JP2018180660A - Biometric authentication apparatus and method - Google Patents

Abstract

Provided is a biometric authentication device capable of performing high-precision authentication even when the illumination environment and the imaging device are different between registration and authentication. A biometric authentication device includes an image input unit that captures a finger under ambient light to acquire at least one biometric image, an authentication processing unit that processes the biometric image, and a plurality of biometric features obtained from the finger. And a storage unit that stores registration information related to at least one of them. The authentication processing unit uses the registration information in the storage unit to generate a color conversion matrix that minimizes the color difference between the reference biometric image and the representative biometric image at the time of authentication, and uses the generated color conversion matrix to generate a color conversion matrix. A converted biometric image is generated (S111), feature amount extraction (S112) is performed, and similarity is calculated with registration information (S113). [Selection] Figure 2B

Description

  The present invention relates to a biometric authentication apparatus and method for authenticating an individual using a living body.

  Among various biometric authentication techniques, finger vein authentication is known as one that can realize highly accurate authentication. Finger vein authentication uses blood vessel patterns inside the finger to achieve excellent authentication accuracy. Since finger vein authentication is more difficult to forge and tamper as compared with fingerprint authentication, high security can be realized.

  In recent years, cases of performing biometrics by photographing a living body with a color camera mounted as a standard on portable terminals such as mobile phones, notebook PCs (Personal Computers), smartphones, tablets, etc. are increasing.

  Individual authentication can be performed using a general-purpose sensor under ambient light in a space where a living body exists, without using a specific light source or a dedicated sensor, so it is not necessary to choose a place such as a home or a place to go, an individual It is expected to be used on various devices such as smartphones and public tablets. Furthermore, realization of mutual biometric authentication between different imaging devices, such as registration with a smartphone and authentication with a tablet, is also expected.

  In Patent Document 1, a finger vein or the like is extracted from an image obtained by photographing a finger with an imaging device including a color camera having sensitivity to visible light and a light emitting diode (LED) light source emitting blue, green and red, and personal authentication It discloses the technology to be implemented. In Patent Document 2, the similarity between finger vein images acquired as a moving image at the time of authentication is calculated, and only finger vein images having high similarity with each other are used for authentication, and imaging of a finger's displacement or shadow of illumination etc. It discloses a technique for avoiding the fluctuation of conditions.

JP, 2016-096987, A JP, 2010-277315, A

  In order to perform high-accuracy biometric authentication using a color camera that is standard-installed on portable terminals such as mobile phones, laptop PCs, smartphones and tablets, it is necessary to change the lighting fluctuation at the time of shooting and the characteristic fluctuation of the imaging device. A robust feature extraction technique is needed.

  In Patent Document 1, the influence of external light is reduced by covering the entire angle of view of the camera with a finger or using a roof. Therefore, the apparatus configuration is premised on arranging the light source in accordance with the position of the camera. Therefore, the authentication with the portable terminal in which the positional relationship between the camera and the light source differs depending on the device or the light source does not exist is not considered. Moreover, the case where mutual authentication is performed with a plurality of devices having individual differences in cameras is not considered.

  According to Patent Document 2, it is possible to avoid fluctuations in shooting conditions during authentication such as finger misalignment and changes in the degree of shadow due to illumination, while shooting conditions such as ambient illumination and individual differences between imaging devices at the time of registration and authentication. Different cases are not considered.

  Therefore, an object of the present invention is to provide a biometric authentication device and method capable of performing authentication with high accuracy even when the illumination environment and imaging conditions of an imaging device are different between registration and authentication.

  In order to achieve the above object, in the present invention, an image input unit for capturing a living body to acquire a living body image, an authentication processing unit for processing a living body image, and registration information on biometric characteristics obtained from the living body are stored. A storage unit, the authentication processing unit performs color conversion to minimize a color difference between the living body image at the time of registration and at the time of authentication using the registration information, and authenticates the living body using the living body image after the color conversion To provide a biometric device to perform.

  Further, in order to achieve the above object, in the present invention, in the biometric authentication method by the authentication processing unit, the authentication processing unit processes a living body image obtained by imaging the living body, and obtains a living body characteristic obtained from the living body The present invention provides a biometric authentication method that performs color conversion to minimize color difference between biometric images at the time of registration and at the time of authentication using the registration information related to (4), and performs biometric authentication using the biometric image after color conversion.

  According to the present invention, by minimizing the color difference of each living body image obtained by photographing the living body at the time of registration and at the time of authentication, the influence of the illumination variation and the individual difference between the devices is reduced and the authentication is performed with high accuracy. be able to.

