JP2009009403A - Biometrics device and living body detection method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a biometrics device having simple sensor structure and having a living body detection method which hardly depends on the surface states of fingers and individual differences in the surfaces of fingers. <P>SOLUTION: The biometrics device acquires an image for biometrics by utilizing an illumination means and scattered light in an object which is generated by the illumination means and has a comparison means for acquiring scattered light distribution information data in the subject at different wavelengths and comparing respective scattered light distribution information data and a living body detection means for executing living body detection on the basis of the comparison result. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to an image acquisition apparatus for biometric authentication and a biometric authentication apparatus using the same, and more particularly to a biometric authentication apparatus that has illumination means and uses scattered light inside a subject. The biometric authentication device is applied as a fingerprint authentication, a palmprint authentication device or the like.

  In recent years, in order to ensure security for personal information and confidential information, biometric authentication systems such as fingerprint authentication devices have attracted attention, and demand for office devices and portable devices is also increasing. Such a biometric authentication system using fingerprints, veins, faces, irises, palm prints, etc., acquires an image of a living body in an image acquisition device, performs feature extraction from the acquired image, and has registered data based on that information And authenticate the identity.

  Here, as a detection method of the image acquisition device used in the biometric authentication system, there are an optical method using an imaging element such as a CCD or a CMOS sensor, a capacitance method, a pressure detection method, a thermal method, an electric field detection method, and the like. is there. Among optical methods, a reflection optical method and a scattered light method are known. Taking a finger as an example of a living body part, a reflective optical system is a method in which a finger is placed on a prism and light is incident from an imaging surface to capture reflected light on the finger surface. On the other hand, a finger is placed on a plate-shaped optical member provided on the image sensor, the finger is illuminated from a region other than the imaging surface, and the light emitted from the inside of the finger to the finger surface is scattered from the inside of the finger. A method of receiving light with an image sensor through a member is a scattered light method inside a finger. The amount of light from the ridge (convex portion) of the fingerprint that is in contact with the plate-like optical member provided on the image pickup device reaches the image pickup device through the plate. On the other hand, light from portions other than the ridge portion (concave portion) of the fingerprint that does not contact the plate-shaped optical member is scattered in the space between the finger surface and the plate, so that the amount of light incident on the image sensor through the plate is small. By using this contrast difference, a fingerprint image is acquired. The optical system that uses scattered light inside the living body (inside the subject) is less susceptible to the state of the living body surface (dry or wet) than the reflective optical system that uses a prism. Known as an image acquisition method.

  In such a biometric authentication apparatus, in addition to the authentication capability, a function for detecting an object other than a living body to be originally authenticated, a fake living body, and the like is required. For example, in the case of fingerprint authentication, the fake living body includes a fake finger image made of a resin or plastic, a fake fingerprint image created with a photograph, or a product obtained by transferring a residual fingerprint to a film.

  Patent Document 1 discloses an example of an authentication device having a function of identifying a fake living body. Here, using an optical image sensor having a block having infrared sensitivity and a block having no infrared sensitivity, the obtained image is subjected to Fourier transform or the like, and a living body and a pseudo living body are discriminated by a difference in clarity. Is disclosed.

Patent Document 2 discloses a configuration that uses the extinction coefficient of hemoglobin. Probe light and reference light are emitted to the finger in contact with the inspection surface. Receives transmitted probe light that has passed through the inside of the finger, receives electrical signals according to the light intensity of the transmitted probe light, and transmitted reference light that has passed through the inside of the finger, and responds to the light intensity of the transmitted reference light Output electrical signals. A transmitted light intensity ratio between the light intensity of the transmitted probe light and the transmitted reference light is obtained from both of the output electrical signals, and whether or not the finger is a living body is identified from the transmitted light intensity ratio.
JP 2002-042117 A JP 2000-20684 A

  In an optical biometric authentication device, there is a concern about the following restrictions in the configuration for discriminating between a living body and a fake living body from the clarity of the image of the living body and light intensity information.

