JP4411149B2 - Fingerprint verification device and fingerprint verification method - Google Patents

Description

  The present invention relates to a fingerprint collation device and a fingerprint collation method, and more particularly to a fingerprint collation device and a fingerprint collation method that perform fingerprint collation using a partial fingerprint image.

  2. Description of the Related Art Conventionally, there has been known a fingerprint collation apparatus that performs fingerprint collation using partial fingerprint images (hereinafter referred to as “divided fingerprint images”) continuously collected by a sensor having a detection area smaller than a finger. .

  For example, Japanese Patent Application Laid-Open No. 2003-248820 discloses an image connecting device that reconstructs an entire fingerprint image for fingerprint collation from divided fingerprint images continuously collected by a sensor. The reconstructed whole fingerprint image is compared with a registered fingerprint image registered in advance, and fingerprint matching is performed.

Further, Patent Document 2 (Japanese Patent Laid-Open No. 2003-51013) discloses a fingerprint that performs fingerprint authentication by comparing all of the divided fingerprint images continuously collected by a sensor with the entire registered fingerprint image or a part thereof. An authentication system is described.
JP 2003-248820 A JP 2003-51013 A

  In the technique described in Patent Document 1, it is necessary to perform a complicated process of reconstructing an entire fingerprint image for fingerprint collation from divided fingerprint images continuously collected by a sensor. Further, in the technique described in Patent Document 1, in order to reconstruct the entire fingerprint image, all of the divided fingerprint images continuously collected by the sensor must be temporarily stored in the memory. As a memory for temporarily storing the memory, a memory having a large storage capacity is required. In addition, since a large-capacity memory is required, there is a problem that power for operating the memory is increased.

  In the technique described in Patent Document 2, it is not necessary to perform a complicated process of reconstructing the entire fingerprint image as in the technique described in Patent Document 1, but all of the divided fingerprint images continuously collected by the sensor. Is compared with the registered fingerprint image, the amount of the comparison operation is enormous. For this reason, a long calculation time is required, or a high calculation capability is required for the calculation unit.

  An object of the present invention is to provide a fingerprint collation apparatus and a fingerprint collation method that do not require reconstruction of the entire fingerprint image and can reduce the amount of comparison calculation.

In order to achieve the above object, a fingerprint collation device of the present invention reads a plurality of different portions of one fingerprint and outputs the images as a plurality of partial fingerprint images, and a plurality of portions output by the sensor. An extractor that extracts a plurality of specific partial fingerprint images that are separated from each other by a predetermined value from the fingerprint image, a specific partial fingerprint image memory that stores a plurality of specific partial fingerprint images extracted by the extractor, and registered in advance A registered fingerprint image memory for storing the registered fingerprint image, a comparator for comparing each of the plurality of specific partial fingerprint images stored in the specific partial fingerprint image memory with a registered fingerprint image stored in the registered fingerprint image memory; based on the comparison result of the comparator, see containing and a determination unit whether fingerprint the sensor read is identical to the registered fingerprint image, the comparison Compares a predetermined one of the plurality of specific partial fingerprint images with the registered fingerprint image, extracts a partial registered fingerprint image corresponding to the predetermined one from the registered fingerprint image, and extracts the partial registered fingerprint image Based on the image and the predetermined value, a partial registered fingerprint image corresponding to each of the remaining parts obtained by removing the predetermined one from the plurality of specific partial fingerprint images is extracted from the registered fingerprint image, and the plurality of specific parts Each fingerprint image is collated with a corresponding partially registered fingerprint image .

