JP2010170319A - Image obtaining device and authentication device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To improve authenticating precision while suppressing the deterioration of comfortability to the absolute minimum even when the state of a living body is not satisfactory without impairing comfortability needlessly for a large majority of uses. <P>SOLUTION: An imaging apparatus 100 picks up a living body image 101 by imaging the finger 901, and a living body image obtaining part 210 obtains the living body image 101 from the imaging apparatus 100. A feature information extraction part 220 extracts feature information 102 from the living body image 101, and an extracted information quantity evaluation part 230 evaluates whether or not the feature information 102 is quality necessary for authentication processing. When the feature information 102 is quality NG, a gain element rate determination part 250 changes a control value (light quantity Gl, accumulation time Gt, diaphragm quantity Gs, analog gain Ga, digital gain Gd, image addition number of sheets Gn, software gain Gs1) of the imaging apparatus 100, and makes the imaging apparatus 100 re-pick up the living body image 101. When the feature information 102 is quality OK, a personal authentication part 240 authenticates a user based on the feature information 102. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  The present invention relates to an image acquisition device and an authentication device used for biometric authentication, for example.

  In a biometric authentication apparatus such as fingerprint authentication, biometric feature information read from a user is stored in advance as registration data 291, and the feature information read from the user at the time of authentication is used as verification data between verification data and registration data 291. The degree of coincidence is evaluated to determine whether or not the user is valid.

  However, high-quality biometric image data is not always obtained during registration and authentication. For example, the fingerprint pattern becomes unclear due to dry skin, moisture, or health condition, and is difficult to read by the imaging device. Then, the fingerprint pattern in the captured image is blurred or crushed. For this reason, the quality of the feature information obtained from the captured image by image processing is lowered, and as a result, the authentication accuracy may be lowered.

As a conventional technique for obtaining a high-quality input image from a living body such as a fingerprint, there is a technique for controlling the gain of amplification of an image signal to increase the luminance of the image until just before saturation (Patent Document 1).
This conventional technique performs imaging a plurality of times, analyzes the luminance distribution of the input image up to the previous time, and determines the optimum gain value at the next imaging. This prior art uses the average luminance value of an image as an index for controlling the gain value.
In the captured image, signal components indicating biological information (ridge information in the case of fingerprints), thermal noise of the image sensor, electrical noise on the electronic circuit, quantization during A / D conversion Noise components such as noise are superimposed. For the authentication performance of the apparatus, it is important how to accurately extract only the signal component from the input image in which the signal component and the noise component are superimposed. That is, high authentication performance can be obtained when the signal component is sufficiently larger than the noise component, but high authentication performance cannot be obtained when the signal component is small compared to the noise component.
The average luminance value used in this conventional technique is a value obtained with the noise component and the signal component superimposed, and it is determined whether the signal component is sufficiently large compared to the noise component. It is not an indicator to represent. For example, when a finger with a thin fingerprint pattern, that is, a small signal component is imaged, the noise component is amplified to the same extent even if the gain is increased to increase the average luminance value. As a result, since the ratio between the noise component and the signal component does not change, it is difficult to accurately extract only the signal component.
Therefore, the conventional technique has a problem that the authentication performance cannot always be improved.

On the other hand, in such a case, in order to improve the authentication accuracy, it is considered that the imaging apparatus may be controlled using the quality of the feature information directly linked to the authentication result as an evaluation criterion.
As a technique for evaluating the quality of feature information, there is a technique for improving authentication accuracy by reducing the weighting of low-quality areas (Patent Document 2).
However, there is no technique that uses the evaluation value for controlling the imaging apparatus and reduces the low-quality area from the input stage and applies it to improve the quality of the feature information.

JP-A-10-240913 JP 2007-102422 A Japanese Patent No. 4020202

“A. Monden, S. Yoshimoto,“ Fingerprint Identification Usage the Accidental Coincidence Probability, ”MVA2002, pp. 124-127, Dec11-13, 2002”.

In order to extract high-quality feature information even when the fingerprint pattern is thin, it is necessary to control to increase the Signal / Noise ratio (S / N ratio) of the image. The imaging device or image processing control element that contributes to the S / N ratio includes a light source light amount, an aperture opening amount (when the image sensor is an optical camera), an exposure light amount control value such as an exposure time, and a plurality of continuous images. There is a process of taking an image and performing addition and averaging operations. However, as a compensation for the improvement of the S / N ratio, the user's comfort is reduced, such as the glare associated with the increase in the amount of light, the increase in the rest time of the user's finger associated with the increase in the exposure time and the number of continuous images. There is.
Therefore, if the amount of light or the number of images is simply increased in order to increase the S / N ratio, comfort will be reduced for many users.
Although it is desirable to reduce the comfort level to the minimum number of users, no technology has been shown so far on how to limit the level of comfort to the minimum number of users.
Therefore, a biometric authentication device capable of gain control is desired while maintaining a balance between the quality of the input image and the operability of the user according to the state of the user's biometric.

  The present invention has been made, for example, in order to solve such problems, and does not unnecessarily impair comfort for the majority of users, and is comfortable even when the living body is in poor condition. The accuracy of authentication is improved while minimizing the deterioration of security.

  An image acquisition device according to the present invention uses an image acquisition unit that acquires an image captured by an imaging device, and a CPU (Central Processing Unit) to determine whether the image acquired by the image acquisition unit satisfies a predetermined condition. And the image determination unit, and the image determination unit determines that the image does not satisfy a predetermined condition, is a control value of the imaging device and is an S / N ratio (Signal of the image captured by the imaging device) / Noise ratio) is changed using a CPU, and an S / N ratio control value changing unit that causes the imaging device to take a new image, and the image acquisition unit includes: An image newly captured by the imaging device is acquired.

According to the present invention, when the quality of feature information extracted from an image is used as an evaluation value and the quality value is not sufficient for authentication, a control element that contributes to the S / N ratio (S / N ratio control). Value) can be increased to increase the S / N ratio of the image.
As a result, for example, it is possible to improve the authentication accuracy without unnecessarily deteriorating comfort for the majority of users and minimizing the decrease in comfort even when the state of the living body is not good. be able to.

  As other control elements, there are an analog amplifier control value for electrically amplifying the signal, a multiple value for multiplying the luminance value of the input image by software, etc., but these do not impair the comfort of the user and the S / N ratio. Therefore, it is not included in the “control element contributing to the S / N ratio”.

