JP2001273497A - Personal identification device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a personal identification device capable of performing accurate identification by one time of identifying operation. SOLUTION: In this personal identification device, plural CCD cameras 12 and 14 for photographing the blood vessel image of a finger are arranged in the device body having finger inserting hole and these CCD cameras 12 and 14 are connected to an information processor 6. The information processor 6 is provided with an image processing circuit 2 for recognizing a three- dimensional blood vessel pattern on the basis of the blood vessel image provided from the CCD cameras 12 and 14 and an arithmetic processing circuit 3 for extracting plural feature amounts from the blood vessel pattern and collating these extracted data with registered data in a memory 5 and in the case of collating both the data, this arithmetic processing circuit 3 applies processing of coordinate system rotation around the central axis of the finger inserting hole and decides similarity between correction data provided thereby and the other data.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an apparatus for performing personal identification based on a blood vessel pattern of a fingertip.

[0002]

2. Description of the Related Art Conventionally, a light source (16) and a CCD camera (17) are provided on both sides of a finger F as shown in FIG.
An apparatus has been proposed in which a finger is perspectively photographed by 7), a geometric feature of a blood vessel pattern is extracted from a blood vessel image obtained thereby, and personal identification is performed based on data representing the extracted feature. (E.g., JP-A-7-21373)
issue). FIG. 8A shows a two-dimensional blood vessel image obtained from a CCD camera, and FIG. 8B shows a geometric feature of a blood vessel pattern obtained from the image (for example, a branch point of a blood vessel). A plurality of data E representing coordinate values, branching directions, etc.)
(x, y), B (x, y), and C (x, y).

[0003] In an apparatus for performing personal identification based on the characteristic amount of a blood vessel pattern as described above, data representing the characteristic amount of a specific person is registered in advance, and the data between the data to be collated and the registered data is registered. The degree of similarity is determined.
When the similarity between the two data exceeds a predetermined threshold,
You are identified as yourself. In the determination of the similarity, as shown in FIG. 7, first, the similarity is initialized in step S31, and then in step S32, the similarity between the collation target data and the registered data is calculated. Here, a correlation coefficient or the like can be adopted as the similarity. Next, in step S33, it is determined whether or not the persons match or not based on whether or not the similarity between the two data exceeds a predetermined threshold. Then, in the case of a match, in step S34, permission information to the effect that the user is issued is issued.
At 35, non-permission information indicating that the user is not the person is issued, and the procedure is terminated.

According to the personal identification device, personal identification can be performed without imposing a burden on the user.

[0005]

However, in the conventional personal identification device utilizing the characteristic amount of the blood vessel pattern, the posture of the finger when the finger is inserted between the light source (16) and the CCD camera (17), In particular, the change in the posture in the rotation direction about the longitudinal axis of the finger also changes the blood vessel pattern image obtained from the CCD camera (17), and accordingly, for example, the coordinate values and branch directions of the branch points of the blood vessels also change. Will shift. Therefore, if the posture of the finger when the user previously registers the feature amount of his or her own blood vessel pattern is different from the posture of the finger at the time of subsequent verification, the user agrees with the feature amount and the permission information is not obtained. Until it is issued, the identification operation must be performed a plurality of times while gradually changing the posture of the finger. Further, in the conventional personal identification device, the fingertip is two-dimensionally photographed by one CCD camera (17), and the feature amount is extracted from the two-dimensional blood vessel pattern. However, there was a problem that the accuracy was low and the identification accuracy was low.

Accordingly, an object of the present invention is to provide a personal identification device capable of performing highly accurate identification by one identification operation.

[0007]

In the personal identification device according to the present invention, a blood vessel image of the finger in the finger insertion hole (15) is three-dimensionally provided on the device body provided with the finger insertion hole (15). An imaging device for photographing is provided, and the imaging device is connected to an information processing circuit. The information processing circuit recognizes a three-dimensional blood vessel pattern based on a blood vessel image obtained from an imaging device, and extracts a plurality of geometric feature amounts from the blood vessel pattern; A feature amount registering unit in which data representing feature amounts of a plurality of specific persons is registered, and a feature amount extracting instruction is given to the feature amount extracting unit in response to a collation instruction, and the data representing the feature amount extracted by this is given. And a feature amount matching unit that compares data registered in the feature amount registration unit to determine a similarity between the two data, and permission information only when the similarity between the two data is determined to be high. Permission information issuing means for issuing, the feature amount collating means, when collating both data, performs a coordinate rotation process around the central axis of the finger insertion hole (15) to one of the data, This gives The similarity between the corrected data and the other data is determined.

[0008] More specifically, the feature amount collating means repeats the coordinate rotation process and the similarity determination for each predetermined rotation angle, and when the similarity of both data exceeds a predetermined threshold, the similarity is high. Is determined.