The fingerprint collation method of the present invention is a fingerprint collation method performed by a fingerprint collation apparatus, and includes a fingerprint reading step for reading a plurality of different portions of one fingerprint, and images of the plurality of portions read in the fingerprint reading step. Extraction of extracting a plurality of specific partial fingerprint images that are separated from each other by a predetermined value from the partial fingerprint image output step that is output as a plurality of partial fingerprint images and the plurality of partial fingerprint images output in the partial fingerprint image output step A step of storing a plurality of specific partial fingerprint images extracted in the extraction step in a specific partial fingerprint image memory; a registered fingerprint image memory storing a pre-registered registered fingerprint image; and storing in the storage step A collation step for collating each of the plurality of specific partial fingerprint images with the registered fingerprint image stored in the registered fingerprint image memory. If, based on the collation result of the collation step, seen including and a determination step of determining whether readings taken fingerprint is identical to the registered fingerprint image in the fingerprint reading step, said checking step, A first extraction step of comparing a predetermined one of the plurality of specific partial fingerprint images with the registered fingerprint image to extract a partial registered fingerprint image corresponding to the predetermined one from the registered fingerprint image; Based on the partial registered fingerprint image extracted in the first extraction step and the predetermined value, the partial registered fingerprint corresponding to each of the remaining ones obtained by removing the predetermined one from the plurality of specific partial fingerprint images from the registered fingerprint image A second extraction step for extracting an image; a comparison step for comparing each of the plurality of specific partial fingerprint images with a corresponding partial registration fingerprint image extracted in the first and second extraction steps; Including.

  According to the above invention, the specific partial fingerprint image is extracted from the partial fingerprint image output from the sensor, and the specific partial fingerprint image is collated with the registered fingerprint image. For this reason, it is not necessary to perform a complicated process of reconstructing the entire fingerprint image from the partial fingerprint image, and the memory necessary for reconstructing the entire fingerprint image can be eliminated. In addition, since the specific partial fingerprint image extracted from the partial fingerprint image is collated with the registered fingerprint image, compared with the case where all the partial fingerprint images output by the sensor are collated with the registered fingerprint image It is possible to reduce the amount of calculation.

  Further, if the predetermined value is increased, the amount of comparison operation associated with the collation can be reduced, and if the predetermined value is shortened, the accuracy of the collation can be increased. Therefore, if the predetermined value is set to a desired value, it is possible to perform fingerprint collation according to the required collation accuracy or the required amount of comparison operation.

  In the fingerprint collation apparatus, the sensor outputs the partial fingerprint images one by one, and the extractor outputs the partial fingerprint image output by the sensor each time the sensor outputs one partial fingerprint image. It is desirable to determine whether or not is the specific partial fingerprint image and extract only the partial fingerprint image determined to be the specific partial fingerprint image.

  In the fingerprint collation method, the partial fingerprint image output step outputs the partial fingerprint images one by one, and the extraction step outputs the partial fingerprint image one by one in the partial fingerprint image output step. It is desirable to determine whether or not the partial fingerprint image is the specific partial fingerprint image and extract only the partial fingerprint image determined to be the specific partial fingerprint image.

  According to the above invention, each time the sensor outputs one partial fingerprint image, it is determined whether or not the output partial fingerprint image is a specific partial fingerprint image. Therefore, it is possible to extract a specific partial fingerprint image necessary for fingerprint collation as needed.

  According to the above invention, when a predetermined one of a plurality of specific partial fingerprint images is compared with a registered fingerprint image and a partial registered fingerprint image corresponding to the predetermined one is extracted, another specific partial fingerprint is obtained. The partially registered fingerprint image corresponding to the image is extracted based on the extracted partially registered fingerprint image and a predetermined distance. For this reason, it is possible to easily extract the partially registered fingerprint image.

  According to the present invention, the specific partial fingerprint image is extracted from the partial fingerprint image output from the sensor, and the specific partial fingerprint image is collated with the registered fingerprint image. For this reason, it is not necessary to perform a complicated process of reconstructing the entire fingerprint image from the partial fingerprint image, and the memory necessary for reconstructing the entire fingerprint image can be eliminated. In addition, since the specific partial fingerprint image extracted from the partial fingerprint image is collated with the registered fingerprint image, compared with the case where all the partial fingerprint images output by the sensor are collated with the registered fingerprint image It is possible to reduce the amount of calculation.

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

  FIG. 1 is a block diagram showing a fingerprint collation apparatus according to an embodiment of the present invention. In FIG. 1, the fingerprint collation device 1 includes a sensor 2, an extractor 3, a memory 4, a comparator 5, and a determiner 6.