1 is a configuration diagram of a biometric authentication device 900 according to Embodiment 1. FIG. 6 is a flowchart showing a biometric authentication method of biometric authentication apparatus 900 according to Embodiment 1. 5 is a flowchart showing a total gain determination process (S110) in the first embodiment. The graph which shows the relationship between Gl _def [level], S / N ratio, and comfort in Embodiment 1. The graph which shows the relationship between _Gtdef [level], S / N ratio, and comfort in Embodiment 1. FIG. 4 is a graph showing a relationship between Gn _def [level], S / N ratio, and comfort in the first embodiment. 9 is a flowchart showing a biometric authentication method of biometric authentication apparatus 900 according to Embodiment 2. 9 is a flowchart showing authentication retry processing (S180) in the second embodiment. 10 is another example of a flowchart showing a biometric authentication method of biometric authentication apparatus 900 in the second embodiment. 10 is another example of a flowchart showing a biometric authentication method of biometric authentication apparatus 900 in the second embodiment. FIG. 10 is a configuration diagram of a biometric authentication device 900 according to Embodiment 3. 10 is a flowchart showing a biometric authentication method of biometric authentication apparatus 900 according to Embodiment 3. 10 is a flowchart showing gain element ratio determination processing (S190) in the third embodiment.

Embodiment 1 FIG.
Take a biological image with the imaging device, evaluate the quality of the feature information obtained from the captured biological image, and if the quality of the feature information is poor, adjust the imaging device to recapture the biological image, A biometric authentication apparatus that performs biometric authentication based on high-quality feature information obtained from the biometric image thus obtained will be described.

FIG. 1 is a configuration diagram of a biometric authentication device 900 according to the first embodiment.
The configuration of biometric authentication apparatus 900 according to Embodiment 1 will be described below with reference to FIG.

  The biometric authentication device 900 includes an imaging device 100, an authentication processing device 200, and a user I / F device 902 (I / F: interface).

  The imaging device 100 includes a light source device 110 and an image generation device 120 and images a user's finger 901.

The light source device 110 includes a first light source 111, a second light source 112, and a light source control unit 113, and irradiates light on a user's finger 901 inserted into the light source device 110.
The 1st light source 111 is provided in the upper part of the light source device 110, and irradiates light to the finger | toe 901 from the upper part. The first light source 111 is transmissive illumination that transmits light to the finger 901.
The 2nd light source 112 is provided in the lower part of the light source device 110, and irradiates light to the finger | toe 901 from the lower part. The second light source 112 is reflective illumination that reflects light to the finger 901.
The light source control unit 113 controls the amount of light emitted from the first light source 111 and the amount of light emitted from the second light source 112 using a CPU (Central Processing Unit). The light quantity Gl is defined in the memory as a control value.

The image generation device 120 includes an imaging element 121, a signal processing circuit 122, an image generation unit 123, and an imaging control unit 124.
The image sensor 121 senses the light emitted from the first light source 111 and transmitted through the finger 901 and the light emitted from the second light source 112 and reflected on the finger 901, and converts the sensed light into electric charges (electrons). The converted charge is accumulated, and the accumulated charge is output as an analog image signal (electric signal representing an image).
The signal processing circuit 122 converts the analog image signal output from the image sensor 121 into a digital image signal. The signal processing circuit 122 includes an operational amplifier and the like.
The image generation unit 123 inputs a plurality of image signals (digital) converted from the image signals (analog) output a plurality of times by the image sensor 121 from the signal processing circuit 122 and superimposes the input plurality of image signals. One image (digital data) is generated using the CPU. The image generated by the image generation unit 123 represents the fingerprint of the user's finger 901.
The imaging control unit 124 adjusts the accumulation time Gt and the aperture amount Gs of the image sensor 121, the analog gain Ga and the digital gain Gd of the signal processing circuit 122, the image addition number Gn and the software gain Gs1 of the image generation unit 123.
The accumulation time Gt is an exposure time during which light from the image sensor 121 is incident, and the aperture amount Gs is an amount for adjusting the amount of light incident on the image sensor 121. The analog gain Ga is an amplification value of the analog image signal, and the digital gain Gd is an amplification value of the digital image signal. The image addition number Gn is the number of image signals to be superimposed to generate one image, and the software gain Gs1 is a value for final adjustment of the image.
The accumulation time Gt, the aperture amount Gs, the analog gain Ga, the digital gain Gd, the image addition number Gn, and the software gain Gs1 are defined in the memory as control values.

In the imaging apparatus 100, the first light source 111 and the second light source 112 emit light, the imaging element 121 outputs an analog image signal, and the signal processing circuit 122 converts the analog image signal into a digital image signal. The image generation unit 123 captures an image by generating an image from a digital image signal.
When the image capturing apparatus 100 captures an image, the image capturing apparatus 100 controls the first light source 111 and the second light source 112 based on the light amount Gl defined by the light source control unit 113, and the accumulation time in which the image capture control unit 124 is defined. Based on Gt, aperture amount Gs, analog gain Ga, digital gain Gd, image addition number Gn, and software gain Gs1, the image sensor 121, the signal processing circuit 122, and the image generator 123 are controlled.
The authentication processing device 200 changes the light amount Gl, the accumulation time Gt, the aperture amount Gs, the analog gain Ga, the digital gain Gd, the image addition number Gn, and the software gain Gs1.

  The authentication processing device 200 (an example of an authentication device) includes a biometric image acquisition unit 210 (an example of an image acquisition unit), a feature information extraction unit 220, an extracted information quality evaluation unit 230 (an example of an image determination unit), and a personal authentication unit 240 ( An example of an authentication processing unit), a gain element ratio determination unit 250 (an example of an S / N ratio control value changing unit), an overall gain determination unit 260 (an example of an overall control value determination unit), and a registered data storage unit 290 are provided.

  The biological image acquisition unit 210 acquires an image (hereinafter referred to as “biological image 101”) obtained by copying the user's finger 901 from the imaging apparatus 100.

The feature information extraction unit 220 extracts specific feature information 102 included in the biological image 101 from the biological image 101 acquired by the biological image acquisition unit 210 using the CPU.
For example, the feature information 102 is a ridge and valley component (branch point, end point, contrast, pattern) on the finger 901.

  Based on the feature information 102 extracted by the feature information extraction unit 220, the extracted information quality evaluation unit 230 determines whether the biological image 101 satisfies a predetermined condition using the CPU. The predetermined condition is that the feature information 102 satisfies a predetermined quality level necessary for the authentication process. For example, the quality level depends on the amount of frequency components of the ridge obtained by expressing the luminance (color component) of the image in terms of frequency, the contrast between the ridges and valleys, the number of branch points and end points of the ridges and valleys, etc. Identified based on.

When the extracted information quality evaluation unit 230 determines that the biometric image 101 satisfies a predetermined condition, the personal authentication unit 240 performs an authentication process using the CPU based on the feature information 102.
In the authentication process, the personal authentication unit 240 collates the feature information 102 with the registration data 291 registered in advance in the registration data storage unit 290. As a result of the collation, if there is registration data 291 that matches the feature information 102, “ “Authentication OK (authentication permitted)” is determined, and if there is no registration data 291 that matches the feature information 102, it is determined “authentication NG (authentication not permitted)”.
“Authentication OK” means that the user is a registered regular user, and “Authentication NG” means that the user is a non-regular user not registered.