In the personal identification apparatus of the present invention, since the feature is extracted from the three-dimensional blood vessel pattern obtained by the fluoroscopic imaging of the finger, the feature is extracted from the conventional two-dimensional blood vessel pattern. Compared with this, more accurate identification information can be obtained, which enables highly accurate identification. Further, even if the posture of the finger when the finger is inserted into the finger insertion hole (15) of the device main body changes in the rotation direction from the posture at the time of registration, the data representing the characteristic amount extracted by the characteristic amount extracting means is not changed. After receiving the coordinate rotation process for each fixed rotation angle, the data is corrected to the data at each rotation angle, and the corrected data and the registered data are sequentially collated. If they belong to the same person, the two data always match.

In a specific configuration, the imaging device includes at least two CCD cameras arranged in a direction perpendicular to the longitudinal direction of the finger insertion hole (15) of the device main body, and the feature amount extracting means includes: Image processing for synthesizing two two-dimensional images obtained from the two CCD cameras and generating a three-dimensional image of the blood vessel with the origin located at or near the back end position of the finger insertion hole (15) of the apparatus main body. With means.

In the above-described specific configuration, the finger is positioned in the longitudinal direction by inserting the finger into the finger insertion hole (15) of the apparatus main body. Therefore, if the finger is photographed from different directions by two CCD cameras, a three-dimensional image of the finger can be obtained by synthesizing two two-dimensional images obtained from these CCD cameras.

[0012]

According to the personal identification device of the present invention, if a finger is inserted into a finger insertion hole of the device body in an arbitrary direction, a single identification operation can be performed without changing the direction of the finger. Can also perform highly accurate identification.

[0013]

DESCRIPTION OF THE PREFERRED EMBODIMENTS Embodiments of the present invention will be specifically described below with reference to the drawings. As shown in FIG. 1, the personal identification device according to the present invention is configured by connecting an information processing device (6) to a detector (1). The detector (1) is a finger F
A first light source (11) and a first CCD camera (12) disposed with a finger F therebetween on a first crossing line orthogonal to the longitudinal direction of
It comprises a second light source (13) and a second CCD camera (14) arranged with a finger F on a second crossing line orthogonal to the crossing line.

The detector (1) has a rectangular parallelepiped main body (10) as shown in FIG. 2, and a finger insertion hole (15) for inserting a finger F is opened in the detector main body (10). The finger insertion hole (1
By inserting the finger F to the depth of 5), the finger F is positioned in the longitudinal direction of the finger insertion hole (15). In this embodiment, the back end of the finger insertion hole (15) is the origin O, and the finger insertion hole (1
5) Three-dimensional coordinates are set, where the longitudinal axis is the Z axis, and two axes orthogonal to the Z axis are the X axis and the Y axis.

As shown in FIG. 1, the information processing device (6) comprises a light source control circuit (4) for controlling lighting of a first light source (11) and a second light source (13), a first CCD camera (12) and An image processing circuit (2) for generating a three-dimensional image by synthesizing a two-dimensional image obtained from the second CCD camera (14), and performing predetermined arithmetic processing on the three-dimensional image obtained from the image processing circuit (2). And a memory (5) for registering identification information (feature amount data) of a specific person.

FIGS. 5 (a-1) and 5 (a-2) show perspective images of a finger obtained from the first and second CCD cameras (12) and (14), respectively. A dimensional blood vessel pattern appears. The broken lines in FIGS. 5 (b-1) and 5 (b-2) respectively represent a two-dimensional blood vessel pattern as a series of coordinate data strings. The image processing circuit (2) synthesizes the coordinate data strings of these two two-dimensional images,
A coordinate data sequence of a three-dimensional image is generated as shown by a broken line in FIG.

Further, the arithmetic processing circuit (3) uses the coordinate data sequence of the three-dimensional image to extract a plurality of feature values (for example, coordinate values of branch points of blood vessels, branch directions, ), And as shown in FIG. 5C, a plurality of data E (x, y, z) and B (x,
y, z) and C (x, y, z). Then, the arithmetic processing circuit (3) executes the registration and collation processing based on the feature amount data.

FIG. 3 shows a series of procedures executed by the information processing device 6 for personal identification. First, in step S1, the ID number input by the user is captured, and in step S2, an image of the finger is captured from the detector (1). Then, after performing image processing for generating a three-dimensional image in step S3, in step S4, a feature amount is extracted from the blood vessel pattern. Then, in step S5,
It is determined whether the user has performed a registration operation or a collation operation. If the registration operation has been performed, the process proceeds to step S6, where the memory (5)
Is registered with the ID number and the feature amount data.

On the other hand, when the collation operation is being performed, the process proceeds to step S7, where the feature amount data to be collated (hereinafter referred to as collation data) is collated with the registration data of the corresponding ID number, and the person is identified. , A mismatch is determined. Here, when the match is determined, the process proceeds to step S8 to issue permission information indicating that the user is the person. When the mismatch is determined, the process proceeds to step S9, and the permission information indicating that the user is not the person is transmitted. Issue.

