JP4232079B2 - Personal identification device - Google Patents

Description

[0001]
[Technical field to which the invention belongs]
The present invention relates to a personal identification device using individual differences in fingerprints or blood vessel patterns under the skin as identification information.
[0002]
[Prior art]
The inventor of the present application has previously made many proposals regarding a personal identification device using fingerprints (Japanese Patent Laid-Open No. 11-296652, etc.).
[0003]
On the other hand, research using biometric information other than fingerprints as personal identification means has also been conducted. For example, a method of performing personal authentication using a blood vessel pattern of a finger (Japanese Patent Laid-Open No. 7-21373), a vein pattern on the back of a hand A method for performing personal authentication using JP-A-10-295674 is disclosed.
[0004]
In addition, these methods fix the finger or back of the hand to a part of the device by any method, limit the degree of freedom of the position, extract the blood vessel pattern of those biological parts, and perform personal authentication using them However, Japanese Patent Laid-Open No. 2002-92616 discloses a method for reading personal blood vessel patterns in a non-contact / non-fixed state without using finger or hand fixation, and performing personal authentication using them. ing.
[0005]
[Problems to be solved by the invention]
However, in the case of personal identification by the former fingerprint, there may be a case where only an amount of information that is insufficient to identify an individual can be extracted due to degradation of the fingerprint or the like. For example, the number of ridge branch points and ridge end points, which are basic feature data of fingerprints, cannot be extracted as necessary for personal identification. Of course, such fingerprints are used for personal identification. It is not possible.
[0006]
Fingers of people who are constantly in contact with chemicals or who require intense wear on their hands or skin may be difficult to extract enough information to identify individuals because of the deterioration of their fingerprints. .
[0007]
In rare cases, all 10 fingers may not be used for personal identification, and such individuals cannot use the fingerprint identification device. It is required to be identifiable.
[0008]
On the other hand, in the case of personal identification by the latter blood vessel pattern, the blood vessel pattern (mainly veins) located in a relatively shallow layer of the skin is much simpler than the pattern of fingerprint ridges. Since it is far less than the number, there is an essential defect that there is little feature data available for personal identification.
[0009]
Further, in the case of a person with thick finger or hand skin, the blood vessel pattern cannot be read sufficiently, and it is often difficult to identify an individual alone.
[0010]
[Means for Solving the Problems]
Accordingly, in view of the above circumstances, the present invention provides a fingerprint image reading unit, means for extracting fingerprint feature data from a fingerprint image read by the unit, a blood vessel pattern image reading unit, and a blood vessel read by the reading unit. A means for extracting blood vessel pattern feature data from the pattern image is provided, and an area for saving the extracted feature data is provided, and the extracted fingerprint feature data or blood vessel pattern feature data or fingerprint feature data and blood vessel pattern The present invention proposes a personal identification device for performing personal identification using data obtained by integrating feature data.
[0011]
BEST MODE FOR CARRYING OUT THE INVENTION
That is, in the present invention, if the fingerprint is determined to be ineligible for personal identification due to lack of extracted fingerprint feature data at the fingerprint identification stage, the fingerprint feature data is temporarily saved and stored. The blood vessel pattern image is collected, the characteristic data of the blood vessel pattern image is extracted, and the individual identification is attempted.
[0012]
Similarly to the case where the extracted blood vessel pattern feature data is a fingerprint, if it can only provide an amount of information that is insufficient for personal identification by itself, the feature data of the temporarily saved fingerprint is derived and referenced. By doing this, the person is identified.
[0013]
As described in Japanese Patent Application Laid-Open Nos. 7-21373 and 10-295674, fixation to a part of a finger or hand device is important for collecting blood vessel pattern images. In the personal identification apparatus according to the above, since the finger is fixed by pressing the finger to the fingerprint image reading unit in use, the blood vessel pattern image can be easily collected.
[0014]
Also, blood vessel pattern collection sites such as fingers or hands may be close to the fingerprint region, so the fusion of both and the compatibility of both technologies are quite good, such as associating the extracted fingerprint feature data with the blood vessel pattern feature data on coordinates. When fingerprint feature data and blood vessel pattern feature data are integrated, feature data with high accuracy in personal identification can be provided.
[0015]
Therefore, in the present invention, it is possible to perform individual identification using only one of the fingerprint feature data or the blood vessel pattern feature data, but the combination of the two may significantly cause the determination of nonconformity to the user. Can be reduced.
