JP4344986B2 - Authentication method and authentication apparatus - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to an authentication method and authentication equipment, and is preferably applied to a case of personal authentication in particular using biometric.
[0002]
[Prior art]
Fingerprints, palm prints, etc., which are widely used for personal authentication, are the parts where the ridge network formed by the skin's epidermal tissue sinking into the uneven structure of the dermis can be seen directly from the outside. It reflects the deep skin structure. For the purpose of making the tactile nerve terminals distributed in the deep skin more easily detect external stimuli and physiological reasons such as the strength against friction, the skin of the palms and soles is different from the skin of other parts, such as the dermis. It has a unique skin structure in which the shape of the deep skin structure matches the shape of the epidermis. Fingerprints conventionally used for personal authentication basically use the permanence of this deep structure.
[0003]
However, the biometric authentication using the above-mentioned fingerprint is not always sufficient for the so-called “spoofing”. For example, since fingerprints easily remain as traces on other objects and are easily visible, the risk of being counterfeited by a third party cannot be denied.
[0004]
Therefore, in recent years, a personal authentication technique using a blood vessel pattern as identification information has been proposed instead of the biometric authentication using the fingerprint. For example, Japanese Patent Laid-Open No. 10-295675 proposes a personal authentication method for viewing the vein pattern of the back of the hand, and Japanese Patent Laid-Open No. 7-21373 proposes a personal authentication method for viewing the blood vessel pattern of the finger. Has been.
[0005]
[Problems to be solved by the invention]
However, in the former case (a method for viewing the vein pattern on the back of the hand), there is a problem in that the data collection area is large and the apparatus is enlarged accordingly. When the apparatus is enlarged, the application is naturally limited, and for example, it is difficult to obtain a wearable configuration.
[0006]
On the other hand, the latter (method of viewing the blood vessel pattern of the finger) is advantageous in terms of miniaturization of the device, but there is a problem that there are few blood vessel patterns that can be used as identification information, and there remains concern about the reliability of authentication.
[0007]
In the case of a living body, unlike the case of an industrial product or the like, for example, the shape of each finger is different, so that there is a problem of alignment when acquiring an image used for authentication. Since there is no mark that serves as a reference for alignment in a living body, alignment during image acquisition is difficult.For example, if the acquisition position of the acquired image is significantly different from the pre-registered blood vessel pattern acquisition position, verification is not possible. May be possible.
[0008]
The present invention has been made in consideration of the above points, and intends to further improve the reliability of the authentication result.
[0009]
[Means for Solving the Problems]
In order to achieve the above-described object, the present invention provides a light source and a light receiving element provided alternately on the inner surface side of a ring-shaped housing to which a finger is fitted along the circumferential direction of the inner surface of the housing. Using the opposed light source and light receiving element as a pair, a transmission image in the direction of each pair is acquired, and a vein pattern of the transmission image is extracted. Then, you against the vein pattern registered with the vein pattern.
[0010]
Since not only images from one direction but also images from a plurality of directions around the finger are collected three-dimensionally, the number of vein patterns that can be used is greatly increased, and highly reliable personal authentication can be realized. Further, since it can be fitted on a finger like a ring, it is possible to easily prevent a positional deviation between the registered image and the collation image, and a so-called wearable downsizing configuration can be realized.
[0012]
DETAILED DESCRIPTION OF THE INVENTION
Hereinafter, an authentication method and an authentication apparatus to which the present invention is applied will be described in detail with reference to the drawings.
[0013]
The authentication method of the present invention uses, for example, a blood vessel pattern such as a finger as authentication information. Therefore, at the time of authentication, it is necessary to take a transmission image of the blood vessel pattern. Here, the transmission of the vascular pattern transmission image may be performed using a known technique. For example, the principle of transmission images of finger blood vessels is described in the literature by Koichi Shimizu et al., “Pathophysiology Vol.11 No.8 1992, 8 Possibility of biological fluoroscopy with light-CT and biological functional imaging”; "Sysmex Journal Vol.2 No.1 1999 Non-invasive measurement technology" and the like.
[0014]
The near-infrared band having a wavelength of 700 to 1200 nm has a specific low absorbance in a living body and is called a “spectral region window”, and penetrates living tissue well. Epidermal tissue, bone, and the like have the property of reflecting and scattering visible light and ultraviolet light, but about 80% of light in this band is transmitted. On the other hand, in the near-infrared band having such characteristics, there is a wavelength that is easily absorbed by hemoglobin in blood. For example, at a wavelength of 805 nm, oxygenated hemoglobin (HbO ₂ ) and reduced hemoglobin (Hb) Absorbance matches. Furthermore, the spectral characteristics of hemoglobin and water in the living body are greatly different. By utilizing this characteristic, a transmission image of the blood vessel pattern can be obtained by distinguishing it from the moisture of the living body.