FIG. 2 is a diagram showing the entire biometric authentication system according to the first embodiment and the configuration of an authentication processing unit thereof. FIG. 6 is a diagram showing an example of a color conversion processing flow at the time of registration of a living body image of the first embodiment. FIG. 7 is a diagram showing an example of a color conversion processing flow at the time of authentication of a living body image of the first embodiment. FIG. 14 is a diagram showing a modification of the processing flow example at the time of registration of a living body image of the first embodiment. FIG. 7 is a diagram showing an example of a data set table according to the first embodiment. 5 is a schematic view showing an example of a living body image of Example 1. FIG. It is a schematic diagram which shows the structure which image | photographs a biological body using the color camera of a portable terminal based on Example 1, and implements personal identification. FIG. 8 is a diagram showing an example of color conversion processing of the biometric authentication system of the second embodiment. FIG. 13 is an explanatory diagram of a method of removing an ambient light component of a living body image of Example 2. FIG. 16 is a diagram showing an example of an authentication processing flow of the biometric authentication system of a third embodiment. It is a schematic diagram of the mechanism which acquires the camera position in the back of a housing based on Example 3. FIG. It is a schematic diagram which shows the structure which image | photographs a biological body using the color camera of the back side of a display screen based on Example 3, and implements personal identification.

  Various embodiments of the present invention will now be described with reference to the accompanying drawings. Although the attached drawings show specific examples according to the principle of the present invention, these are for understanding the present invention and are used for limiting interpretation of the present invention. is not. The same reference numerals are attached in principle to the common configuration in each drawing.

  The first embodiment includes an image input unit that captures a living body and acquires a living body image, an authentication processing unit that processes a living body image, and a storage unit that stores registration information on biological features obtained from the living body. The processing unit performs color conversion to minimize color difference between biometric images at registration and authentication using registration information, and is an embodiment of a biometric authentication system that performs biometric authentication using a biometric image after color conversion. is there. In particular, the authentication processing unit uses the registration information to generate a color conversion matrix that minimizes the color difference between the biometric image at registration and at authentication, and generates a transformed image from the generated color conversion matrix and the biometric image, It is an Example of a structure which calculates and authenticates the similarity degree of a conversion image and registration information.

  The present embodiment may not be a system, but may be a device in which all or a part of the configuration is mounted on a housing. The device may be a biometric authentication device including an authentication processing unit, or the authentication processing unit may be performed by an external device, and the device is a blood vessel image acquiring device or a blood vessel image extracting device specialized for acquiring a blood vessel image. It is also good. Therefore, the various functions described below may be configured by a plurality of devices present on the network. The device may also be implemented as a mobile terminal such as a mobile phone, a laptop PC, a smartphone or a tablet.

  FIG. 1 is a diagram showing an entire configuration of a biometric authentication system using a blood vessel of a finger according to a first embodiment and a configuration of an authentication processing unit thereof. The biometric authentication system shown in FIG. 1A includes an image input unit 1, an authentication processing unit 2, a storage unit 3, a display unit 4 and an input unit 5. The image input unit 1 acquires an image including a biometric feature from a living body of a person to be authenticated, and inputs the acquired image to the authentication processing unit 2. The authentication processing unit 2 performs image processing on the image input from the image input unit 1 and executes authentication processing. In other words, the authentication processing unit 2 has an image authentication processing function.

  As shown in FIG. 1A, the authentication processing unit 2 includes a CPU (Central Processing Unit) 6, a memory 7, and various IFs (Interfaces) 8. The CPU 6 performs various processes by executing the program stored in the memory 7. The memory 7 stores a program executed by the CPU 6. The memory 7 also temporarily stores the image input from the image input unit 1.

  FIG. 1B shows a functional block diagram of the authentication processing unit 2. The authentication processing unit 2 having an image authentication processing function includes an authentication unit 9 and a registration unit 10. The authentication unit 9 performs image processing on a living body image which is input data input from the image input unit 1 at the time of authentication, and collates it with registration information registered in the storage unit 3 to authenticate the user. The registration unit 10 creates registration information of the biological image acquired by the image input unit 1 at the time of registration, and stores the registration information in the storage unit 3.

  Each process of the above-mentioned authentication processing unit 2 can be realized by various programs. For example, various programs stored in the storage unit 3 are expanded in the memory 7. The CPU 6 executes the program loaded into the memory 7 to execute the processing and operation of the authentication processing unit 2 described below.

  The IF 8 connects the authentication processing unit 2 to an external device. Specifically, the IF 8 is connected to the storage unit 3, the display unit 4, the input unit 5, the image input unit 1, and the like. The storage unit 3 stores in advance registration information of the user. The registration information is information on a biometric feature for matching the user, and is, for example, image information of a living tissue such as a finger vein pattern or a fat print pattern. Usually, the image of the finger vein pattern is an image obtained by imaging the finger vein, which is a blood vessel mainly distributed subcutaneously on the palm side of the finger, as a dark shadow pattern.

