rectangular area having the same size as the rectangular area Pi is set on the compressed image of the verification fingerprint, and the rectangular area is two-dimensionally moved in pixel units, and the phase between the rectangular area and the rectangular area Pi is determined. The number of relations is calculated for each movement. As a result, the rectangular region at the position where the maximum correlation coefficient is obtained is defined as a rectangular region PRi, and coordinates serving as a reference for specifying the positions of the rectangular region Pi and the rectangular region PRi on the image, for example, the upper left corner of both rectangles The coordinates of the vertices of the end are obtained, and Pi (Xi, Y
i) and PRi (XRi, YRi) are stored in the registered fingerprint position memory 13f and the collation fingerprint position memory 13g in the RAM 13 as shown in FIG. 8 (steps S65 to S67).

Then, the relative distance between K and Pi, KR and P
Δi, which is the difference from the relative distance to Ri, is calculated for all i (= 1, 2, 3, 4) based on the following equation (step S68).

[0057]

(Equation 1)

Then, it is determined whether or not all the calculated Δi are within a predetermined value. If all of the calculated Δi are within the predetermined value, it is determined that the registered fingerprint and the collation fingerprint coincide with each other. It is determined that it does not match the collation fingerprint (step S69). The predetermined value used here is, for example, a value that actually calculates Δi from fingerprint image data obtained from a plurality of persons and that obtains a desired collation accuracy based on the distribution of the calculation result.

In the comparison process between the registered fingerprint compressed image and the collated fingerprint rotated compressed image, K, KR, P1
Although the regions P4 and PR1 to PR4 are rectangular regions, the shapes of these regions are not limited to rectangles and may be arbitrary.
It is desirable that the sizes and shapes of K and KR and that of P1 to P4 and PR1 to PR4 are the same, but even if there are some differences, they are acceptable as long as the required fingerprint matching accuracy can be obtained. it can.

In the collation judgment, in addition to the method of evaluating Δi in the above-described processing, for example, a difference in shape or area between a figure formed with Pi as a vertex and a figure formed with PRi as a vertex is used. It is also possible to employ various methods, such as a determination based on.

Next, the calculation of the correlation coefficient used in steps S62 and S66 of the fingerprint collation processing shown in FIG. 6 will be described. Here, a description will be given as calculation of a correlation coefficient between the rectangular area A and the rectangular area B.

First, pixels included in each of the rectangular area A and the rectangular area B are represented by A (i, j) and B (m,
n). However, the total number of pixels included in each of the rectangular area A and the rectangular area B is equal. Also, let _{Xij and Ymn} be signal intensities, which are multi-tone values indicating shading for these pixels.

When these signal intensities are generalized and expressed as Zpq, the following equation is defined.

[0064]

(Equation 2)

In the above equation, N indicates the total number of pixels included in the rectangular area. In the above equation, Σ indicates the sum of all the pixels included in the rectangular area. That is, the above expression indicates the average value of the signal intensity for the pixels included in the rectangular area.

Next, the following expression is further defined.

[0067]

(Equation 3)

The above equation shows the mean square value of the signal intensity for the pixels included in the rectangular area.

Here, the correlation coefficient C _AB between the rectangular area A and the rectangular area B can be calculated by the following equation expressed by using the definition of the above equation.

[0070]

(Equation 4)

A correlation coefficient between regions is calculated using the above equation.

In the calculation of the correlation coefficient by the above equation, instead of using the signal intensities of all the pixels in the rectangular area, for example, the pixels arranged on an arbitrary line in the rectangular area may be used. Calculation using only pixels included in a part of the rectangular region, or calculation using only pixels selected by arbitrarily thinning out the rectangular region. However, there is no problem as long as the required fingerprint matching accuracy can be obtained. Use of such a calculation method is advantageous because the number of pixels for which the correlation coefficient is to be calculated is further reduced and the calculation amount is reduced. Further, another calculation method of the correlation coefficient may be adopted in the fingerprint collation processing described above.