  The gain element ratio determination unit 250 improves the S / N ratio (Signal / Noise ratio) of the biological image 101 when the extracted information quality evaluation unit 230 determines that the biological image 101 does not satisfy the predetermined condition. In addition, the control values (light quantity Gl, accumulation time Gt, aperture amount Gs, analog gain Ga, digital gain Gd, number of added images Gn, software gain Gs1, etc.) of the imaging apparatus 100 are changed using the CPU, The image 101 is newly imaged.

The total gain determination unit 260 determines a specific value as a total gain G _all using the CPU based on the pixel level of the biological image 101 acquired in advance by the biological image acquisition unit 210.
The overall gain is a value corresponding to the brightness of the biological image 101. If the overall gain G _all is the same, the brightness of the biological image 101 does not change even if the control value of the imaging device 100 is changed. The overall gain determination unit 260 determines the overall gain G _all so that the luminance distribution of the biological image 101 is in an optimal state that is not insufficient or saturated. The optimal state is a state in which the dynamic range of the luminance distribution of the biological image 101 is expanded to the maximum value range from which the feature information 102 can be extracted, and the image pattern of the biological image 101 is not crushed due to luminance saturation. .
Each control value of the imaging apparatus 100 is determined so that the product of the control values is the same value as the total gain G _all .

  The registration data storage unit 290 is a storage unit in which feature information extracted from a biological image of a specific user (regular user) captured in advance is stored in advance as registration data 291, and the registration data 291 is stored as a storage medium. It is a storage device that uses and stores.

The user I / F device 902 is a device that notifies a user of a predetermined message.
For example, the user I / F device 902 includes a display device, and displays a message notifying the result of the authentication process and the retry of the authentication process on the display device.

  The imaging device 100 and the authentication processing device 200 include a CPU, a ROM, a RAM, and a communication board (not shown). The CPU controls these hardware devices via the bus.

  The communication board is connected to a communication network such as a LAN (Local Area Network), the Internet, or a telephone line by wire or wireless.

The ROM or RAM stores an OS (Operating System), programs, and various data.
The program is executed by the CPU, and causes the imaging device 100 and the authentication processing device 200 (an example of a computer) to execute a function described as “˜unit” in the embodiment.
Each control value, biological image 101, feature information 102, authentication result, and registration data 291 of the imaging apparatus 100 are examples of various data.

  The “˜unit” is configured by firmware, software, hardware, or a combination thereof.

FIG. 2 is a flowchart showing a biometric authentication method of biometric authentication apparatus 900 according to the first embodiment.
A biometric authentication method of biometric authentication apparatus 900 according to Embodiment 1 will be described below with reference to FIG.
Each “unit” executes each process described below using a CPU.

The total gain determination unit 260 determines a total gain (S110), and the gain element ratio determination unit 250 determines a gain element ratio (control value of the imaging device 100) based on the total gain (S120).
The imaging device 100 captures the biological image 101 based on the control value corresponding to the gain element ratio (S130), and the biological image acquisition unit 210 acquires the biological image 101 from the imaging device 100 (S131).
The feature information extraction unit 220 extracts the feature information 102 from the biological image 101 (S140), and the extracted information quality evaluation unit 230 evaluates the quality of the feature information 102 (S150).
When the characteristic information 102 is “quality NG” (S151) and the gain element ratio can be changed (S152), the gain element ratio determination unit 250 changes the gain element ratio (S160). The process returns to S130.
When the feature information 102 is “quality OK” (S151), the personal authentication unit 240 executes an authentication process based on the feature information 102, and notifies the user of the authentication result using the user I / F device 902 ( S170).
Details of each process (S110 to S170) will be described below.

<S110: Comprehensive gain determination process>
The overall gain determination unit 260 determines an overall gain that determines the brightness of the biological image 101.

FIG. 3 is a flowchart showing the overall gain determination process (S110) in the first embodiment.
The overall gain determination process (S110) in the first embodiment will be described below based on FIG.

The imaging device 100 captures the biological image 101 for determining the overall gain (S111), the biological image acquisition unit 210 acquires the biological image 101 from the imaging device 100 (S112), and the overall gain determination unit 260 determines the maximum of the biological image 101. The pixel level is specified (S113), and the overall gain determination unit 260 determines the overall gain based on the maximum pixel level of the biological image 101 (S114).
Below, the detail of each process (S111-S114) is demonstrated.

<S111>
The imaging device 100 captures the biological image 101 for determining the total gain. At this time, a predetermined control value (hereinafter referred to as “total gain determination control value”) is used as the control value of the imaging apparatus 100.
The overall gain determination control value is a low value that does not include saturated pixels in the biological image 101.
Since the total gain can be determined even when the resolution of the biological image 101 is not high, the imaging apparatus 100 may capture the biological image 101 for determining the total gain in the high-speed imaging mode. The high-speed imaging mode means reducing the transfer time and exposure time of image data to the CPU by specific operations such as thinning the scan line of the image, thinning pixel values, transferring partial images to the CPU, and reducing the exposure time. In this mode, high-speed imaging is performed. The high-speed imaging mode is a function that is generally provided in an imaging apparatus whose imaging element is a CMOS type or a CCD type. An image captured in the high-speed imaging mode has a low image quality and is called a draft image.
After S111, the process proceeds to S112.

<S112>
The biological image acquisition unit 210 acquires the biological image 101 captured in S111 from the imaging device 100.
After S112, the process proceeds to S113.

<S113>
The overall gain determination unit 260 scans the biological image 101 acquired in S112 and identifies the maximum pixel level of the biological image 101. The pixel level is a value indicating the luminance (color component value) of the pixel (for example, “0 to 255” when monochrome is represented by 8 bits).
For example, the total gain determination unit 260 specifies the largest pixel level among all the pixels of the biological image 101 as the maximum pixel level.
Further, for example, if the dynamic range can be expanded without allowing the biometric image 101 to be blurred by allowing a few saturated pixels, the total gain determination unit 260 can increase the upper N percentile value ( N is preferably “1” or less) as the maximum pixel level.
The upper N percentile value is the value of the next largest data after the upper N% data when data (here, the pixel level of each pixel) are arranged in order of value. That is, the number of data having a value larger than the upper N percentile value corresponds to N percent of the total number of data. For example, the upper first percentile value of the 100-pixel biological image 101 is the second largest pixel level value.
After S113, the process proceeds to S114.

<S114>
The total gain determination unit 260 calculates the following Expression 1 using the maximum pixel level I _peak and determines the total gain G _all .

G _all = (I _max / I _peak ) × G ₀ <Formula 1>

In Equation 1, “I _max ” is the maximum value of the pixel level (“255” when the pixel level is expressed in 8 bits), and “G ₀ ” is the control value for determining the total gain used in S111. It is the product of each control value included (light quantity, light quantity, aperture amount, analog gain, digital gain, number of added images, etc.).
After S114, the overall gain determination process (S110) ends.

In FIG. 2, the description of the biometric authentication method is continued.
After S110, the process proceeds to S120.