  The sensor 2 is an image reading unit that reads a plurality of different portions of one fingerprint and outputs images of the plurality of read portions as a plurality of partial fingerprint images (hereinafter referred to as “divided fingerprint images”).

  For example, when the user slides his / her finger on the reading area (not shown) of the sensor 2 while the user slides the finger on the reading area of the sensor 2, in other words, the user's finger moves against the sensor 2. When relatively moved, a plurality of different portions are sequentially read from the fingerprint of the finger, and the images of the read portions are output one by one in the order of reading as a plurality of divided fingerprint images.

  The sensor 2 outputs the divided fingerprint image to the extractor 3. The area of the reading region of the sensor 2 is preferably smaller than the area of the fingerprint of one human finger.

  FIG. 2 is an explanatory diagram showing an example of a plurality of divided fingerprint images output from the sensor 2. In FIG. 2, a divided fingerprint image 7a, a divided fingerprint image 7b, and a divided fingerprint image 7c are divided fingerprint images output from the sensor 2. For example, the sensor 2 includes the divided fingerprint image 7a, the divided fingerprint image 7b, and the divided fingerprint image. The divided fingerprint images are output successively one by one in the order of the image 7c. The number of divided fingerprint images 7 output from the sensor 2 is not limited to three and can be changed as appropriate.

  Returning to FIG. 1, the extractor 3 is a distance calculator, for example. The extractor 3 extracts a plurality of specific partial fingerprint images that are separated from each other by a predetermined value from the plurality of divided fingerprint images output from the sensor 2. For example, each time a divided fingerprint image is input from the sensor 2, the extractor 3 determines whether or not the input divided fingerprint image is a specific partial fingerprint image, and determines that the input is a specific partial fingerprint image Extract only the partial fingerprint image.

  The predetermined value is desirably changed by the user. For example, it is desirable that an input unit for receiving user input is provided in the extractor, and the user operates the input unit to set a predetermined distance in the extractor 3. If the predetermined value is increased, the amount of comparison operation associated with the collation can be reduced, and if the predetermined value is shortened, the accuracy of the collation can be increased. Therefore, if the user arbitrarily changes the predetermined value, it is possible to perform fingerprint collation according to the required collation accuracy or the required amount of comparison operation.

  FIG. 3 is an explanatory diagram showing an example of the specific partial fingerprint image 8 extracted by the extractor 3. In FIG. 3, the specific partial fingerprint image 8 includes a specific partial fingerprint image 8a, a specific partial fingerprint image 8b, a specific partial fingerprint image 8c, a specific partial fingerprint image 8d, and a specific partial fingerprint image 8e. The specific partial fingerprint image 8a, the specific partial fingerprint image 8b, the specific partial fingerprint image 8c, the specific partial fingerprint image 8d, and the specific partial fingerprint image 8e have a predetermined (fixed) distance from the adjacent specific partial fingerprint image. Yes.

  Returning to FIG. 1, the memory 4 includes a specific partial fingerprint image memory 41 and a registered fingerprint image memory 42. The specific partial fingerprint image memory 41 stores the specific partial fingerprint image (see FIG. 3) extracted by the extractor 3. The registered fingerprint image memory 42 stores registered fingerprint images registered in advance.

  FIG. 4 is an explanatory diagram showing an example of the registered fingerprint image 9 stored in the registered fingerprint image memory 42.

  Returning to FIG. 1, the comparator 5 converts the specific fingerprint image 8 (see FIG. 3) stored in the specific fingerprint image memory 41 into the registered fingerprint image 9 (see FIG. 4) stored in the registered fingerprint image memory 42. ).

  The determination unit 6 determines whether the fingerprint read by the sensor 2 (specifically, the specific partial fingerprint image 8) matches the registered fingerprint image 9 based on the comparison result of the comparator 5.

  Next, an outline of the operation will be described.