  The display unit 4 is, for example, a liquid crystal display, and is an output device that displays the information received from the authentication processing unit 2. In addition to the display unit 4, an audio output unit such as a speaker (not shown) can be connected as an output device, and information received from the authentication processing unit 2 can be transmitted as an acoustic signal, for example, voice. The input unit is, for example, a touch panel, and transmits information input by the user to the authentication processing unit 2.

  The image input unit 1 includes a color camera and has a plurality of light receiving sensors sensitive to a wavelength band of visible light. The color camera has, for example, three types of CMOS or CCD elements sensitive to blue (B), green (G), and red (R), and these are arranged in a grid shape for each pixel of the image. That is, the color camera has three light receiving sensors having different peak wavelengths of light receiving sensitivity. The light receiving sensitivity of each light receiving sensor has peaks at, for example, around 480 nm for blue, around 550 nm for green, and around 620 nm for red, and the spatial color distribution of light emitted from a living body can be acquired by the sensitivity of each wavelength.

  According to the above configuration, the image input unit 1 can acquire a plurality of biological features of the living body by measuring the light emitted from the living body with the light receiving sensor having peaks at different light receiving sensitivities. Here, the method of acquiring the information of the biometric feature used for personal identification from the image obtained by photographing the biometric with a color camera will be described in detail.

  Images of skin photographed in color by transmitting or reflecting visible light such as white light, shadows due to unevenness of the skin surface, white or yellowish brown of melanin in the epidermis, red of blood in the capillaries of the dermis, and It is observed in a superimposed state of colors exhibited by various living tissues existing inside and outside the skin, such as blue or gray of veins present in the subcutaneous tissue. In particular, as the color of the shallow skin layer from the epidermis to the dermis, the part where a large amount of blood is distributed is red, and the other part is dominated by the pigment of melanin and white or yellow.

  Since the patterns of these living tissues differ depending on the individual, the difference can be used to carry out personal identification. For that purpose, it is necessary to extract independently the pattern of each biological tissue as information of a biological feature. Generally, pattern extraction of living tissue utilizes differences in color or shape characteristics between living tissues. There is a finger as one of the living tissues. The finger can simultaneously measure a plurality of living tissues, and each of the plurality of fingers can measure living tissues such as veins, fingerprints, joint marks, fat marks and the like. Therefore, it is possible to obtain abundant biological information used for personal identification. In other words, abundant biological information can be obtained by using a plurality of finger images of a living body as the biological image processed by the authentication processing unit.

  The veins have a blue or gray pattern flowing toward the fingertips, and the fatty print has a white or tan mottled pattern. For separation of blood vessels such as veins and patterns of fat streaks, attention is paid to the fact that pigment concentrations of melanin and hemoglobin that constitute living tissue differ between veins and fat streaks. By solving the relationship between color information of the image and melanin and hemoglobin concentration by independent component analysis (ICA) or the like, it is possible to separate blood vessel and fat pattern patterns.

  In addition, the joint pattern is a low frequency pattern flowing in a direction orthogonal to the direction of the fingertip exhibiting redness, and is extracted, for example, from an image acquired by a light receiving sensor having sensitivity to blue or green by direction emphasis processing such as a two-dimensional Gabor filter. be able to. The fingerprint on the fingertip is a relatively high frequency spiral pattern, and a fingerprint pattern can be obtained, for example, by high frequency component emphasis processing.

  Here, since the color camera constituting the image input unit 1 in the present embodiment is normally mounted on a portable terminal such as a smartphone or a tablet, personal authentication by an image obtained by photographing a living body with the color camera Is expected to be implemented mutually between multiple devices. For example, it is expected that biometric information is registered with a personal smartphone at home and biometric authentication is performed with a common tablet in a public place. At this time, the optical characteristics of the lens, the spectral sensitivity characteristics of the light receiving sensor, the characteristics of the internal signal processing, and the like differ depending on the device that captures the living body. In addition, as illumination light at the time of shooting, various light sources such as white LED, fluorescent lamp, and incandescent lamp are assumed. The overall or local color tone of the living body changes based on the individual differences between the spectral sensitivity and the characteristics of the internal signal processing and differences in imaging conditions due to illumination variations and the like.

  As described above, when the color tone of a living body image obtained by shooting a living body fluctuates due to fluctuation of imaging conditions such as illumination at the time of imaging or imaging device, each living body tissue based on color information such as veins and fat prints described above. It becomes difficult to perform pattern extraction with high accuracy.