<S120>
The gain element ratio determination unit 250 determines the gain element ratio based on the total gain G _all .
The gain element is each control value of the imaging apparatus 100, and the gain element ratio is the ratio of each control value of the imaging apparatus 100 to the total gain G _all (an example of the total control value). The gain element ratio determination unit 250 determines each control value of the imaging device 100 so that a value obtained by multiplying the control values of the imaging device 100 becomes the total gain G _all . Determination of the gain element ratio means determination of each control value of the imaging apparatus 100.

  The control values of the imaging device 100 determined in S120 are the light amount Gl of the light source device 110, the accumulation time Gt of the imaging device 121, the aperture amount Gs of the imaging device 121, the analog gain Ga of the signal processing circuit 122, and the signal processing circuit 122. The digital gain Gd, the image addition number Gn of the image generation unit 123, and the software gain Gs1 of the image generation unit 123.

The S / N ratio of the biological image 101 increases (improves / improves) as the values of the light amount Gl, the accumulation time Gt, the aperture amount Gs, and the number of added images Gn (each example of S / N ratio control value) are larger.
As the values of the analog gain Ga, the digital gain Gd, and the software gain Gs1 (each example of a non-S / N ratio control value) are larger, the S / N ratio of the biological image 101 becomes lower (deteriorates). The analog gain Ga is an analog amplification value at the electric signal level, and the digital gain Gd and the software gain Gs1 are amplification coefficients after being discretized into digital quantities. The digital gain Gd and the software gain Gs1 have such a low S / N ratio that a large value is required because the dynamic range of the image before amplification is narrow due to discretization rounding error noise.

  The gain element ratio determination unit 250 calculates the following Expressions 2 to 8, and determines each control value of the imaging apparatus 100 in the order of use at the time of imaging.

_{Gl = Max (Gl min, Min} (G all, Min (Gl set, Gl max))) < Formula 2>
Gt = Max (Gt _min , Min (G _all / G1, Min (Gt _set , Gt _max ))) <Expression 3>
Gs = Max (Gs _min , Min (G _all / (Gl * Gt), Gs _max )) <Expression 4>
_{Ga = Max (Ga min, Min} (G all / (Gl * Gt * Gs), Ga max)) < Equation 5>
_{Gd = Max (Gd min, Min} (G all / (Gl * Gt * Gs * Ga), Gd max)) < Equation 6>
_{Gn = Max (Gn min, Min} (G all / (Gl * Gt * Gs * Ga * Gd), Min (Gn set, Gn max))) < Equation 7>
_{Gs1 = Max (Gs1 min, Min} (G all / (Gl * Gt * Gs * Ga * Gd * Gn), Gs1 max)) < Equation 8>

In Expressions 2 to 8, Max (a, b) means selecting the larger value of “a” and “b”, and Min (a, b) is “a” and “b”. It means that the smaller value is selected.
“Gx _min ” and “Gx _max ” (Gx: Gl, Gt, GS, Ga, Gd, Gn, Gs1) are preliminarily set as values having no problem in terms of hardware of the biometric authentication device 900 and image quality of the biometric image 101. the minimum and maximum values of a defined G _x.
“Gx _set ” (Gx: Gl, Gt, Gn) is a value of Gx at a level selected from a plurality of levels predetermined as the level of Gx. For example, the value of each level of Gx is defined by a function table in which the level and the value are associated (hereinafter referred to as “Gx _def [level]”).

FIG. 4 is a graph showing the relationship between Gl _def [level], S / N ratio, and comfort in the first embodiment.
FIG. 5 is a graph showing the relationship among Gt _def [level], S / N ratio, and comfort in the first embodiment.
FIG. 6 is a graph showing the relationship between Gn _def [level], S / N ratio, and comfort in the first embodiment.

As shown in FIGS. 4 to 6, Gl _def [level], Gt _def [level], and Gn _def [level] increase in value and the S / N ratio of the biological image 101 as the level increases. However, the user's comfort (or operability) such as glare and extended rest time is reduced.

  When the number of added images is “N”, there is an effect of improving the S / N ratio by √N times (when the main cause of noise is due to the analog gain Ga). In the image sensor 121, since the accumulated charge of the pixel is saturated when the accumulation time Gt exceeds a certain length, the accumulation time Gt cannot be extended particularly when a bright image is taken. Qualitatively, it is similar to the increase in the accumulation time Gt in the image sensor 121. However, if the bit depth of the image addition buffer is sufficiently kept, ideally, the accumulation time Gt can be obtained by several tens of times. The S / N ratio can be improved.

In each function table, the maximum level value is set such that the user's comfort does not fall below a predetermined allowable value.
Further, each function table can extract feature information 102 as a minimum level (level 1) value sufficient to perform authentication processing in the majority of users and usage environments having the S / N ratio of the biological image 101. A value is set so that an S / N ratio of a certain degree is obtained.

In S120 of FIG. 2, the gain element ratio determination unit 250 uses the level 1 value of Gx _def as Gx _set as shown in the following formulas 9 to 11.

Gl _set = Gl _def [level (1)] <Expression 9>
Gt _set = Gt _def [level (1)] <Expression 10>
Gn _set = Gn _def [level (1)] <Formula 11>

The above is an example of a method for determining the gain factor ratio.
For example, the gain element ratio determination unit 250 may determine the gain elements (G1, Gt, Gs, Gn) that improve the S / N ratio first.

The gain element ratio determination unit 250 outputs the determined control value (gain element) to the imaging device 100, and the imaging device 100 sets the control value output by the gain element ratio determination unit 250 in the memory as a control value for authentication processing. To do.
After S120, the process proceeds to S130.

<S130>
The imaging apparatus 100 captures the biometric image 101 for authentication processing based on the control value for authentication processing set in S120.
After S130, the process proceeds to S131.

<S131>
The biological image acquisition unit 210 acquires the biological image 101 captured in S130 from the imaging device 100.
After S131, the process proceeds to S140.

<S140>
The feature information extraction unit 220 extracts the feature information 102 from the biological image 101 acquired in S131.
For example, the feature information extraction unit 220 extracts the feature information 102 from the biological image 101 by the technique disclosed in Patent Document 3.
After S140, the process proceeds to S150.

<S150>
The extracted information quality evaluation unit 230 performs quality evaluation (numerical evaluation) on the feature information 102 extracted in S140, and the quality of the feature information 102 is quality (quality OK) with which a correct authentication result can be obtained in the authentication process (S170). Or not (quality NG).

  For example, the extracted information quality evaluation unit 230 artificially calculates the S / N ratio as follows. When the luminance (color component) of an image is represented by frequency, the noise component appears in a high frequency band, and the signal component (component other than the noise component) appears in a low frequency band. The extracted information quality evaluation unit 230 performs frequency analysis on the biological image 101 and calculates the ratio between the amount of the high frequency band component corresponding to the noise component and the amount of the frequency band component corresponding to the signal component. The extracted information quality evaluation unit 230 determines that the quality of the feature information 102 is “quality OK” when the ratio of the signal components satisfies a predetermined value, and the quality of the feature information 102 when the ratio of the signal components does not satisfy the predetermined value. Is determined as “quality NG”. For example, when the extraction target of the feature information 102 is a fingerprint, the luminance of the pixel representing the ridge is a signal component.