  When the user's finger moves relative to the sensor 2, the sensor 2 sequentially reads a plurality of different portions from the fingerprint of the finger, and reads a plurality of different divided fingerprint images according to the read portion. Output one by one in the order specified. The extractor 3 extracts a plurality of specific partial fingerprint images 8 that are separated from each other by a predetermined value from the plurality of divided fingerprint images output from the sensor 2, and extracts the plurality of specific partial fingerprint images 8 that have been extracted Stored in the fingerprint image memory 41.

  The comparator 5 collates the specific fingerprint image 8 stored in the specific fingerprint image memory 41 with the registered fingerprint image 9 stored in the registered fingerprint image memory 42. The determiner 6 determines whether or not the specific partial fingerprint image 8 matches the registered fingerprint image 9 based on the comparison result of the comparator 5.

  According to the present embodiment, it is not necessary to reconstruct the entire fingerprint image from the divided fingerprint images continuously collected by the sensor 2, and the divided fingerprint images stored in the specific partial fingerprint image memory 41 are continuously collected. Since a plurality of specific partial fingerprint images extracted from all the divided fingerprint images based on a predetermined value are obtained, the storage capacity for temporarily storing the divided fingerprint images can be reduced.

  Moreover, since the divided fingerprint images used for collation are divided fingerprint images extracted based on a predetermined value from all the divided fingerprint images collected continuously, all the divided fingerprint images collected continuously are Compared with the case of collating with a registered fingerprint image, the amount of comparison calculation accompanying collation can be reduced, and the time required for collation can be shortened.

  Next, the operation will be specifically described.

  FIG. 5 is a flowchart for explaining the operation of the fingerprint collation device 1. Hereinafter, the operation of the fingerprint collation apparatus 1 will be described with reference to FIG.

  In step 51, for example, when a fingerprint input operation is started by operating a mode switch (not shown), the sensor 2 executes step 52.

  In step 52, the sensor 2 reads an object existing on the reading area of the sensor 2, and acquires an image of the read object. When the sensor 2 acquires the video, the sensor 2 executes Step 53.

  In step 53, the sensor 2 determines whether or not a finger is placed on the reading area of the sensor 2, in other words, whether or not a fingerprint is being input.

  Specifically, the sensor 2 compares the image acquired in step 52 with the image acquired immediately before the image acquired in step 52, and if the two images are different, the finger is read by the sensor 2. It is determined that the image is placed on the area, and the image acquired in step 52 is output to the extractor 3 as a divided fingerprint image 7 (see FIG. 2). When the extractor 3 receives the divided fingerprint image 7 from the sensor 2, the extractor 3 executes Step 54. Note that the sensor 2 executes Step 52 when the two compared images are the same.

  In step 54, the extractor 3 stores the divided fingerprint image 7 input from the sensor 2 in a memory (not shown) in the extractor 3, and determines whether the stored divided fingerprint image 7 is a continuous divided fingerprint image. Judging. Note that the memory in the extraction unit 3 can store two divided fingerprint images 7. In step 54, the extractor 3 extracts the divided fingerprint image 7 input from the sensor 2. Store in the memory area.

The step 3 will be described in detail. The extractor 3 uses the divided fingerprint image input from the sensor 2 stored in the memory in the extractor 3 and the divided fingerprint image stored in the memory in the extractor 3 as well. An overlapping region is searched while relatively moving the divided fingerprint image input from the sensor 2 immediately before. For example, when the divided fingerprint image 7b in FIG. 2 is input from the sensor 2, the extractor 3 relatively moves the divided fingerprint image 7a in FIG. 2 input from the sensor 2 and the divided fingerprint image 7b immediately before that. A region where the divided fingerprint image 7a and the divided fingerprint image 7b overlap is searched.

  When the extractor 3 searches for the overlapping area, it determines that the two divided fingerprint images are continuous divided fingerprint images and executes step 55. If the overlapping area cannot be searched, the fingerprint input is completed. And step 60 is executed.

  In step 54, when the extractor 3 inputs a divided fingerprint image for the first time after determining that the sensor 2 is inputting a fingerprint, for example, the extractor 3 determines that the sensor 2 is inputting a fingerprint. When 7a is input, the divided fingerprint image 7a is stored in a memory (not shown) in the extractor 3, and step 55 is executed.