  One is that there are cases where it is difficult to discriminate depending on individual differences and race differences. The method of achieving only the intensity ratio and the clearness between infrared and visible light has a problem that “the output obtained by the skin color and finger surface condition is greatly influenced” (problem α). The transmittance varies greatly depending on individual differences and races. In addition, the sharpness varies greatly depending on the dry and wet state of the finger and the rough state of the individual finger. As a result, if the allowable range of variation is increased, the false living body is erroneously recognized as a living body, and if the allowable range of variation is decreased, the living body is erroneously recognized as a false living body. Furthermore, if this is abused, false detection may be made by using a material that matches the vividness of the infrared and visible light transmittance and the fingerprint pattern of the finger, such as a film printed with a color fingerprint image. It is possible to create a fake finger to wake up (task α).

  Furthermore, when identifying wavelengths that are close to infrared sensitivity and red, for example, it is necessary to control the sensitivity to wavelength with high accuracy, and a pixel block with a difference in infrared sensitivity is required. There was a problem ”(Problem β).

  Also, when captured near a window with a lot of infrared light in the western sun, and under fluorescent lamps, “the infrared ratio changes due to the entry of external light, and the sharpness and light intensity change. There is also a concern (problem γ).

  Therefore, the present invention uses a method different from the intensity ratio and the sharpness according to the wavelength, and has a simple sensor structure and is less dependent on the finger surface condition and individual differences of the finger surface (solving the problems α and β). An object of the present invention is to provide a biometric authentication device having a biometric detection method. It is another object of the present invention to provide a living body detection method that is resistant to external light (solves the problem γ).

  In order to solve the above-described problems, the present invention focuses on the fact that in the biometric authentication device using scattered light inside the finger, a difference occurs in the scattered light distribution depending on the location of the subject, such as the center or the periphery of the finger. In particular, in visible light compared to infrared light used for authentication image acquisition, the ratio of the amount of light attenuated with respect to distance increases as the wavelength becomes shorter, and scattering is likely to be hindered by components such as bone. Become. It has been found that a fake finger and a residual fingerprint can be discriminated using a change in the scattered light distribution (location dependence, luminance shading) depending on the wavelength.

  By comparing the scattered light distribution at each wavelength using a plurality of visible wavelengths, three-dimensional information of the subject can be obtained, so that a fake finger or a residual fingerprint can be easily identified. Since it is not the absolute value or ratio of light intensity, it is difficult to depend on individual differences such as the color of the finger surface. In addition, since the use of shading is used, which does not use a fine image pattern such as sharpness, and uses a relatively gentle place dependency = shading that is close to the shape of the finger, it is difficult to depend on individual differences such as the finger surface condition (problem α Is resolved).

  Moreover, since an image of a plurality of visible wavelengths can be obtained with an easily available LED light source having a plurality of visible wavelengths, a color filter provided in a pixel shape, etc., the sensor structure is not complicated and can be realized with a simple configuration ( Problem β is solved).

  Further, by controlling the light source and the image reading means, it is possible to perform biometric detection and authentication after removing the influence of external light (the problem γ is solved).

  Utilizing this fact, the following configuration was adopted.

  A first biometric authentication device according to the present invention is a biometric authentication device that acquires an image for biometric authentication using illumination means and scattered light inside a subject generated by the illumination means, and has different wavelengths. A comparison means for acquiring the scattered light distribution information in the subject and comparing the scattered light distribution information, and a biological detection means for performing biological detection based on a result of the comparison.

  In the biometric authentication apparatus, the illumination unit includes a first illumination unit having an infrared wavelength for biometric authentication and a second light unit having a visible light wavelength for comparing scattered light distribution information in the subject. And illumination means.

  In the biometric authentication device, the second illumination unit includes two light sources having different wavelengths arranged in the vicinity of each other, and the biometric authentication device further switches and emits the light sources sequentially. You may make it have.

  In the above biometric authentication device, the second illumination unit emits at least light having a wavelength band longer than 610 nm (about 610 to 800 nm) and light having a wavelength shorter than 570 nm (about 400 to 570 nm). You may make it radiate | emit.