  Further, for example, the extracted information quality evaluation unit 230 calculates a region where the contrast between the ridge portion and the valley portion of the fingerprint is less than a predetermined value because the biological image 101 is unclear based on the feature information 102. The extracted information quality evaluation unit 230 determines that the quality of the feature information 102 is “quality OK” when the calculated area ratio of the area is larger than the predetermined ratio, and when the calculated area ratio of the area is equal to or lower than the predetermined ratio. The quality of the information 102 is determined as “quality NG”.

  Further, for example, the extracted information quality evaluation unit 230, when the number of feature points (branches and end points of ridges and valleys) included in the feature information 102 satisfies a minimum value required for the authentication process, the feature information When the quality of the feature information 102 is determined as “quality OK” and the number of feature points included in the feature information 102 is less than the minimum value necessary for the authentication process, the quality of the feature information 102 is determined as “quality NG”. . For example, the extracted information quality evaluation unit 230 calculates the probability that the feature information 102 coincides with the other person's registered data 291 based on the number of feature points included in the feature information 102. Then, if the calculated probability is less than the predetermined value, the extracted information quality evaluation unit 230 assumes that the number of feature points included in the feature information 102 satisfies the minimum value necessary for the authentication process, and the quality of the feature information 102 Is determined to be “quality OK”. In addition, when the calculated probability is equal to or greater than a predetermined value, the extracted information quality evaluation unit 230 determines that the number of feature points included in the feature information 102 is less than the minimum value necessary for the authentication process, and The quality is determined as “quality NG”. As a method for calculating the probability that the feature information 102 coincides with the registered data 291 of another person, Non-Patent Document 1 can be applied.

  The extracted information quality evaluation unit 230 may evaluate the quality of the intermediate processing result of the feature information extraction process in S140 instead of the feature information 102.

  After S150, the process proceeds to S151.

<S151>
If it is determined that “quality is OK” in S150, the process proceeds to S170.
When it is determined as “quality NG” in S150, the process proceeds to S152.

<S152>
The gain element ratio determination unit 250 determines whether the gain element ratio can be changed.
For example, in S120, the gain element ratio determination unit 250 defines Gx _def (Gx: Gl, Gt, Gn) as a value used as the light amount Gl _set , the accumulation time Gt _set, or the image addition number Gn _set . When the value is not the maximum level, it is determined that the gain element ratio can be changed.
Further, for example, the gain element ratio determination unit 250 determines that the gain element ratio cannot be changed when the gain element ratio has already been changed a predetermined number of times or more in S160.
Further, for example, the gain element ratio determination unit 250 determines that the gain element ratio can be changed when the elapsed time since the finger 901 is placed on the imaging apparatus 100 does not exceed a predetermined time.
If the gain element ratio can be changed (YES), the process proceeds to S160.
When the gain factor ratio cannot be changed (NO), the personal authentication unit 240 notifies the user that the authentication has failed (authentication NG) using the user I / F device 902, and the process ends.

<S160>
When the feature information 102 is “quality NG” (NO) in S151 and the gain element ratio can be changed (YES) in S152, the gain element ratio determination unit 250 causes the imaging apparatus 100 to re-image the biological image 101. The gain element for authentication processing (control value of the imaging apparatus 100) is redetermined so as to improve the S / N ratio.
The gain element ratio determination unit 250 outputs the re-determined gain element for authentication processing to the imaging device 100, and the imaging device 100 sets the gain element for authentication processing output from the gain element ratio determination unit 250 in the memory.
In S160, the light amount of the light source device 110, the accumulation time of the image sensor 121, the aperture amount of the image sensor 121, the analog gain of the signal processing circuit 122, the digital gain of the signal processing circuit 122, the number of images added by the image generation unit 123, and the image generation The software gain of the unit 123 is changed.

For example, the gain element ratio determination unit 250 re-determines a gain element for authentication processing in the same manner as in S120. At this time, the gain element ratio determining unit 250 uses Gx _def (Gx on the d level) as Gl _set for the light amount, Gt _{set for} the accumulation time, or Gn _set for the number of added images, as shown in Expressions 12 to 15 below. : Gl, Gt, Gn) values are used. The d level “d” may be a fixed value of one or two or more, considering the quality evaluation result of the feature information 102 in S150 as a binary logic of “quality OK” and “quality NG”. In addition, “d” may be determined based on the amount by which the feature information 102 falls below a predetermined reference value (such as the ratio of signal components) in S150. “D” becomes a larger value as the amount by which the feature information 102 falls below a predetermined reference value is larger.

level ← level + d <expression 12>
Gl _set = Gl _def [level] <Expression 13>
Gt _set = Gt _def [level] <Expression 14>
Gn _set = Gn _def [level] <Expression 15>

In S160, when the levels of the light amount Gl _def , the accumulation time Gt _def and the image addition number Gn _def increase, as shown in FIGS. 4 to 6, each gain element (light amount Gl, accumulation time Gt, image addition) The S / N ratio corresponding to the number of sheets Gn) increases. The S / N ratio of the biological image 101 is proportional to the product of the S / N ratios corresponding to the respective gain elements (sum in the case of decibel expression). For this reason, the S / N ratio of the biological image 101 increases as the S / N ratio corresponding to each gain element increases.
At the next imaging by the imaging apparatus 100, the overall gain G _all does not change, so that the brightness of the biological image 101 maintains the previous brightness, but the S / N ratio of the biological image 101 is higher than the previous time.

  After S160, the process returns to S130, and the biological image 101 is re-imaged by the imaging device 100 based on the changed control values (gain elements).

<S170>
If the feature information 102 is “quality OK” (YES) in S151, the personal authentication unit 240 performs an authentication process based on the feature information 102.
In the authentication processing, the personal authentication unit 240 compares the feature information 102 with the registration data 291 stored in the registration data storage unit 290, and calculates the similarity (authentication score) between the feature information 102 and the registration data 291. To do. The personal authentication unit 240 determines “authentication OK” if the similarity between the feature information 102 and the registration data 291 is equal to or higher than a predetermined level, and determines that “similarity” between the feature information 102 and the registration data 291 is less than the predetermined level. “Authentication NG”.
The personal authentication unit 240 notifies the user of the authentication result using the user I / F device 902. Further, when the authentication result is “authentication OK”, the personal authentication unit 240 notifies a predetermined device (for example, an electronic lock device) that “authentication is OK”. For example, when receiving an “authentication OK” notification from the personal authentication unit 240, the electronic lock device unlocks the locked door and allows the user to enter the building (or room) from the door. To do.
After S170, the process ends.