  In step 55, the extractor 3 calculates the distance that the divided fingerprint image input from the sensor 2 has moved from the position of the divided fingerprint image input from the sensor 2 immediately before the divided fingerprint image.

  Specifically, the extractor 3 moves the common region by performing a correlation calculation of feature shapes in overlapping regions while relatively moving two temporally divided fingerprint images stored in the memory of the extractor 3. To calculate the relative movement amount between the images. When the extractor 3 calculates the movement amount, the extractor 3 executes Step 56.

  In step 55, the extractor 3 receives the divided fingerprint image for the first time after determining that the sensor 2 is inputting a fingerprint. For example, the extractor 3 determines that the sensor 2 is inputting a fingerprint after the determination. When 7a is input, step 56 is executed.

  In step 56, the extractor 3 determines that the divided fingerprint image input from the sensor 2 is the first divided fingerprint image input after the sensor 2 determines that the fingerprint is being input (for example, the divided fingerprint image 7a in FIG. 2). It is judged whether it is.

  The extractor 3 executes step 57 if the divided fingerprint image input from the sensor 2 is the first divided fingerprint image input after determining that the sensor 2 is inputting the fingerprint, and inputs the input from the sensor 2 If the divided fingerprint image is not the first divided fingerprint image input after the sensor 2 determines that the fingerprint is being input, step 58 is executed.

  In step 57, the extractor 3 stores the divided fingerprint image input from the sensor 2 in the specific partial fingerprint image memory 41 as the specific partial fingerprint image 8. The divided fingerprint image (for example, the divided fingerprint image 7a shown in FIG. 2) first input after the sensor 2 determines that the fingerprint is being input by the operation of the extractor 3 in step 57 is a specific partial fingerprint image ( For example, it is stored in the specific partial fingerprint image memory 41 as the specific partial fingerprint image 8a) shown in FIG. The extractor 3 executes Step 59 when Step 57 ends.

  In step 58, the extractor 3 determines whether the divided fingerprint image input from the sensor 2 has moved by a certain distance (predetermined value) with respect to the specific partial fingerprint image stored most recently in the specific partial fingerprint image memory 41. .

  The extractor 3 executes step 57 when the divided fingerprint image input from the sensor 2 has moved a certain distance (predetermined value) with respect to the specific partial fingerprint image stored in the specific partial fingerprint image memory 41 most recently. If the divided fingerprint image input from the sensor 2 has not moved a certain distance (predetermined value) with respect to the specific partial fingerprint image stored most recently in the specific partial fingerprint image memory 41, the divided fingerprint image input from the sensor 2 is Step 59 is executed without storing in the specific partial fingerprint image memory 41.

  In step 59, the extractor 3 deletes the non-latest divided fingerprint image stored in the memory of the extractor 3 from the memory of the extractor 3, and subsequently receives the next input from the sensor 2. Are received, and step 54 is executed.

  When the extractor 3 repeatedly executes Step 54 to Step 59, the specific partial fingerprint image memory 41 stores a plurality of specific partial fingerprint images 8 that are separated from each other by a predetermined value (a constant distance).

  In step 60, the extractor 3 outputs a fingerprint input end signal to the comparator 5. When the comparator 5 receives the fingerprint input end signal, it executes step 61.

  In step 61, the comparator 5 specifies a partial registered fingerprint image on the registered fingerprint image corresponding to a predetermined one of the plurality of specific partial fingerprint images stored in the specific partial fingerprint image memory 41. Specifically, the comparator 5 specifies, from the registered fingerprint image, an area having a fingerprint closest to a predetermined one as a partially registered fingerprint image corresponding to the predetermined one.

  For example, the comparator 5 detects from the registered fingerprint image 9 a partial registered fingerprint image corresponding to the specific partial fingerprint image 8a shown in FIG. In this case, the comparator 5 compares the specific partial fingerprint image 8a with the registered fingerprint image 9, and detects the partial registered fingerprint image corresponding to the specific partial fingerprint image 8a shown in FIG. When the comparator 5 specifies the partially registered fingerprint image, the comparator 5 executes Step 62.