[0015]
This blood vessel blood flow pattern is unique to the living body, and when the tissue is cut from the living body, it immediately disappears due to vascular atrophy, blood flow cessation, blood loss, or the like. Therefore, biometric authentication and biometric affiliation recognition coincide with each other. For this reason, even if the cut tissue is immersed in physiological saline or the like to save the cells, the blood flow does not exist, so that authentication can be excluded. The tissue to be certified must have the pulmonary circulation and pulsation and the correct blood flow and blood hemoglobin ratio. Even if the arm is surgically cut and used, the blood vessels of the arm are surgically removed. It is necessary to connect to a heart-lung machine and accurately reproduce the pulsation waveform. For example, it is difficult to realize in a present situation where a portable heart-lung machine has not been put into practical use. Even if it is put to practical use in the future, it begins with the cutting of the arm, connection to each blood vessel and device, treatment for the cut microvessels and nerves, elimination of tissue changes due to living reactions to cutting, and resumption of blood flow It requires advanced surgical techniques and medical equipment, such as stabilization of later tissues, and is not a realistic task. On the other hand, it is almost impossible to artificially reproduce a three-dimensional structure of fine capillaries without using a living body.
[0016]
In the present invention, in order to capture a transmission image of a blood vessel pattern, for example, as shown in FIG. 1, a light source 1 and an imaging camera 2 are arranged facing each other with a finger 3 as a measurement target portion interposed therebetween, and the finger 3 is Imaging is performed using transmitted light. As the light source 1, a light source that emits long-wavelength light (near infrared light) that is specifically absorbed by hemoglobin in blood that passes through bones and flows through blood vessels, for example, a near infrared LED or the like is used.
[0017]
At this time, it is preferable to block the external light by mounting the device in the box 4. Further, for example, it is preferable to suppress as much as possible light that is reflected from the side of the finger 3 and enters the camera, thereby suppressing unevenness in photographing. For this reason, the slits 5 and 6 that are open only in the vicinity of the center are arranged to restrict the imageable area of the imaging camera 2, and the transmitted light is imaged by the imaging camera 2 through these slits.
[0018]
In addition, in order to obtain an image of the entire periphery of the finger 3, the finger 3 is arranged so that the center of the finger 3 is located on the center axis of the turntable 7 in which the imaging camera 2 and the light source 1 are paired. A moving image is acquired while rotating. Among the acquired moving images, a vertically continuous pixel column of the central portion in the horizontal direction is acquired from each frame, and these are combined to obtain an all-around developed image of one finger. As the imaging camera 2, it is preferable to use a camera having a visual field in the range of the first to third joints of the finger.
[0019]
The image obtained as described above is subjected to image processing and collated with a previously registered blood vessel pattern for personal authentication. Prior to this, the image must be aligned. The collation is meaningless unless the blood vessel patterns at the same place are compared. Therefore, here, alignment is performed using the texture of the finger. Specifically, since the texture of the skin surface of the second joint part of the finger (second joint part skin surface pattern) has a very clear contrast, calibration is performed based on this.
[0020]
The second joint skin surface pattern has a greater absorption in near-infrared region light (for example, light having a wavelength of 880 nm) than other sites, such as a relatively large skin thickness. For this reason, when near-infrared light is imaged as a light source, the image is clearly captured as a region with low brightness. Using this feature, this region can be used as a reference point for alignment for image standardization.
[0021]
As a specific method of alignment, the acquired image is divided into, for example, 32 × 32 blocks, the average brightness in the area is calculated, and a portion where dark areas are gathered similarly around the darkest part. Mark as a candidate. Then, it is examined whether or not a characteristic texture (a plurality of dark lines in the lateral direction) of wrinkles on the skin surface is detected in each region.
[0022]
After the above alignment, a vein pattern is extracted. During the extraction process, the acquired image is subjected to appropriate image processing, so that the positional relationship between the specimen, the light source, and the imaging camera is blurred, the skin surface texture, adhering dust, measurement instrument noise, light from the outside world, etc. A standardized image from which disturbance elements such as contamination are removed is acquired. An example of the steps for obtaining this standardized image is shown in FIG.
[0023]
In this example, first, a rectangular area that can be used as data is cut out from the captured image (area cutout).
[0024]
Next, background lightness correction is applied to the cut-out image area. For example, average the intensity of a 48 × 48 square block, set the intensity as the background brightness, extend the size while complementing with the bicubic method (bicubic method), invert the intensity, take the sum, and calculate the intensity average by half Correction is performed so that it becomes (127 at 256 gradations) (background lightness correction). The bicubic method is an image interpolation method called a cubic interpolation method, and has a feature that a natural image can be obtained with little loss of information.
[0025]
Further, noise removal is performed by applying a median filter to the image corrected as described above (noise removal). The median filter gives a median value in the local region as an output density, and can remove noise without blurring the edge of the image.
[0026]
Next, since the dispersion of the maximum value and the minimum value of the intensity of the obtained image varies, in order to correct it, the intensity is normalized so that the minimum value is 0 and the maximum value is 255. (Image intensity information normalization).
[0027]
Then, a band pass filter is applied in the spatial frequency domain (high frequency removal filter application and low frequency removal filter application). This is because the blood vessel image is affected by the scattering of the living body, so it contains almost no high-frequency components, and the low-frequency components are mostly due to uneven brightness during imaging and differences in finger thickness depending on the part. This is because the high transmission intensity is globally different. In particular, since many skin wrinkle patterns are included as high-frequency components, the skin surface pattern can be effectively removed by cutting the high-frequency components.