  As described above, the biometric authentication device 900 increases the S / N ratio of the biometric image 101 at the next imaging when the quality evaluation value of the feature information 102 is not at a level at which the authentication process is successful (in the case of quality NG). Each control amount of the imaging apparatus 100 is changed. As a result, the biometric authentication device 900 can perform re-imaging even when a biometric pattern (ridge or valley) is thin and the feature point is buried in noise and the feature point cannot be sufficiently detected. The detection accuracy of the feature information 102 is gradually increased, and the accuracy of the authentication process can be increased. In addition, the biometric authentication device 900 performs authentication processing without re-imaging when many biometrics whose biometric patterns are not thin are targeted (when the first feature information 102 is “quality OK”). For this reason, the biometric authentication device 900 has an advantage that many standard users are not inconvenienced and the number of unauthenticated users can be reduced.

The imaging apparatus 100 is an example of an apparatus that captures the biological image 101.
Although the optical imaging device 100 has been described above, the imaging device 100 may be a device that captures the biological image 101 by other imaging principles such as a capacitance type, a radio wave type, and an ultrasonic type. The device that drives a device that applies a voltage in the capacitance type corresponds to the light source device 110, the device that drives the radio wave transmission antenna in the radio wave type, and the device that drives the ultrasonic wave transmitter in the ultrasonic type.
The light is an example of physical energy that is irradiated or injected into a living body (for example, a finger 901) that is a subject. For example, as physical energy, in addition to light, a radio wave in a radio wave imaging device, a voltage or current in a capacitive imaging device, and an ultrasonic wave in an ultrasonic imaging device can be given.
The amount of light is an example of a control value that controls the amount of physical energy irradiated or injected into the living body.
The accumulation time, the aperture amount, and the number of added images are examples of control values for controlling the sensitivity of the image sensor 121 or the physical energy accumulation amount.

  The biometric authentication method described above is a controllable hardware element or software element that can affect the S / N ratio of the biometric image 101 (for example, the light amount of the light source device 110 described above, the accumulation time of the image sensor 121, the image sensor 121). Can be applied to the biometric authentication device 900 including one or more of the aperture amount and the number of images added by the image generation unit 123.

  The image sensor 121, the signal processing circuit 122, and the image generation unit 123 may be collectively referred to as an image sensor (for example, a CMOS image sensor).

  The above authentication process performed based on the fingerprint feature information 102 is an example of biometric authentication, and the authentication process may be performed based on the feature information 102 such as a vein, an iris, and a face.

The authentication processing method in S170 may be “1: N authentication” or “1: 1 authentication”.
“1: N authentication” is a method in which feature information 102 is sequentially checked against all registered data 291 until registered data 291 that matches feature information 102 is found.
In “1: 1 authentication”, an ID for identifying the user in association with the registration data 291 is stored as a registration ID, and the user is allowed to input the ID, and a registration ID that matches the input ID is searched. The feature information 102 is collated only with the registration data 291 corresponding to the registration ID.
When the authentication processing method is “1: 1 authentication”, a level suitable for the user is stored in association with the registration ID, and in S120, the gain element ratio determination unit 250 sets the level corresponding to the registration ID. The Gx _set (Gx: Gl, Gt, Gn) may be determined by using this method. As a result, the number of times that the feature information 102 becomes quality NG in S151 can be reduced, and the repetition of imaging (S160 → S130) can be reduced.

  The biometric authentication device 900 may be either a contact type in which the finger 901 is in contact with the imaging apparatus 100 or a non-contact type in which the finger 901 is not in contact with the imaging apparatus 100.

In the first embodiment, the following biometric authentication device 900 has been described.
The biometric authentication device 900 includes a total gain determination unit that determines a total gain suitable for extraction of the feature information 102, an image acquisition unit (imaging device 100) that captures a biometric image using the determined total gain, and an image. Feature information extraction means for extracting the feature information 102 from the information, and personal authentication means for evaluating the degree of similarity with the registration data 291 using the feature information 102 and performing personal authentication.
The image acquisition unit includes at least one of an imaging energy irradiation unit (light source device 110) or an image storage unit (image generation device 120). The imaging energy irradiating means is a means capable of controlling the input energy amount when irradiating or injecting at least one of physical energy of light, radio wave, voltage, current, and ultrasonic wave to a living body. The image accumulating means is a means capable of controlling the temporal or spatial or electrical or software logical accumulation amount of the physical energy that is emitted from the living body by reflection, refraction, and transmission and enters the image sensor 121. It is.
Furthermore, the biometric authentication device 900 includes an extracted information quality evaluation unit that evaluates the quality of the feature information 102 extracted by the feature information extraction unit.
Further, the biometric authentication device 900 can control among the control amount of the physical energy and the accumulation control amount in the required total gain in association with the reduction level of the quality evaluation value by the extracted information quality evaluation means. And a gain element ratio determining means for changing the ratio of the control amount to increase the amount of accumulated physical energy.
The biometric authentication device 900 performs the above-determined ratio at the time of the next imaging or the first imaging at the next authentication trial among a plurality of imaging operations that can be performed continuously during one authentication trial by the device user. The imaging energy irradiating means or the image accumulating means is controlled using the control amount, and an image with an improved S / N ratio is taken.

The total gain determining means determines the total gain using an image taken in the draft mode with a low resolution but a short image acquisition time.
In the biometric authentication device 900, the total gain determination unit determines the total gain using a draft captured image having the same luminance distribution as the image input to the feature information extraction unit. As a result, the biometric authentication device 900 is not required to have another environment measurement sensor such as a thermometer or a hygrometer that measures the surrounding environment of the living body, and is affected by a difference between the actual imaging state of the surrounding environment and the living body. The overall gain can be determined without any problem.

  When the captured image does not satisfy the predetermined condition, the control value of the imaging device is changed so as to improve the S / N ratio, and the image acquisition device that causes the imaging device to acquire a new image (in Embodiment 1) The authentication processing device 200) is an effective device even when used in other image processing systems other than the biometric authentication device 900.

  A description of correction of nonlinear elements such as the clamp level value and γ characteristic of the image sensor 121 is omitted.

Embodiment 2. FIG.
Also in the case of authentication NG, the form which changes the control value of the imaging device 100, re-images the biological image 101, and performs an authentication process based on the re-imaged biological image 101 is demonstrated.
Hereinafter, items different from the first embodiment will be mainly described. Matters whose description is omitted are the same as those in the first embodiment.

FIG. 7 is a flowchart showing a biometric authentication method of biometric authentication apparatus 900 according to the second embodiment.
A biometric authentication method of biometric authentication apparatus 900 according to Embodiment 2 will be described below with reference to FIG.

The flowchart shown in FIG. 7 is obtained by adding S171 and S180 to the biometric authentication method (see FIG. 2) described in the first embodiment.
Hereinafter, S171 and S180 will be described, and description of other processes (S110 to S170) will be omitted.

  After S170, the process proceeds to S171.

<S171>
If the result of the authentication process in S170 is “authentication OK” (YES), the process ends.
If the result of the authentication process in S170 is “authentication NG” (NO), the process proceeds to S180.

FIG. 8 is a flowchart showing the authentication retry process (S180) in the second embodiment.
The authentication retry process (S180) in the second embodiment will be described below with reference to FIG.