  In step 62, the comparator 5 determines whether or not there is a specific partial fingerprint image that does not specify the corresponding partial registered fingerprint image among the plurality of specific partial fingerprint images stored in the specific partial fingerprint image memory 41. To do. If there is a specific partial fingerprint image that does not specify the corresponding partial registered fingerprint image, the comparator 5 executes step 61 and there is a specific partial fingerprint image that does not specify the corresponding partial registered fingerprint image. If so, step 63 is executed.

  When executing step 61 via step 62, the comparator 5 has already determined the position of the partial registered fingerprint image 9a corresponding to the specific partial fingerprint image 8a in step 61 before executing step 62. Therefore, when specifying the partial registered fingerprint images corresponding to the remaining specific fingerprint images 8b, 8c, 8d, and 8e, the remaining specific portions are determined based on the partial registered fingerprint image 9a and a certain distance (predetermined value). Partially registered fingerprint images corresponding to the fingerprint images 8b, 8c, 8d, and 8e are automatically determined, respectively.

  Specifically, the comparator 5 specifies a partial registration fingerprint image 9b that is separated from the partial registration fingerprint image 9a by a predetermined value, and then specifies a partial registration fingerprint image 9c that is separated from the partial registration fingerprint image 9b by a predetermined value. Then, the partial registration fingerprint image 9d separated from the partial registration fingerprint image 9c by a predetermined value is specified, and then the partial registration fingerprint image 9e separated from the partial registration fingerprint image 9d by a predetermined value is specified.

  In step 63, the comparator 5 collates each of the plurality of specific partial fingerprint images stored in the specific partial fingerprint image memory 41 with the corresponding partial registered fingerprint image.

  The collation method performed by the comparator 5 in step 63 may be a method of comparing the feature amount of the specific partial fingerprint image with the feature amount of the partial registered fingerprint image corresponding to the specific partial fingerprint image. A method of comparing with a partially registered fingerprint image corresponding to the specific partial fingerprint image may be used.

  In the case of the method of extracting and comparing feature amounts, there is a possibility that the feature amount is not included in both the specific partial fingerprint image and the partial registered fingerprint image. In that case, the comparator 5 compares the fingerprint image of the portion and performs collation, and sends the collation result to the determination unit 6.

  The determiner 6 obtains an average value of the matching results of the individual specific partial fingerprint images and the corresponding partial registered fingerprint images sent from the comparator 5, and compares the average value with a predetermined threshold value. The determiner 6 determines whether or not the specific partial fingerprint image 8 (specifically, the fingerprint read by the sensor 2) is the same as the registered fingerprint image 9 depending on whether or not the average value is larger than a predetermined threshold value. Judgment is made.

  According to the present embodiment, the following effects can be obtained.

  The first effect is that the entire apparatus can be downsized. The reason is that a sensor that reads a plurality of different portions from one fingerprint is used, so that a sensor that reads the entire fingerprint at once can be eliminated.

  The second effect is that the entire apparatus can be saved in power. The reason is that the divided fingerprint image temporarily stored for fingerprint collation becomes a specific partial fingerprint image extracted from the divided fingerprint image output from the sensor 2 based on a predetermined value, so that all the divided fingerprint images are stored. This is because the storage capacity of the memory required for storing the divided fingerprint image is smaller than in the case where the divided fingerprint image is stored. For this reason, it becomes possible to achieve power saving of the memory.

  The third effect is that the verification time can be shortened. The reason is that the divided fingerprint image to be compared with the registered fingerprint image is a specific partial fingerprint image extracted from the divided fingerprint image output from the sensor 2 based on a predetermined value. This is because the amount of comparison calculation is smaller than in the case of comparison.

  In the embodiment described above, the illustrated configuration is merely an example, and the present invention is not limited to the configuration.