  In the biometric authentication apparatus, the second illuminating unit may emit red light and green light.

  The biometric authentication apparatus may further include an image reading unit, and the image reading unit may include a wavelength separation unit.

  The biometric authentication apparatus may include means for detecting luminance shading caused by the shape of the subject as the scattered light distribution information in the subject.

  In the biometric authentication device, the biometric authentication device includes an image reading unit having an optical member that is in close contact with the subject and an image sensor that is in close contact with the optical member, and the second illumination unit is configured to display the subject image. You may make it arrange | position in the position which illuminates the said to-be-photographed object from the other area | region of the imaging surface of the optical member made to closely_contact | adhere for acquisition.

  In the biometric authentication apparatus, the second illumination unit is disposed on the same surface as the imaging surface of the optical member with respect to the subject, and is disposed at a position where light is emitted from a lower portion that is substantially in contact with the subject. It may be.

  A second biometric authentication device according to the present invention is a biometric authentication device that includes an illuminating unit and an image reading unit that acquires an image for biometric authentication using scattered light inside the subject generated by the illuminating unit. The illumination means has an external light removal means for acquiring a difference image between an image when the light is emitted and an image when the light is not emitted.

  In the biometric authentication device, the image reading unit includes a pixel group provided with a unit for separating visible light, and the biometric authentication device performs switching for switching a wavelength band of an image acquired by switching the pixel group. You may make it have a means and a means to compare the acquired image mutually.

  A third biometric authentication device according to the present invention is a biometric authentication device that includes an illuminating unit and an image reading unit that acquires an image for biometric authentication using scattered light inside the subject generated by the illuminating unit. The image reading unit includes a pixel group provided with a wavelength separation unit, and the biometric authentication device mutually exchanges the acquired image with a switching unit that switches the wavelength band of the acquired image by switching the pixel group. It has the means to compare with.

  The living body detection method according to the present invention includes a first step of acquiring a subject image having a first wavelength component, a second step of acquiring a subject image having a second wavelength component, the first step, and the second step. A third step of comparing the scattered light distribution information in a plurality of subject surfaces having different wavelengths from the subject image obtained in the step of step 4, and a fourth step of discriminating whether or not the subject is a living body based on the comparison result It is characterized by having.

  In the living body detection method, a fifth step of illuminating the subject with a first light amount, a sixth step of illuminating the subject with a second light amount, the fifth step, and a sixth step A seventh step of removing an external light component by obtaining a difference from the image obtained in step (b), and a step of performing living body detection of the subject using the image obtained from the seventh step. Also good.

  Since the present invention can acquire three-dimensional information from the difference in wavelength dependence of the internal scattered light according to the subject shape by acquiring the scattered light distribution information in the subject plane, It is possible to distinguish between a living body and a non-living body without being affected by individual surface conditions or individual differences on the finger surface.

  In addition, a living body detection method that is resistant to external light can be provided.

  Next, embodiments of the present invention will be described with reference to the drawings.

(First embodiment)
Next, the configuration of the first embodiment to which the present invention is applied will be described with reference to FIGS.

  Here, as the biometric authentication device, an authentication device using an optical fingerprint sensor that reads a fingerprint by giving scattered light to the inside of the finger and bringing the finger into close contact with the imaging unit is shown.

  FIG. 1 is a block diagram showing a schematic configuration of a fingerprint authentication device as Embodiment 1 of the present invention. FIG. 2 is a schematic diagram showing an external configuration.

  The fingerprint authentication apparatus according to the present embodiment includes an image acquisition unit 101 and an authentication unit 102 as an image acquisition apparatus. For example, the image acquisition unit 101 is an imaging unit having an image sensor, and the authentication unit 102 is a combination of functions executed by a personal computer. Alternatively, the image acquisition unit 101 and the authentication unit 102 may be combined as one fingerprint authentication unit and may be an independent device connected to a personal computer (not shown).