[0028]
The vein pattern extracted as described above is stored as image information and registered as identification information. Alternatively, personal authentication is performed by collating with a vein pattern registered in advance. This personal authentication step is shown in FIG.
[0029]
For personal authentication, first, the previously stored image information is acquired from the image database (image acquisition from the image database). The resolution of the acquired image is optimized (image resolution optimization), and alignment is performed using the reference point (position shift).
[0030]
After registration, personal authentication is performed by collation. In collation, a correlation value is calculated (correlation value calculation), and a maximum correlation value is acquired (maximum correlation value acquisition). Then, an image having the maximum correlation value is acquired, and personal authentication is performed based on the acquired image.
[0031]
In the above-described authentication method and authentication apparatus, the blood vessel pattern of the finger is not photographed only from one direction, but is collected three-dimensionally, for example, as an all-around developed image. Significant increase and reliable authentication is possible. Moreover, since sufficient identification information can be obtained even in a small measurement target site such as a finger, the apparatus can be easily downsized.
[0032]
By the way, the configuration of the authentication device is not limited to the above example. For example, a plurality of light source-imaging camera pairs may be prepared and arranged around a finger or the like. Alternatively, a ring-shaped device can be used.
[0033]
FIG. 4 shows an example of an authentication device configured in a ring shape. The authentication device 10 can be fitted on a finger like a ring, and can have a so-called wearable configuration. In the ring-shaped authentication device 10, as shown in FIG. 5, the light source elements 11 and the light receiving elements 12 are alternately arranged along the circumferential direction on the inner surface facing the finger 3, and the light source elements 11 facing each other. And the light receiving element 12 capture a transmission image at each angle.
[0034]
In the case of the above wearable configuration, a blood vessel pattern is detected and compared with a pattern registered in advance in an authentication device or a server connected to the authentication device over a network, and the result is provided at least from the authentication device or the network. It is possible to realize so-called access control that permits or restricts at least a part of the service.
[0035]
【The invention's effect】
According to the present invention as described above, by which is adapted to the image three-dimensionally collected from a plurality of directions in the finger around is fitted on a finger as rings, from a limited stbm many patterns Information can be acquired, and highly reliable personal authentication can be realized. At the same time, it is possible to easily prevent misregistration between the registered image and the collation image, and to realize a so-called wearable downsizing configuration.
[Brief description of the drawings]
FIG. 1 is a schematic diagram showing an example of an authentication apparatus to which the present invention is applied.
FIG. 2 is a flowchart illustrating an example of steps for acquiring a standardized image.
FIG. 3 is a flowchart illustrating an example of steps for performing personal authentication.
FIG. 4 is a schematic perspective view showing an example of a ring-shaped authentication device.
FIG. 5 is a schematic plan view showing the configuration of the inner peripheral surface side of the ring-shaped authentication device.
[Explanation of symbols]
DESCRIPTION OF SYMBOLS 1 Light source, 2 Imaging camera, 3 Finger, 5, 6 Slit, 7 Turntable, 10 Ring-shaped authentication apparatus, 11 Light source element, 12 Light receiving element

Claims (5)

A ring-shaped housing in which fingers can be fitted;
A light source that is provided on the inner surface side of the housing and that emits near-infrared light; and
A light receiving element provided on the inner surface side of the housing and facing the light source;
An acquisition means for acquiring a transmission image using the light source and the light receiving element;
Extracting means for extracting the vein pattern of the transparent image;
Using the vein pattern extracted by the extraction means, and a matching means for matching with a registered vein pattern,
The light source and the light receiving element are alternately provided along the circumferential direction of the inner surface of the housing,
The acquisition means is
An authentication device that acquires a transmission image in the direction of each pair by using a light source and a light receiving element that face each other as a pair. Generating means for combining the transmission images in the direction of each pair and generating a developed image around the finger;
Position determining means for determining position information indicating an area serving as a reference when extracting the vein pattern of the developed image;
The extraction means is
The authentication apparatus according to claim 1, wherein a vein pattern is extracted based on the position information. The extraction means is
The authentication apparatus according to claim 1, further comprising a step of normalizing the transmission image of the blood vessel in the direction of each pair so that the maximum value and the minimum value of the intensity in the image are the same. The extraction means is
The authentication apparatus according to claim 1, further comprising a step of removing a high-frequency component corresponding to a skin wrinkle pattern from a transmission image of a blood vessel in the direction of each pair. Of the light source and the light receiving element that are alternately provided along the circumferential direction of the inner surface of the casing on the inner surface side of the ring-shaped casing on which the finger is fitted, the opposed light source and the light receiving element are used as a pair. An acquisition step of acquiring a transmission image in the direction of each pair;
An extraction step of extracting the vein pattern of the transmission image;
An authentication method comprising: a verification step of verifying with a registered vein pattern using the vein pattern extracted in the extraction step.

Images