FIG. 4 shows a specific procedure of the collation procedure in step S7. First, step S1
In step 1, the similarity as a criterion for determining whether a person matches or not matches is initialized. Next, the collation data is set to the data of the initial angle. Then, in step S13, the similarity between the collation data and the registered data is calculated. Note that a correlation coefficient is used as the similarity. Then, step S1
At 4, it is determined whether or not the calculated similarity has reached the maximum value exceeding the previous value. If the determination is yes, the similarity is stored at step S15. When it is determined NO in step S14, the process goes to step S16, bypassing step S15.

In step S16, it is determined whether or not the coordinate rotation processing of the collation data has been completed by the specified rotation angle. If the determination is no, the process proceeds to step S17, where the Z axis is added to the collation data. After performing a coordinate rotation process at a fixed angle centered on, the process returns to step S13. Then, when it is determined YES in step S16, the process proceeds to step S18. In this way, the maximum similarity among a plurality of similarities obtained in a series of coordinate rotation processing is stored.

In step S18, the stored similarity is compared with a predetermined threshold, and when the similarity exceeds the threshold, it is determined that there is a match, and when the similarity does not exceed the threshold, there is no match. judge. Then, when it is determined that they match, permission information is issued (step S8), and when it is determined that they do not match, non-permission information is issued (step S9).

In the personal identification device, two CCs
The three-dimensional blood vessel pattern is recognized based on the finger images using the D cameras (12) and (14), and the feature amount is extracted from the blood vessel pattern. Therefore, the feature amount is extracted from the two-dimensional blood vessel pattern as in the related art. More accurate identification information can be obtained as compared with the case where the identification is performed, thereby enabling highly accurate identification. Even if a finger is inserted in an arbitrary direction into the finger insertion hole (15) of the detector body (10) of the detector (1), the collation processing involving the coordinate rotation processing around the Z axis is performed. Action is 1
It is only necessary to repeat the identification operation while changing the direction of the finger as in the related art.

The configuration of each part of the present invention is not limited to the above embodiment, and various modifications can be made within the technical scope described in the claims. For example, in the above embodiment, two C
Although the three-dimensional blood vessel pattern is recognized using the CD cameras (12) and (14), three or more CCD cameras can be used. Further, in the above embodiment, the coordinate rotation processing is performed on the collation data, but a method of performing the coordinate rotation processing on the registered data can also be adopted.

[Brief description of the drawings]

FIG. 1 is a diagram showing a configuration of a personal identification device according to the present invention.

FIG. 2 is a partially cutaway perspective view of a detector main body.

FIG. 3 is a flowchart showing the operation of the personal identification device according to the present invention.

FIG. 4 is a flowchart showing a data collation procedure.

FIG. 5 is a diagram illustrating a series of image processing and feature amount extraction processing in the personal identification device according to the present invention.

FIG. 6 is a diagram showing a configuration of a conventional personal identification device.

FIG. 7 is a flowchart showing the operation of a conventional personal identification device.

FIG. 8 is a diagram illustrating a feature amount extraction process in a conventional personal identification device.

[Explanation of symbols]

 (1) Detector (10) Detector body (11) First light source (12) First CCD camera (13) Second light source (14) Second CCD camera (15) Finger insertion hole (2) Image processing circuit (3) Arithmetic processing circuit (4) Light source control circuit (5) Memory (6) Information processing device

Claims (3)

[Claims] 1. An apparatus body provided with a finger insertion hole (15),
An imaging device for three-dimensionally capturing a blood vessel image of a finger in the finger insertion hole (15) is provided, the imaging device is connected to an information processing circuit, and the information processing circuit is obtained from the imaging device. A feature amount extracting means for recognizing a three-dimensional blood vessel pattern based on the obtained blood vessel image and extracting a plurality of geometric feature amounts from the blood vessel pattern; and representing in advance a feature amount of one or more specific persons. A feature amount registration unit in which data is registered, and a feature amount extraction instruction is given to the feature amount extraction unit in response to the collation instruction, and data representing the extracted feature amount is registered in the feature amount registration unit. Feature value matching means for comparing the existing data and the similarity between the two data, and permission information issuing means for issuing permission information only when the similarity between the two data is determined to be high. , The feature collator , Upon verification of both the data, either one of the data, performing coordinate rotation processing center axis of the finger insertion hole (15),
A personal identification device characterized by determining the similarity between the correction data obtained thereby and the other data. 2. The feature amount matching unit repeats the coordinate rotation process and the similarity determination for each fixed rotation angle, and determines that the similarity is high when the similarity of both data exceeds a predetermined threshold. The personal identification device according to claim 1. 3. The image pickup apparatus comprises at least two CCD cameras arranged in a direction perpendicular to the longitudinal direction of the finger insertion hole (15) of the apparatus main body, and the feature amount extracting means includes a finger insertion part of the apparatus main body. An image processing means is provided that combines two two-dimensional images obtained from the two CCD cameras and generates a three-dimensional image of a blood vessel with the back end position of the hole (15) or a position in the vicinity thereof as the origin. The personal identification device according to claim 1 or 2.