  In order to solve this problem, as a method of setting an image obtained by shooting the same subject under different devices or illumination to a certain color tone, a common color chart is shot by each device, and the obtained image is obtained. There is a method of generating a color conversion matrix based on image signal values corresponding to each color in a color chart. Although this method is a typical method for acquiring an image of constant color even if it is a combination of different devices and illuminations, it is necessary to capture a color chart for each combination of the devices and the illumination, and this embodiment It is not suitable when carrying out biometrics authentication anywhere, such as at home or in a public place as expected.

  Therefore, in the present embodiment, color calibration is performed based on color information of a plurality of portions of at least one living body image of registration time and authentication time, not color calibration between different devices and illuminations. That is, using the registration information stored in the storage unit, the authentication processing unit generates a color conversion matrix that minimizes the color difference between the biometric image at registration and at authentication, and converts the generated color conversion matrix and the biometric image Color calibration is performed by generating an image, that is, a converted living body image.

  In the color calibration of the living body image of the present embodiment, both the color of the living body image at the time of registration and the color of the living body image at the time of authentication are converted to be similar to the color of the living body image as a reference. The living body image serving as a reference is stored in the storage unit 3 as registration information related to the living body feature. As the living body image serving as the reference, a general living body image set in advance can be used. For example, a living body image obtained by photographing a living body with a combination of a plurality of devices and illumination can be used as a living body image serving as a reference. Alternatively, one of the biometric images at the time of registration or authentication may be selected and used as a reference biometric image. Hereinafter, the color calibration method of the present embodiment using a general living body image as a reference living body image will be described in detail.

  FIGS. 2A, 2B, and 2C show an example of a color conversion process flow at the time of registration and authentication of a living body image using a finger as a living body, and a modification thereof as one of the processing flows of the color calibration method of this embodiment. Is shown. The color conversion process flow is mainly processed by the CPU 6 of the authentication processing unit 2. Hereinafter, the color conversion processing flows of FIG. 2A and FIG. 2B will be sequentially described. First, an image obtained by photographing a living body used at the time of registration or authentication, from which an area other than the living body area such as the background is removed is taken as a living body image (S100, S107).

  Next, the detected living body image is divided into a plurality of areas (S101, S108). As a method of dividing this area, as shown in FIG. 4, even if it is divided into a plurality of finger areas, it is a joint print of finger 17 which is biological information that can be extracted relatively robustly against variations in illumination and individual differences of devices. Good. Since the finger except the thumb generally has two joint marks, it may be divided into a total of four areas of the three finger areas 12 separated by the joint marks and the joint area 11.

  Next, the representative color of each of the four regions obtained by dividing the living body image, that is, the representative color is obtained (S102, S109). The representative color may be, for example, an average value or a median value of colors in each region, and determined so as to minimize the influence of image noise due to back light etc. and specular reflection or shadow on the living body surface due to illumination. .

  Next, a data set is created in which the representative living body image serving as the reference and the representative color data for each region of the living body image at the time of registration or authentication are vector pairs (S103, S110). In FIG. 3, a hand image represented by three bands of RGB is acquired by color photography, four fingers are detected from this hand image, and each finger N is divided into four regions R. An example is shown. As illustrated in the table 18 in the figure, a color vector of N × R (= 16) pairs of length B (= 3) is a pair of a living body image serving as a reference and a living body image at the time of registration or authentication. You can define the data set you

  As described above, in the case where the regions are first divided by each finger of a plurality of fingers, the representative color may be further regenerated from the representative color of the same portion between the fingers. For example, the average value of representative colors of the areas of the fingertips of four fingers is used as the representative color. At this time, as a representative color of the living body image serving as a reference, for example, an average value of representative colors obtained from an image obtained by photographing a living body with a combination of a plurality of devices and illumination can be used. In any case, a representative color is acquired for each of a plurality of regions, and a color conversion matrix is generated using the acquired representative colors.

From the obtained data set, a color conversion matrix of a living body image at registration or at authentication is generated, and color conversion is performed on each living body image (S104, S111). In other words, the authentication processing unit generates a color conversion matrix using the reference biological image and the biological image. That is, generation of the color conversion matrix is performed, for example, by squaring the representative color vector (x _rn ) of the region r of the finger n, which is a living body image at registration or authentication, and the color vector of a representative color of the reference living body image. The error can be determined to be small and preferably minimized. Equations 1 and 2 show examples of the color conversion matrix. Here shows the A is uniform color transformation matrix does not depend on the body part of the formula 1, M _rn of formula 2 shows a different color conversion matrix for each body part.