[0016]
【Example】
Hereinafter, the present invention will be described in detail with reference to the embodiments shown in the drawings. The personal identification device according to the present invention has a CPU 1 as an overall control area, a first memory 2 as a work area, fingerprint feature data or blood vessel pattern feature data. The second memory 3 and the external interface 4 are provided as a saving area, and a key input device 36 and registered biometric data 38 are connected to the external interface 4.
[0017]
Further, the personal identification device includes a blood vessel pattern reading unit 9 including a fingerprint image reading unit 5, a blood vessel pattern image reading near infrared light source unit 6, a blood vessel pattern image reading conversion control unit 7, and a blood vessel pattern image reading camera 8. Have.
[0018]
It is convenient to collect the blood vessel pattern of the finger near the fingerprint collecting position or the blood vessel pattern of the back of the hand because the apparatus can be miniaturized.
[0019]
As shown in FIG. 2, a method for collecting a blood vessel pattern of a finger irradiates light from the abdomen of the finger and receives transmitted light on the dorsal side of the finger and light from the dorsal side of the finger. A reflection reading method (B) in which irradiation is performed and reflected light is received on the back side of the same finger can be considered.
[0020]
Specifically, referring to FIG. 2, the personal identification device includes a fingerprint image reading unit 5 provided inside the fingertip and a blood vessel pattern reading unit 9 provided on the finger. The blood vessel pattern reading unit 9 shields the finger from light. Covered with a cover 11 and provided with a light receiving unit 10 comprising a light source unit 6 and a finger blood vessel pattern reading camera 8 and the like, and in the transmission reading method (A), a light source unit 6 is provided on the ventral side of the finger, A camera 10 is provided on the back side of the finger. In the reflection reading method (B), a light source unit 6 and a light receiving unit 10 are provided on the back side of the finger with a reflecting plate 12 around.
[0021]
FIG. 3 shows a specific example of a method for collecting the blood vessel pattern on the back of the hand. This apparatus is composed of a fingerprint image reading unit 5 provided inside the fingertip and a blood vessel pattern reading unit 9 provided on the back of the hand. The blood vessel pattern reading unit 9 includes a light source unit 6 and a light receiving unit 10 provided with a reflection plate 12 around the back of the back of the hand, as in the reflection reading method.
[0022]
Here, as the light source unit 6, a light source unit in the near-infrared wavelength region having a high transmittance with respect to the skin can be used, and in order to increase the sensitivity to image acquisition of the blood vessel pattern, the light shielding cover 11 is used. Can cut the ambient light in the near-infrared wavelength region, which has a high transmittance to the skin.
[0023]
Even if the light-shielding cover 11 that cuts off ambient light in the near-infrared wavelength region, which has a high transmittance to the skin, is installed so that other visible light can be transmitted, the user can read the fingerprint when reading the fingerprint. There will be no inconvenience or trouble in positioning.
[0024]
Next, the processing of the personal identification device according to the present invention will be described based on the embodiment of FIG. 3. First, one processing process for registering personal feature information will be described with reference to FIG.
[0025]
The fingerprint image reading unit 5 is activated by the processor 1 that controls the entire apparatus, and the fingerprint image is collected (FIG. 6-24).
[0026]
From the collected fingerprint image, the fingerprint ridge end point and the branch point where the ridge branches are extracted as fingerprint feature points (Fig. 6-25), and a fingerprint feature data group is created from them.
[0027]
In detail, the extracted end points and branch points have different ridge thicknesses, and this difference can relate the reliability of the end points and branch points. An average is taken and assigned to the feature point as direction information.
[0028]
Reference numeral 19 in FIG. 5 shows a fingerprint feature data group in which end points and branch points are represented by direction information, vertices of an isosceles triangle, and reliability is represented by the size of the area of the triangle.
[0029]
This extraction operation is performed until the number of feature points consisting of the end points and branch points of the ridges found and extracted exceeds the limit number (fingerprint feature amount threshold, for example, 12 to 15 in the judiciary). If judged (Fig. 6-26), the fingerprint feature data group is registered (Fig. 6-27).
[0030]
If the number of feature points consisting of the end points and branch points of the ridges found and extracted does not reach the threshold value, it is judged that it is insufficient to identify the person based on this. In this case, if registration is allowed with a shortage of information, a trouble that may be mistaken for another person at the time of collation may occur, so the feature data group registration of the extracted fingerprint is rejected (FIG. 6-26).