<S181>
The gain element ratio determination unit 250 determines whether the gain element ratio can be changed, as in S152.
If the gain element ratio can be changed (YES), the process proceeds to S182.
If the gain element ratio cannot be changed (NO), the process ends.

<S182>
The gain element ratio determination unit 250 changes the gain element ratio using the value of Gx _def (Gx: Gl, Gt, Gn) on the d level, similarly to S160.
“D” may be determined according to the degree of similarity between the biometric image 101 and the registration data 291 in the authentication process (S170 or S186 described later). “D” increases as the similarity between the biological image 101 and the registered data 291 decreases.
After S182, the process proceeds to S183.

<S183>
The personal authentication unit 240 notifies the user using the user I / F device 902 of a message requesting that the finger 901 be inserted into the imaging device 100 in order to retry the authentication process.
The imaging apparatus 100 re-images the biometric image 101 for authentication processing.
After S183, the process proceeds to S184.

<S184>
The biological image acquisition unit 210 acquires the biological image 101 re-imaged in S183 from the imaging device 100.
After S184, the process proceeds to S185.

<S185>
The feature information extraction unit 220 extracts the feature information 102 from the biological image 101 acquired in S184. The extraction method of the feature information 102 is the same as S140.
After S185, the process proceeds to S186.

<S186>
The personal authentication unit 240 performs an authentication process based on the feature information 102 extracted in S185. The authentication process is the same as S170.
After S186, the process proceeds to S187.

<S187>
If the result of the authentication process in S186 is “authentication OK” (YES), the process ends.
When the result of the authentication process in S186 is “authentication NG” (NO), the process returns to S181, and the authentication retry process is repeatedly executed.

The biometric authentication apparatus 900 evaluates the quality of the feature information 102 (S150), and executes the authentication process (S170) when the feature information 102 is “quality OK” (S151). However, even if the feature information 102 is evaluated as “quality OK”, the result of the authentication process is not always a correct result. That is, there is a possibility that the feature information 102 evaluated as “quality OK” is actually insufficient quality, and the authorized user is determined as “authentication NG”.
Therefore, the biometric authentication device 900 according to the second embodiment increases the accuracy of the authentication process by executing the authentication retry process (S180) in the case of “authentication NG”.
The biometric authentication method shown in FIGS. 7 and 8 is an example of the biometric authentication method of biometric authentication apparatus 900 in the second embodiment.

FIG. 9 and FIG. 10 are other examples of flowcharts showing the biometric authentication method of the biometric authentication apparatus 900 according to the second embodiment.
As shown in FIG. 9, the biometric authentication method of the biometric authentication apparatus 900 according to the second embodiment may be a method of executing the processing from S152 onward in the case of “authentication NG” in S171.
In the biometric authentication method of biometric authentication apparatus 900 according to the second embodiment, as shown in FIG. 10, S150 to S160 may not be executed, and the gain factor ratio may be changed only in the case of “authentication NG”. (S173).

In the second embodiment, the following biometric authentication device 900 has been described.
The biometric authentication device 900 includes a total gain determination unit that determines a total gain suitable for extraction of the feature information 102, an image acquisition unit (imaging device 100) that captures a biometric image using the determined total gain, and an image. Feature information extraction means for extracting the feature information 102 from the information, and personal authentication means for evaluating the degree of similarity with the registration data 291 using the feature information 102 and performing personal authentication.
The image acquisition unit includes at least one of an imaging energy irradiation unit (light source device 110) or an image storage unit (image generation device 120). The imaging energy irradiating means is a means capable of controlling the input energy amount when irradiating or injecting at least one of physical energy of light, radio wave, voltage, current, and ultrasonic wave to a living body. The image accumulating means is a means capable of controlling the temporal or spatial or electrical or software logical accumulation amount of the physical energy that is emitted from the living body by reflection, refraction, and transmission and enters the image sensor 121. It is.
Furthermore, the biometric authentication device 900 associates the control amount of the physical energy and the accumulation control in the required total gain in association with the level of decrease in the similarity with the registration data 291 evaluated by the personal authentication means. Gain factor ratio determining means is provided for changing the controllable control amount ratio in the direction of increasing the physical energy storage amount.
The biometric authentication device 900 performs the above-determined ratio at the time of the next imaging or the first imaging at the next authentication trial among a plurality of imaging operations that can be performed continuously during one authentication trial by the device user. The imaging energy irradiating means or the image accumulating means is controlled using the control amount, and an image with an improved S / N ratio is taken.

The gain element ratio determining unit uses the imaging energy irradiating unit and the image acquiring unit to capture an image with an increased S / N ratio at the next imaging using the similarity with the registration data 291 obtained from the personal authentication unit. Change the control amount.
The biometric authentication device 900 includes a gain element ratio determining unit, and controls a imaging energy irradiation unit or an image storage unit to target a living body in which a biometric pattern is thin and feature points are buried in noise and cannot be sufficiently detected. Even in this case, the detection accuracy of the feature information 102 is gradually increased each time re-imaging is performed, and the accuracy of the authentication process can be increased. In addition, the biometric authentication apparatus 900 performs authentication processing without re-imaging when a large number of biometrics whose biometric patterns are not thin are targeted (when the first feature information 102 is “quality OK”). For this reason, the biometric authentication apparatus 900 can reduce the number of unauthenticated users without causing inconvenience to many standard users.

Embodiment 3 FIG.
A mode in which the gain factor ratio is determined based on the total gain biometric image 101 before capturing the authentication biometric image 101 will be described.
Hereinafter, items different from the first embodiment will be mainly described. Matters whose description is omitted are the same as those in the first embodiment.

FIG. 11 is a configuration diagram of the biometric authentication device 900 according to the third embodiment.
The configuration of biometric authentication apparatus 900 according to Embodiment 3 will be described below based on FIG.

The biometric authentication device 900 includes a feature information simple extraction unit 221 and an extraction information simple quality evaluation unit 231 in addition to the configuration described in the first embodiment (see FIG. 1).
The feature information simple extraction unit 221 easily extracts feature information from the biological image 101 using a CPU.
The extracted information simple quality evaluation unit 231 performs quality evaluation of the feature information extracted by the feature information simple extraction unit 221 using the CPU.

FIG. 12 is a flowchart showing a biometric authentication method of biometric authentication apparatus 900 according to the third embodiment.
A biometric authentication method of biometric authentication apparatus 900 according to Embodiment 3 will be described below with reference to FIG.

The flowchart shown in FIG. 12 is obtained by changing S120 of the biometric authentication method (see FIG. 2) described in Embodiment 1 to S190.
Hereinafter, S190 will be described, and description of other processes (S110, S130 to S170) will be omitted.

  After S110, the process proceeds to S190.

FIG. 13 is a flowchart showing gain element ratio determination processing (S190) in the third embodiment.
The gain element ratio determination process (S190) in the third embodiment will be described below based on FIG.