In Equations 1 and 2, x _rn ^en is a representative color vector (length B) of the region r of the finger n at the time of registration, and x _rn ⁱⁿ is a representative color of the region r of the finger n at the time of authentication The vectors (length B), x 1 _rn , _..., X _Brn respectively indicate the elements (B pieces exist per vector) of the representative color vector of the region r of the finger n.
Also, generation of the color conversion matrix may be performed for each divided area. Furthermore, when the pixel performing the color conversion belongs to the area between the first joint and the second joint of the finger, weighting is applied to the data set in the area between the first joint and the second joint of the finger and its neighboring area Then, a color conversion matrix may be obtained. Then, using the color conversion matrix obtained, the signal value of each pixel of the biometric image at the time of authentication is converted. Then, the feature quantities are extracted using the color-converted raw image (S105, S112), and at the time of registration, the extracted feature quantities are registered in the storage unit 3 (S106), and the process is ended. On the other hand, at the time of authentication, the extracted feature amount is compared with the registered feature amount (S113), and the result is output and the like, and the process is ended.

  As described above, in the configuration of the present embodiment, the authentication processing unit preferably divides the finger image into a plurality of areas for each finger, generates a color conversion matrix for each of the plurality of divided areas, and performs conversion. An image is generated, and the degree of similarity between the converted image, which is a living body image after conversion, and the registration information is calculated to perform authentication. In the color calibration method of the present embodiment described above, color conversion is performed on the living body image at the time of registration and authentication using the living body image serving as a reference created in advance, but as described above, the living body image serving as a reference The color conversion may be performed only on the living body image at the time of authentication using the living body image at the time of registration. That is, the registration information includes a living body image at the time of registration as a living body image serving as a reference, and the authentication processing unit generates a color conversion matrix using the living body image serving as the reference and the living body image at the time of authentication. FIG. 2C shows an example of the registration process flow in that case. As clearly shown in FIG. 2C, FIG. 2B is different except that creation of a reference-registered color data set (S110) and color conversion of a living body image (S111) are omitted and creation of registered color data (S117) is added. Since it is the same as the above, the explanation is omitted. Conversely, the biometric image serving as the reference may be a biometric image at the time of authentication, and color conversion may be performed only on the biometric image at the time of registration. In this specification, also in this case, color conversion is performed to minimize the color difference between the biometric image at the time of registration and at the time of authentication.

  The configuration for performing color calibration using the color conversion matrix of the present embodiment described above minimizes the color difference between the biometric image at the time of registration and the biometric image at the time of authentication as much as possible, and captures images such as illumination variations and errors between devices It is possible to extract stable feature quantities that do not depend on the difference in conditions, and to realize high accuracy of authentication.

  The second embodiment is an embodiment of a biometric authentication apparatus further provided with an auxiliary light source that generates light different from ambient light in addition to the color camera. That is, the apparatus further includes an auxiliary light source that can be turned on and off, the image input unit acquires a living body image when the auxiliary light source is on and off, and the authentication processing unit uses the living body image when on and off This is an embodiment of a biometric authentication device capable of removing ambient light components.

  FIG. 5 is a schematic view showing biometrics by the device of the second embodiment. In the present embodiment, using the portable terminal 15 having the color camera 13 and the auxiliary light source 14, biometric authentication that is robust against illumination variations at the time of shooting is performed. As the portable terminal 15, one having a color camera 13 and an auxiliary light source 14 such as a portable telephone or a smartphone can be used.

  FIG. 6 shows an example of a processing flowchart of the apparatus of the second embodiment. The CPU in the authentication processing unit 2 mainly processes the process flow in the same figure as the process flow in FIGS. 2A to 2C. First, a finger is photographed while the auxiliary light source 14 is turned off, and a living body area such as a finger is detected from the obtained image (S200). For detection of a living body region, rule-based extraction processing may be performed using feature quantities such as a skin-specific color and a shape such as a skeleton. Alternatively, images obtained by photographing a sufficient number of living bodies may be prepared, and living body regions may be extracted from the images using deep learning or the like represented by Convolutional Neural Networks (CNN).

  Next, as in the first embodiment, the obtained living body image is divided into one or more regions (S201), representative colors of the respective regions are obtained (S202), and temporarily stored in the memory 7. As for the details of the segmentation of the living body image and the acquisition of the representative color, as described in the first embodiment, when each finger of a plurality of fingers such as four is segmented, the representative color of the same part between the fingers is further extracted The representative color may be regenerated. For example, an average value of representative colors of areas of fingertips of four fingers is used as a representative color of the area. At this time, as a representative color of a living body image serving as a reference, an average value of representative colors obtained from an image obtained by photographing a living body with a combination of a plurality of devices and illumination may be used as in the first embodiment. .

  Next, it is checked whether shooting is completed in each of the states where the auxiliary light source 14 is turned on and off (S203). If either one of the shooting has not been completed, the lighting state of the auxiliary light source 14 is switched (S204), and processing from shooting to acquisition of the representative color (S202) is performed again, and the auxiliary light source 14 is turned on and off again. In each of the above states, it is confirmed whether the imaging is completed (S203).