[0031]
If this fingerprint information registration refusal occurs, the extracted fingerprint feature data is temporarily saved in the second memory 3 (FIG. 6-28).
[0032]
Then, the fingerprint reading unit 5 is turned off, the blood vessel pattern image reading light source unit 6, the blood vessel pattern image reading conversion control unit 7, the blood vessel pattern image reading camera 8 and the blood vessel pattern reading unit 9 are turned on to turn on the blood vessel pattern. The process proceeds to reading of blood vessel pattern feature information in the angle of view (image area) 13.
[0033]
Here, the angle of view (image area) 13 of the blood vessel pattern irradiates a part of the back of the hand with the light source unit 6 having the reflector 12 and reads the image information reflected in the fixed area 13 by the light receiving unit 10. (Figure 6-30).
[0034]
After the collection of the blood vessel pattern image is completed, image preprocessing such as noise removal and blood vessel image enhancement processing is performed on the image data (Fig. 6-31), and branch points are extracted from the blood vessel pattern image (Fig. 6-32), personal feature data of blood vessel pattern image information is extracted by the following method.
[0035]
That is, the blood vessel pattern is regarded as an aggregate of line segments (hereinafter referred to as blood vessel line segments) divided by branch points and image region boundaries, and blood vessel pattern image information collected thereon is used as personal feature data as blood vessel data. Flow direction distribution, blood vessel thickness distribution, branch point position, distance and direction between branch points, connectivity, and the like (hereinafter referred to as blood vessel pattern feature data).
[0036]
This outline will be explained with reference to FIG. 4. From the blood vessel pattern image of the back of the hand, branch points such as branch point (a) 14, branch point (b) 15, branch point (c) 16, branch point (d) 17 are extracted. To do.
[0037]
Here, when the blood vessel line segment 19 connecting the branch point 14 (a) and the branch point 15 (b) and the blood vessel line segment 18 connecting the branch point 16 (c) and the branch point 17 (d) are compared, Although the thickness is different, this difference can be related to the reliability of the bifurcation point.
[0038]
That is, it is determined that the reliability of the branch point belonging to the thick blood vessel line segment is high, and in the example of FIG. 4, the branch point 14 belonging to the relatively thin blood vessel line segment is less reliable than the branch points 16 and 17. It will be.
[0039]
In this way, the weights of feature points (only branch points in the case of blood vessels) are defined, and the average of the directions of connected blood vessel line segments is taken as with fingerprints, and this is given to the feature points as direction information. A feature data group of a blood vessel pattern is used.
[0040]
It should be noted that the blood vessel pattern collection site such as a finger or a hand is close to the fingerprint site, suggesting that the extracted fingerprint feature data and the blood vessel pattern feature data can be correlated on the coordinates as shown in FIG. is doing.
[0041]
Then, by extracting the feature data of both in association with, for example, the coordinate information of the branching point, the fusion of both is good, and it is possible to obtain a feature data group with high accuracy when integrating.
[0042]
For this reason, the feature point data group 19 extracted from the fingerprint is schematically shown in the embodiment shown in FIG. 5, but the coordinate information of the fingerprint branch point is also obtained for the angle of view (image area) 13 of the blood vessel pattern. The feature data group of the blood vessel pattern is based on the direction (vertices of the isosceles triangle) introduced and extracted from the blood vessel pattern and the reliability (the size of the area of the triangle).
[0043]
The blood vessel pattern feature data is extracted until the blood vessel pattern feature amount threshold value is exceeded, and if it is judged sufficient (FIG. 6-33), the blood vessel pattern feature registration is performed (FIG. 6-34).
[0044]
Since the feature data extracted from the fingerprint and the blood vessel pattern feature data can be handled in the same manner as described above, if the feature data is insufficient in the determination of FIG. By integrating the blood vessel pattern feature data obtained in FIG. 6-32 (FIG. 6-35), it is possible to create feature data having an information amount sufficient for personal identification.
[0045]
In addition, the case where it was decided that personal identification was possible only with the fingerprint feature data and the information registration was completed, the case where only the feature data of the blood vessel pattern was registered, or the combination of both was used as the feature data for personal identification. Set an identifier (type flag) to distinguish the type in each case (Figs. 6-27, -34, -35), and input personal information such as name from the keyboard 36 connected to the external interface 4 of the device Then, they are added and registered as registered biometric data in the database 38 (FIG. 6-37).