Therefore, according to the fingerprint collation apparatus having the above-described configuration, when the collation fingerprint image is read by the image reading device 14, the registered fingerprint image previously stored and registered in the storage device 12 is read out, and the registered fingerprint image Number of pixels,
After the matching fingerprint image is normalized according to the pixel pitch, both the registered fingerprint image and the matching fingerprint image
One-third compression processing is performed. Then, by performing the collation judgment using the calculation of the correlation coefficient between the compressed registered fingerprint image and the collation fingerprint image, a judgment display of the coincidence / mismatch between the registered fingerprint and the collation fingerprint is made. Even when the correlation coefficient for each pixel is calculated on the collation fingerprint image as a reference, the calculation is performed because the registered fingerprint image and the collation fingerprint image to be subjected to collation determination are compressed to one-fourth. The load associated with the processing is greatly reduced, and the time required for the entire fingerprint collation can be significantly reduced.

Further, according to the fingerprint collating apparatus having the above-mentioned structure,
The quarter compression processing of the registered fingerprint image and the collation fingerprint image is performed by adding a weight of each pixel to a 3 × 3 9-pixel area and compressing it into one pixel in both the vertical and horizontal directions of the target image. Since the processing is performed by repeating the processing every other pixel, the load associated with the compression processing can be minimized without using a complicated image compression technique.

Further, according to the fingerprint collating device having the above configuration,
If the collation match determination between the registered fingerprint compressed image and the collated fingerprint compressed image is not made, the collated fingerprint image which is the original uncompressed image is compressed while rotating by Δθ, and each time the collated fingerprint image is compared with the registered fingerprint compressed image. Since the determination is repeated, the reading angle (finger placement angle) of the fingerprint image of the subject to be registered in the image reading device 14 is
Even when the reading angle of the fingerprint image of the subject to be collated varies, the collation determination of the fingerprint image can be accurately performed.

Further, according to the fingerprint collating device having the above configuration,
Since the collation fingerprint image read by the image reading device 14 is normalized according to the number of pixels and the pixel pitch of the registered fingerprint image as a reference for collation, for example, when the registered fingerprint image is read and registered by the image reading device 14, Even when the resolution of the imaging device differs from that of the imaging device at the time of reading and collating the collation fingerprint image, the fingerprint image collation determination can be accurately performed.

The method described in the above embodiment, that is, each method such as the fingerprint collation processing shown in the flowcharts of FIGS. 5 and 6, is a program that can be executed by a computer as a memory card (ROM card, RAM card). Card, magnetic disk (floppy disk, hard disk, etc.), optical disk (CD-ROM,
DVD and the like, and can be stored in an external storage medium 18a such as a semiconductor memory and distributed. The computer of the fingerprint matching device reads the program stored in the external storage medium 18a into the storage device 12, and controls the operation by the read program, thereby realizing the fingerprint matching function described in the above embodiment. However, similar processing can be performed by the above-described method.

The data of the program for realizing each of the above methods can be transmitted over a network (public line) 20 in the form of a program code. The program data is fetched by the unit 19,
The above-described fingerprint collation function can also be realized.

[0079]

As described above, according to the image collating apparatus according to the first aspect of the present invention, the reference image serving as the reference for collation acquired by the reference image acquiring means and the reference image acquired by the collation image acquiring means. The collation image to be collated with the reference image is both compressed by image compression means, and image collation is performed based on the compressed reference image and the collation image. This reduces the load associated with the matching.

Further, according to the image collating apparatus according to the second aspect of the present invention, since the correlation coefficient of the compressed reference image on the compressed collation image is calculated, the image collation is performed. Even when calculating the correlation coefficient for each pixel on the reference image with respect to the reference image, the load of the calculation process can be reduced by compressing the reference image and the reference image to be subjected to the comparison determination. become.

Thus, according to the present invention, it is possible to reduce the load involved in the collation processing and to significantly reduce the collation time.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a configuration of an electronic circuit of a fingerprint matching device according to an embodiment of the image matching device of the present invention.

FIG. 2 is a diagram showing a data memory secured in a RAM of the fingerprint collation apparatus.

FIG. 3 is a view for explaining quarter compression processing of fingerprint image data by the fingerprint collation apparatus.