  When both imaging ends (YES), the living body image obtained by imaging the living body in a state in which the auxiliary light source 14 is turned on, the pixel value belonging to each area divided as described above, and the living body in a state in which the auxiliary light source 14 is turned off The environmental light component is removed by obtaining the difference in the value of the representative color of each area obtained by dividing the living body image obtained by photographing the image (S205). An image obtained by performing this removal process on each pixel can be approximately regarded as an image obtained by imaging a living body in an environment in which only the auxiliary light source is turned on.

  Next, the representative color of the living body image obtained by removing the ambient light component is acquired in this manner, and a data set consisting of a color vector pair with the representative color of the living body image serving as a reference is generated as in the first embodiment. (S206). A color conversion matrix is generated from the determined data set, and color conversion of each pixel of the living body image at registration or at authentication is performed (S207).

  Then, feature quantities are extracted using the living body image after color conversion, and registration or collation is performed. With the above-described auxiliary light source on / off processing according to the present embodiment, it is possible to acquire a living body image that is being photographed under constant illumination even when environmental light differs between registration and authentication, for example. .

  FIG. 7 shows an example of a color space in which pixel values of a living body image obtained by photographing the living body 16 under different illuminations in the configuration of the present embodiment. For convenience, although shown in a two-dimensional space provided with an X-axis and a Y-axis, it is actually a three-dimensional space provided with an X-axis, a Y-axis and a Z-axis. In this color space, for example, environmental light at registration is environmental light A, and environmental light at authentication is environmental light B. At the time of registration, the auxiliary light source 14 is turned on under ambient light A to photograph the living body and the color space 21 of the living body image obtained by the photographing processing of this embodiment described above and the auxiliary light source under ambient light A By turning off the light 14 and synthesizing a living body image obtained by shooting a living body, the color space 23 of the living body image obtained by shooting the living body under the auxiliary light source 14 can be generated in a pseudo manner. Similarly, also at the time of authentication, the effect of the ambient light B is generated by combining the color space 22 of the living body image obtained by capturing the living body by turning on the auxiliary light source 14 and the image capturing the living body by turning off the auxiliary light source 14. Alternatively, a living body image 23 obtained by capturing a living body under an auxiliary light source can be artificially generated.

  According to the configuration of the present embodiment that performs auxiliary light source on / off switching as described above, a biometrics authentication device capable of performing biometrics without being affected by ambient light A at registration and ambient light B at authentication. It will be possible to provide.

  In the third embodiment, a guidance display is provided as a guiding mechanism for generating a color conversion matrix stably and with high accuracy when using a device having different structures such as the size of the housing and the position of the camera at registration and at authentication. It is an Example of the biometrics apparatus of the structure to perform. That is, the authentication processing unit controls the image input unit to display guidance for acquiring a living body image on the display unit.

  In portable terminals such as mobile phones and smart phones owned by individuals, users are often familiar with the position of the camera of these terminals, and it is considered that a living body can be easily presented to the camera. However, in a common tablet or the like installed in a public place such as a POS (Point of Sales) terminal, the user does not know the position of the camera of the terminal and struggles to present a living body to the camera. Sex may be reduced.

  When a plurality of fingers are used as an example of a living body used for authentication, if it is a portable terminal such as a portable telephone or a smart phone that is a relatively small device, as shown in FIG. By holding it over the back of the device, it is possible to shoot multiple fingers with an out camera that is typically mounted on the back of the device. Mobile devices such as mobile phones and smartphones are relatively small compared to notebook PCs and tablets, and are generally designed so that the position where the user holds the chassis by hand and the out camera do not interfere, so most of the out cameras are It is located at the top of its frame. Therefore, the user can shoot a plurality of fingers almost uniformly by holding the hand over the back of the housing of the portable terminal 15 as shown in FIG.

  Next, consider the case of photographing a finger with a relatively large terminal such as a laptop PC or a tablet. Since the user can see the in-camera arranged on the same side as the display screen, the user can intuitively present a plurality of fingers by holding the hand toward the screen. However, in fact, it is necessary to correct the position and orientation of the hand, because the first hand raised is too small or too large, and the hand holding position is shifted. At this time, even if it is attempted to correct the position or orientation of the holding hand according to the image of the hand displayed on the screen, for example, the hand overlapping the display screen overlaps the position and orientation of the hand. In addition, shooting with an in-camera is often in a backlit state, and the quality of a living body image tends to deteriorate, such as an increase in image noise. Furthermore, an object of a color similar to a finger, such as a face, is often reflected as a background, which makes it difficult to detect only a finger from an image obtained by photographing.