<S191>
The feature information simple extraction unit 221 extracts feature information from the biological image 101 for total gain captured in the total gain determination processing (S110) by simple processing. Hereinafter, the feature information extracted in S191 is referred to as “simple feature information 103”.
For example, the feature information simple extraction unit 221 extracts the simple feature information 103 from the biological image 101 for total gain by the Solve filter process.
After S191, the process proceeds to S192.

<S192>
The extracted information simple quality evaluation unit 231 evaluates the quality of the simple feature information 103 extracted in S191.
At this time, the extracted information simple quality evaluation unit 231 simply evaluates the quality of the simple feature information 103, for example, by reducing the number of evaluation items compared to the quality evaluation of the extracted information quality evaluation unit 230 in S150. However, the extracted information simple quality evaluation unit 231 may evaluate the quality of the simple feature information 103 in the same manner as the extracted information quality evaluation unit 230 in S150.
After S192, the process proceeds to S193.

<S193>
The gain element ratio determination unit 250 determines the gain element ratio for authentication processing based on the amount that the simple feature information 103 falls below a predetermined reference value in S192, and changes the control value of the imaging apparatus 100. The method for determining the gain element ratio is the same as in S160.
After S193, the gain element ratio determination process (S190) ends.

  The biometric authentication device 900 determines the gain element ratio based on the biometric image 101 for total gain before capturing the biometric image 101 for authentication processing, thereby reducing the number of imaging and maintaining authentication accuracy. User operability and comfort can be improved.

  As in the second embodiment, the biometric authentication device 900 changes the control value of the imaging device 100 in the case of authentication NG, re-images the biometric image 101, and executes authentication processing based on the re-imaged biometric image 101. It doesn't matter.

In the third embodiment, the following biometric authentication device 900 has been described.
The biometric authentication device 900 includes second feature information extraction means (a feature information simple extraction unit 221) that extracts feature information (simple feature information 103) using a draft captured image that has been captured immediately before, and second feature information. 2nd extraction information quality evaluation means (extraction information simple quality evaluation part 231) which evaluates the quality of the feature information extracted by the extraction means is provided.
The gain element ratio determining means uses the second evaluation value by the second extracted information quality evaluation means at the time of the first imaging, and controls the ratio of controllable control amounts between the physical energy control amount and the accumulation control amount. To decide.
The biometric authentication device 900 includes a high-speed feature information extraction unit (feature information simple extraction unit 221) and an extraction information quality evaluation unit (extraction information simple quality evaluation unit 231) dedicated to changing the control amount for the imaging energy irradiation unit and the image storage unit. And a feature information extraction means (feature information extraction section 220) for personal authentication processing and an extraction information quality evaluation means (extraction information quality evaluation section 230). As a result, the biometric authentication device 900 can shorten the time required to determine the gain factor ratio, reduce the burden on the device user, and improve the authentication accuracy.

  DESCRIPTION OF SYMBOLS 100 Imaging device, 101 Living body image, 102 Feature information, 103 Simple feature information, 110 Light source device, 111 1st light source, 112 2nd light source, 113 Light source control part, 120 Image generation device, 121 Imaging device, 122 Signal processing circuit, 123 image generation unit 124 imaging control unit 200 authentication processing device 210 biological image acquisition unit 220 feature information extraction unit 221 feature information simple extraction unit 230 extraction information quality evaluation unit 231 extraction information simple quality evaluation unit 240 Personal authentication unit, 250 gain element ratio determination unit, 260 total gain determination unit, 290 registration data storage unit, 291 registration data, 900 biometric authentication device, 901 finger, 902 user I / F device, 911 CPU, 912 bus, 913 ROM , 914 RAM, 915 communication board, Gl light quantity, G Storage time, Gs aperture amount, Ga analog gain, Gd digital gain, Gn image addition number, Gs1 software gain.

Claims (10)

An image acquisition unit that acquires an image captured by the imaging device;
An image determination unit that determines whether an image acquired by the image acquisition unit satisfies a predetermined condition using a CPU (Central Processing Unit);
When the image determination unit determines that the image does not satisfy a predetermined condition, the control value of the imaging device and the S / N ratio (Signal / Noise ratio) of the image captured by the imaging device are changed. A S / N ratio control value changing unit that changes the S / N ratio control value to be performed using a CPU and causes the imaging device to newly capture an image,
The image acquisition device, wherein the image acquisition unit acquires an image newly captured by the imaging device. The S / N ratio control value changing unit updates the S / N ratio control value to a high value and sets a non-S / N ratio control value that is a control value of the imaging apparatus other than the S / N ratio control value. Update to a lower value,
The image acquisition unit includes the S / N ratio control value updated by the S / N ratio control value change unit and the non-S / N ratio control value updated by the S / N ratio control value change unit. The image acquisition device according to claim 1, wherein the image acquisition device is controlled based on the acquired image to acquire an image from the image pickup device. The S / N ratio control value changing unit matches the product of the S / N ratio control value and the non-S / N ratio control value before and after the update. Image acquisition device. The image acquisition device further includes:
A comprehensive control value determination unit that determines a specific value as a total control value using a CPU based on a pixel level of an image acquired in advance by the image acquisition unit;
The S / N ratio control value changing unit matches the product of the S / N ratio control value and the non-S / N ratio control value with the total control value determined by the total control value determining unit. The image acquisition apparatus according to claim 3. 5. The S / N ratio control value changing unit updates the S / N ratio control value based on a difference between the image and the predetermined condition. 6. Image acquisition device. The imaging control value determination unit determines at least one of a control value that determines the intensity of physical energy input to the imaging device of the imaging apparatus and a control value that determines the amount of physical energy to be accumulated in the imaging device as the S The image acquisition apparatus according to claim 1, wherein the S / N ratio control value is updated as an / N ratio control value. The image acquisition device according to any one of claims 1 to 6,
An authentication apparatus comprising: an authentication processing unit that performs an authentication process using a CPU based on the image when the image determination unit determines that the image satisfies a predetermined condition. The S / N ratio control value changing unit further includes:
8. The authentication apparatus according to claim 7, wherein when the authentication result by the authentication processing unit is authentication NG, the S / N ratio control value is updated. The authentication device further includes:
A feature information extraction unit that extracts specific feature information included in the image from the image using a CPU;
The authentication processing unit determines an authentication result based on the similarity between the feature information extracted by the feature information extraction unit and registered information registered in the storage device in advance,
The S / N ratio control value changing unit updates the S / N ratio control value based on the similarity between the feature information and the registration information when the authentication result by the authentication processing unit is authentication NG. The authentication apparatus according to claim 8, wherein: An image acquisition unit that acquires an image captured by the imaging device;
An authentication processing unit that performs an authentication process using a CPU based on the image acquired by the image acquisition unit;
When the result of the authentication process by the authentication processing unit is authentication NG, the S / N ratio that changes the S / N ratio (Signal / Noise ratio) of the image that is the control value of the imaging device and is captured by the imaging device A control value is changed using a CPU, and an S / N ratio control value changing unit that causes the imaging device to newly capture an image,
The authentication apparatus according to claim 1, wherein the image acquisition unit includes an image newly captured by the imaging apparatus.

Images