FIG. 4 is a diagram showing an example of compressing a fingerprint image in the fingerprint collation device.

FIG. 5 is a flowchart showing a fingerprint matching process by the fingerprint matching device.

FIG. 6 is a flowchart showing details of a matching process step accompanying the fingerprint matching process of the fingerprint matching device.

FIG. 7 is a view showing an example of a fingerprint image used for explaining the collation processing of the fingerprint collation apparatus. FIG. 7A shows a subject read out from a storage device, compressed, and stored in a registered fingerprint compressed image memory. FIG. 4B is a diagram showing a registered fingerprint quarter-compressed image, and FIG. 4B is a diagram showing a subject's collated fingerprint, which is read by an image reader, compressed, rotated as necessary, and stored in the collated fingerprint compressed image memory. The figure which shows one compressed image.

FIG. 8 shows a RAM associated with the fingerprint collation processing of the fingerprint collation apparatus.
FIG. 6 is a diagram showing position data of a collation area stored in a registered fingerprint position memory and a collation fingerprint position memory.

[Explanation of symbols]

 REFERENCE SIGNS LIST 11 control unit (CPU) 12 storage device 13 RAM 13 a read image memory 13 b registered fingerprint compressed image memory 13 c reference fingerprint image memory 13 d reference fingerprint rotation image memory 13 e reference fingerprint compressed image memory 13 f registered fingerprint Position memory 13g Reference fingerprint position memory 13h Rotation angle memory 14 Image reading device 15 Display unit 16 Input unit 17 Bus 18 Storage medium reading unit 18a External storage medium 19 Transmission control unit 20 Network 21 Program server (external computer terminal) 22 ... storage device of external terminal

Claims (7)

[Claims] A reference image acquisition unit for acquiring an image serving as a reference for collation; a collation image acquisition unit for acquiring an image to be collated with the reference image acquired by the reference image acquisition unit; Image compression means for compressing both the reference image acquired by the means and the collation image acquired by the collation image acquisition means; and collating the image based on the reference image and the collation image compressed by the image compression means. Image collating means for performing image collation. 2. The image matching means according to claim 1, wherein the image matching means calculates a correlation coefficient of the compressed reference image on the compressed matching image and matches the images. Item 2. The image matching device according to Item 1. 3. An image rotation unit for rotating the collation image by a predetermined rotation angle, wherein the image collation unit is provided on the compressed reference image on the collation image rotated and compressed by the image rotation unit. Is an image matching unit that calculates a correlation coefficient in and performs image matching.
The image matching device according to claim 1 or 2, wherein: 4. A comparison image normalizing unit that normalizes a comparison image acquired by the comparison image acquisition unit together with a reference image acquired by the reference image acquisition unit, wherein the image compression unit includes: An image compression unit that compresses both the reference image acquired by the reference image acquisition unit and the collation image normalized by the collation image normalization unit.
The image collating apparatus according to claim 1, wherein: 5. The image compression means according to claim 1, wherein said image compression means is an image compression means for adding and compressing both said reference image and said collation image by weighting each pixel value for each predetermined number of pixel regions. The image collating apparatus according to claim 1, wherein the image collating apparatus performs the processing. 6. A reference image acquiring step for acquiring an image serving as a reference for collation, a collation image acquiring step for acquiring an image to be collated with the reference image acquired in the reference image acquiring step, An image compression step of compressing both the reference image acquired in the acquisition step and the collation image acquired in the collation image acquisition step; and a reference image compressed in the image compression step, on the compressed collation image. An image matching step of calculating a correlation coefficient in step (a) to perform image matching. 7. A storage medium storing an image collation processing program for controlling a computer to collate images, the computer comprising: a reference image acquisition unit for acquiring an image serving as a reference for collation; Collation image acquisition means for acquiring an image to be collated with the reference image acquired by the image acquisition means; compressing both the reference image acquired by the reference image acquisition means and the collation image acquired by the collation image acquisition means A computer-readable image processing means that functions as an image compression unit that calculates a correlation coefficient of the reference image compressed by the image compression unit on the compressed verification image and performs image verification. Storage medium that stores various image collation processing programs.