  On the other hand, when shooting a finger using an out-camera on the back of a tablet or the like, it becomes less likely to be backlit or the living body other than the hand is reflected in the background, and a small device such as a mobile phone or smartphone It is possible to carry out shooting under conditions similar to those for shooting. However, when putting a hand over an out-camera such as a tablet installed in a public place, the user does not know the position of the out-camera on the back of the tablet and does not know how to hold his hand. Is considered to occur. Therefore, in order to perform biometric authentication on a large device such as a tablet under the same shooting conditions as a small device such as a mobile phone or a smartphone, the user can easily take a living body to the out camera which is the back camera of the tablet. It is important to design a guiding mechanism that can be held up.

  FIG. 8 shows an example of the authentication processing flow of the biometrics authentication system of the present embodiment provided with a guiding mechanism for holding the hand over the out camera of a relatively large frame, for example, a tablet. This processing flow can also be realized by the CPU 6 of the authentication processing unit 2 mainly controlling the screen display of the display unit 4. In this embodiment, first, the out-camera position is acquired (S300), and the guide screen at the time of authentication is set and displayed on the display unit 4 of the apparatus based on the acquired out-camera position (S301), and biometric authentication is performed ( S302).

  That is, in the biometric device according to the present embodiment, the position of the out camera of the tablet, which is a relatively large housing compared to a mobile phone or a smartphone, is acquired. As a method of acquiring the out-camera position, for example, the software developer or the operation manager may investigate the camera position of each case in advance and store it in the memory in the tablet, but in the present embodiment, A method of automatically acquiring an out-camera position using a touch operation on the back of a housing will be described.

  FIG. 9 shows an example of a guidance display prompting the user 20 to touch the back of the tablet housing 19 in order to obtain the out-camera position. In the example of FIG. 9, the lower left of the back of the tablet housing 19 is urged to touch the user 20, and the user 20 touches the lower left of the back of the tablet housing 19 to the out camera Shoot. The guidance screen shown in FIG. 9 is an example for prompting the user 20 to touch a specific part of the housing 19, and the figures and characters in the drawing are display examples of guidance as a guiding mechanism for guiding the user. If there is an out camera at the position touched by the user 20, an image in which a part or the whole of the angle of view is blocked by the hand is obtained, and if there is no out camera at the touched position, an image with less blocking is obtained.

  As a method of detecting that the back surface of the tablet case 19 is touched using the obtained image, for example, if the tablet is fixed, multiple frames of the moving image taken by the out-camera are used. A moving object may be detected by applying a background subtraction method or the like. If the tablet is not fixed, it is assumed that the living body as the user 20 touches the tablet, the ratio of skin color and red pixels to the image taken with the out-camera is determined, and the frame with a large ratio of the pixels is used It may be an image obtained by photographing while the person 20 is touching the tablet.

  The shooting is repeated while changing the position where the user 20 is touched to, for example, the lower middle, lower right, upper left, upper middle, upper right of the back of the tablet housing 19 and the images obtained in each case are compared By doing this, you can get the rough position of the out camera.

  FIG. 10 shows a screen display example for guiding the living body 16 to be presented to the user at the time of authentication based on the position of the out camera acquired using the guidance screen display of FIG. 9 in the configuration of the present embodiment. This example is an example in the case where the position of the out camera 24 of the tablet is in the lower left area 25 on the back surface of the housing 19.

  By displaying the image obtained by photographing at the same position as the position where the out camera 24 of the photographing screen of the tablet is located, the deviation between the position where the user holds the living body 16 and the position displayed on the photographing screen is reduced. Intuitively, the living body 16 can be presented.

  The example of the screen display for acquiring the position of the out camera 24 described above and guiding the living body 16 to be presented on the back of the housing 19 is not limited to the housing having the horizontally long screen shown in FIGS. As shown in FIG. 5, the mobile terminal 15 such as a tablet can be used by standing or using it, a mobile phone, a smartphone or the like.

  The present invention is not limited to the embodiments described above, but includes various modifications. The above embodiments are described in detail in order to explain the present invention in an easy-to-understand manner, and are not necessarily limited to those having all the configurations described. For example, in the configuration of the second embodiment, when the illuminance of ambient light can be considered to be sufficiently smaller than the illuminance of the auxiliary light source, an image obtained by photographing the living body by lighting the auxiliary light source is an image obtained after removing the ambient light component It may be regarded. Also, part of the configuration of one embodiment can be replaced with the configuration of another embodiment. Also, the configuration of another embodiment can be added to the configuration of one embodiment. In addition, with respect to a part of the configuration of each embodiment, another configuration can be added, deleted, or replaced.

  Each of the above configurations, functions, processing units, etc. may be realized by hardware, for example, by designing part or all of them with an integrated circuit. Further, each configuration, function, etc. described above may be realized by software by the processor interpreting and executing a program that realizes each function. Information such as a program, a table, and a file for realizing each function can be stored in a memory, a hard disk, a recording device such as a solid state drive (SSD), a recording medium such as an IC card, an SD card, or a DVD.