It is the block diagram which showed the fingerprint collation apparatus of one Example of this invention. It is explanatory drawing which showed an example of the division | segmentation fingerprint image. It is explanatory drawing which showed an example of the specific partial fingerprint image. It is explanatory drawing which showed an example of the registered fingerprint image. It is a flowchart for demonstrating operation | movement of a fingerprint collation apparatus.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Fingerprint collation apparatus 2 Sensor 3 Extractor 4 Memory 41 Specific partial fingerprint image memory 42 Registered fingerprint image memory 5 Comparator 6 Judgment device

Claims (6)

A sensor that reads a plurality of different portions of one fingerprint and outputs the images as a plurality of partial fingerprint images;
An extractor for extracting a plurality of specific partial fingerprint images that are separated from each other by a predetermined value from a plurality of partial fingerprint images output by the sensor;
A specific fingerprint image memory for storing a plurality of specific fingerprint images extracted by the extractor;
A registered fingerprint image memory for storing a registered fingerprint image registered in advance;
A comparator for comparing each of a plurality of specific partial fingerprint images stored in the specific partial fingerprint image memory with a registered fingerprint image stored in the registered fingerprint image memory;
Based on the comparison result of the comparator, see containing and a determination unit whether fingerprint the sensor read is identical to the registered fingerprint image,
The comparator compares a predetermined one of the plurality of specific partial fingerprint images with the registered fingerprint image to extract a partial registered fingerprint image corresponding to the predetermined one from the registered fingerprint image; Based on the partial registered fingerprint image and the predetermined value, a partial registered fingerprint image corresponding to each of the remaining portions obtained by removing the predetermined one from the plurality of specific partial fingerprint images is extracted from the registered fingerprint image; A fingerprint collation device that collates each of the specific partial fingerprint images with a corresponding partial registered fingerprint image . The fingerprint verification apparatus according to claim 1,
The sensor outputs the partial fingerprint images one by one,
Each time the sensor outputs one partial fingerprint image, the extractor determines whether or not the partial fingerprint image output by the sensor is the specific partial fingerprint image. A fingerprint collation device that extracts only the determined partial fingerprint image. The fingerprint collation device according to claim 1 or 2,
The fingerprint collating apparatus , wherein the predetermined value can be changed . A fingerprint verification method performed by a fingerprint verification device,
A fingerprint reading step of reading a plurality of different portions of one fingerprint;
A partial fingerprint image output step for outputting a plurality of partial images read in the fingerprint reading step as a plurality of partial fingerprint images;
An extraction step of extracting a plurality of specific partial fingerprint images that are separated from each other by a predetermined value from the plurality of partial fingerprint images output in the partial fingerprint image output step;
A storing step of storing a plurality of specific partial fingerprint images extracted in the extraction step in a specific partial fingerprint image memory;
A registered fingerprint image memory for storing a registered fingerprint image registered in advance;
A collation step of collating each of the plurality of specific partial fingerprint images stored in the storing step with a registered fingerprint image stored in a registered fingerprint image memory;
On the basis of the matching step of matching results, see containing and a determination step of determining whether readings taken fingerprint is identical to the registered fingerprint image in the fingerprint reading step,
The matching step includes
A first extraction step of comparing a predetermined one of the plurality of specific partial fingerprint images with the registered fingerprint image and extracting a partial registered fingerprint image corresponding to the predetermined one from the registered fingerprint image;
Partial registration corresponding to each of the remaining portions obtained by removing the predetermined one from the plurality of specific partial fingerprint images from the registered fingerprint image based on the partial registered fingerprint image extracted in the first extraction step and the predetermined value A second extraction step for extracting a fingerprint image;
A comparison step of comparing each of the plurality of specific partial fingerprint images with a corresponding partial registration fingerprint image extracted in the first and second extraction steps . The fingerprint collation method according to claim 4, wherein
The partial fingerprint image output step outputs the partial fingerprint images one by one,
The extraction step determines whether or not the partial fingerprint image is the specific partial fingerprint image every time one partial fingerprint image is output in the partial fingerprint image output step. A fingerprint verification method that extracts only the determined partial fingerprint image. In the fingerprint collation method according to claim 4 or 5,
The fingerprint collating method , wherein the predetermined value can be changed .

Images