[0046]
In this embodiment, the extracted fingerprint feature data is temporarily saved in the second memory 3. However, the blood vessel pattern feature data is temporarily saved in the second memory 3 to obtain the fingerprint feature data. Extraction may be performed.
[0047]
Next, an embodiment of the personal authentication process will be described with reference to FIG. 7. A user who needs to verify his / her identity is asked to input personal information such as a name, and the biometric related information is registered from the registered biometric database 38 based on the personal data. Read the information (Figure 7-41).
[0048]
The type of biometric information registered from this data is read, and the biometric reading process for each type is performed according to the type (FIG. 7-42).
[0049]
When extracting only fingerprint feature data, perform fingerprint image data collection (Figure 7-24) and fingerprint feature data extraction (Figure 7-25). When extracting only blood vessel pattern feature data, a blood vessel pattern image When collecting (Figure 7-30), blood vessel pattern image processing (Figure 7-31), blood vessel pattern feature data extraction (Figure 7-32), and integrating both, fingerprint and blood vessel pattern features After extracting the data, each data is integrated (Figure 7-35). After each data is extracted, the feature data is collated (Figure 7-44) and the collation is finished (Figure 7-45). .
[0050]
In this embodiment, the blood vessel pattern of the fingerprint and the back of the hand has been described with reference to FIG. 3. However, when the blood vessel pattern of the fingerprint and the finger is used as shown in FIG. The feature data is extracted from the blood vessel pattern image, but the finger direction 20 is calculated from the direction distribution of the blood vessel line segment of the extracted feature data. The fingerprint information area is on the extension of the finger direction. .
[0051]
That is, it is possible to associate the finger direction 20 that can be calculated in this way with the feature information of the fingerprint and the blood vessel pattern of the finger on the same coordinates. For example, the user who fixed the finger to the fingerprint reader When the purpose is not completed due to inconvenience of the finger state, by extracting the finger direction from the average of the blood vessel direction in the feature amount of the blood vessel pattern and referring to it when extracting the feature data group of the fingerprint When integrating the two, it can be used as more accurate feature data of a fingerprint.
[0052]
【The invention's effect】
In short, according to the present invention, it is possible to perform personal identification using only one of fingerprint feature data or blood vessel pattern feature data, but combining them can cause incompatibility to the user. Can be significantly reduced.
[Brief description of the drawings]
FIG. 1 is a block diagram of a personal identification device according to an embodiment of the present invention. FIG. 2A is a diagram illustrating an embodiment in which light is emitted from the ventral side of a finger and transmitted light is transmitted from the back side of the finger. Fig. 3B is a schematic diagram of a transmission reading system that receives light from the back of the finger. Fig. 3B is a schematic diagram of a reflection reading system that irradiates light from the back side of the finger and receives reflected light on the back side of the same finger. FIG. 4 is a diagram showing an example of a blood vessel pattern image of the back of the hand collected by the above-described device. FIG. 5 is a schematic diagram of feature data extracted from the blood vessel pattern image. FIG. 6 is a flowchart of one process for registering biometric feature information according to the present invention. FIG. 7 is a process for verifying biometric feature information according to the present invention. Flowchart [Fig.8] Diagram relating characteristics data of the blood vessel pattern and a fingerprint of a finger in the invention of

Claims (3)

A fingerprint image reading unit provided inside the fingertip, a blood vessel pattern reading unit provided on the finger, means for extracting fingerprint feature data from a fingerprint image read by the fingerprint image reading unit, and the blood vessel pattern And a means for extracting blood vessel pattern feature data from the blood vessel pattern image read by the reading unit. When extracting the fingerprint feature data from the fingerprint image read by the fingerprint image reading unit, the blood vessel pattern reading unit reads the fingerprint feature data. A personal identification device that calculates a finger direction from a direction distribution of blood vessel line segments of blood vessel pattern feature data extracted from a blood vessel pattern image, and extracts fingerprint feature data by referring to the finger direction. 2. The personal identification device according to claim 1, wherein the blood vessel pattern reading unit comprises a light source unit and an image pickup unit for reading a blood vessel pattern image under the skin of the finger. The personal identification device according to claim 1, wherein a near-infrared light source unit that transmits a finger is used as the light source unit, and a light-shielding cover that shields a near-infrared wavelength region from outside light in a light source wavelength component as a light-shielding cover.

Images