  In the configuration of the above-described embodiment, the control lines and the information lines indicate what is considered necessary for the description, and not all the control lines and the information lines in the product are necessarily shown. All configurations may be connected to each other.

  The present invention is capable of mutually performing biometric authentication using a general-purpose color camera with different lighting devices and is useful as a personal authentication technology.

1 image input unit 2 authentication processing unit 3 storage unit 4 display unit 5 input unit 6 CPU
7 Memory 8 IF
9 authentication unit 10 registration unit 11 joint area 12 finger area other than joint area 13 color camera 14 auxiliary light source 15 portable terminal 16 living body 17 finger 18 table 19 case 20 user 21, 22, 23 color space 24 out camera 25 area

Claims (15)

An image input unit that captures a living body and acquires a living body image;
An authentication processing unit that processes the living body image;
A storage unit for storing registration information on biological features obtained from the living body, the authentication processing unit including:
Using the registration information, perform color conversion to minimize color difference between biometric images at registration and at authentication, and authenticate the living body using the biometric image after color conversion.
A biometric authentication device characterized by The biometric authentication device according to claim 1, wherein
The authentication processing unit
Using the registration information, generate a color conversion matrix that minimizes the color difference between the living body image at registration and authentication time, and generate a converted image from the generated color conversion matrix and the living body image, and the converted image Calculate the degree of similarity between and the registered information and perform authentication
A biometric authentication device characterized by The biometric authentication device according to claim 1, wherein
The living body image acquired by the image input unit is a finger image of the living body,
A biometric authentication device characterized by The biometric apparatus according to claim 2, wherein
The living body image acquired by the image input unit is an image of a plurality of fingers of the living body,
The authentication processing unit
The finger image is divided into a plurality of areas for each finger,
The color conversion matrix is generated for each of a plurality of divided areas to generate the converted image.
A biometric authentication device characterized by The biometric authentication device according to claim 4, wherein
The authentication processing unit
Acquiring a representative color for each of the plurality of regions, and generating the color conversion matrix using the representative color;
A biometric authentication device characterized by The biometric authentication device according to claim 4, wherein
The plurality of regions include a finger region and a finger joint region.
A biometric authentication device characterized by The biometric apparatus according to claim 2, wherein
The registration information stored in the storage unit includes a biometric image as a reference,
The authentication processing unit
Generating the color conversion matrix using the reference living body image and the living body image;
A biometric authentication device characterized by The biometric apparatus according to claim 2, wherein
The registration information stored in the storage unit includes, as a reference living body image, the living body image at the time of registration,
The authentication processing unit
Generating the color conversion matrix using the reference living body image and the living body image at the time of authentication;
A biometric authentication device characterized by The biometric authentication device according to claim 1, wherein
It has an auxiliary light source that can be turned on and off to generate light different from ambient light,
The image input unit acquires the living body image when the auxiliary light source is on and off;
The authentication processing unit
The ambient light component is removed using the on / off biological images
A biometric authentication device characterized by The biometric authentication device according to claim 1, wherein
A display unit capable of displaying the biological image acquired by the image input unit;
The authentication processing unit
The image input unit is controlled to display a guidance for acquiring the living body image on the display unit.
A biometric authentication device characterized by A biometric authentication method by an authentication processing unit,
The authentication processing unit
A living body image is taken and processed, and a living body image obtained is processed, and color conversion is performed to minimize color difference between the living body image at the time of registration and at the time of authentication using registration information on the living body feature obtained from the living body. Perform authentication of the living body using the living body image of
Biometric authentication method characterized by The biometric authentication method according to claim 11, wherein
The authentication processing unit
Using the registration information, generate a color conversion matrix that minimizes the color difference between the living body image at registration and authentication time, and generate a converted image from the generated color conversion matrix and the living body image, and the converted image Calculate the degree of similarity between and the registered information and perform authentication
Biometric authentication method characterized by The biometric authentication method according to claim 12, wherein
The authentication processing unit
Using a plurality of finger images of the living body as the living body image,
The finger image is divided into a plurality of areas for each finger,
Acquiring a representative color for each of the plurality of divided regions, and generating the color conversion matrix using the representative color;
Biometric authentication method characterized by The biometric authentication method according to claim 11, wherein
The authentication processing unit
The environmental light component is removed using the biological image acquired by switching on and off the auxiliary light source that generates light different from the environmental light.
Biometric authentication method characterized by The biometric authentication method according to claim 11, wherein
The authentication processing unit controls to display a guidance for acquiring the biological image on a display unit.
Biometric authentication method characterized by

Images