  In the image acquisition unit 101 in FIG. 1, reference numerals 103a and 103b denote LED light sources as illumination means. Reference numeral 103a denotes a light source 1 as a first illumination unit having an infrared wavelength for acquiring a subject image for authentication. Reference numeral 103b denotes a light source 2 as a second illuminating means having a visible light wavelength for acquiring a subject image for discriminating whether a living body or a non-living body is a false finger or the like. Here, 103b uses a white LED including both a red wavelength (first wavelength component) and a green wavelength (second wavelength component). Reference numeral 108 denotes a light source control unit that controls the luminance and lighting timing of each LED, and reference numerals 112a and 112b denote control signal lines.

  Reference numeral 104 denotes an image sensor such as a CMOS type or a CCD type as image reading means. In this embodiment, a CMOS image sensor is exemplified. Reference numeral 105 denotes a drive pulse generator (TG unit) that controls the sampling timing of the image sensor 104 and the ADC (analog / digital converter) 107. A signal switching unit 106 switches between two analog outputs from the image sensor 104. The image sensor 104 can output an output for each different color filter using a color filter as a wavelength separation unit described later. Reference numeral 107 denotes an analog / digital converter (ADC). 110a, 110b, 110c are analog image data signal lines, 110a is an output signal line for a pixel with a red color filter, and 110b is an output signal line for a pixel with a green color filter. Reference numeral 110c denotes an output signal line in which one of them is selected. Reference numeral 113 denotes a digital image data signal line. Reference numerals 112d and e denote signal lines of drive pulses sent from the drive pulse generator to the image sensor 104 and the analog / digital converter (ADC) 107.

  In the authentication unit 102, an image of visible light components of red and green (components having a wavelength of about 610 to 800 nm and components of about 400 to 570 nm) acquired separately by the authentication unit 102 using a color filter as a wavelength separation unit described later. It is a difference production | generation part as a comparison means to compare. While temporarily storing a plurality of images in the frame memory 117, a difference between the images is generated. Reference numeral 114 also serves as an external light component removing unit to be described later. Reference numeral 115 denotes a living body detection unit serving as a living body detection unit for identifying whether a living body or a living body other than a living body such as a pseudo living body. The living body detection unit 115 determines the difference between the red and green images using the difference in luminance distribution in the subject plane.

  Reference numeral 116 denotes a pre-processing unit that performs image processing such as edge enhancement in order to perform feature extraction at a later stage. Reference numeral 118 denotes a feature extraction unit, and reference numeral 119 denotes a registration collation unit that registers the individual features extracted in 118 in the database or compares and collates with registered data. A database 120 stores personal data. A control unit 123 controls the operation of each unit.

  Reference numerals 124a, b, c, d, and e denote data lines for transmitting image data. Reference numeral 125 denotes a data line and a control line between the database and the registration / verification unit. 126 is a signal line for controlling the operation of the light source control unit 108, and 127 is a signal line for controlling the operation of the drive pulse unit (TG unit). Reference numeral 128 denotes a switching signal line for the drive pulse unit (TG unit) 105 to switch the signal switching unit 106 under the control. Reference numerals 129a to 129 denote control lines for the control unit 123 to control each unit of the authentication unit 102. Reference numeral 130 denotes a signal line that transmits the living body detection result to the control unit 123.

  In the schematic view of FIG. 2, FIG. 2 (a) is a cross-sectional view, and FIG. 2 (b) is a top view.

  Reference numeral 1 denotes a finger to be subjected to fingerprint authentication as a subject, which is placed in close contact with the imaging surface. Reference numerals 2a and 2b denote LED light sources as illumination means for irradiating the finger part. Among these, 2a is 103a of FIG. 1, and is the light source 1 as a 1st illumination means of the infrared wavelength for acquiring the to-be-photographed image for authentication.

  Reference numeral 2b denotes a light source 2 as a second illumination unit having a visible light wavelength for acquiring a subject image for discriminating whether a living body or a non-living body, a false finger, or the like. Here, 2b uses a white LED including both a red wavelength and a green wavelength.

  Reference numeral 3 denotes an optical member having a plate shape that guides the optical difference of the uneven pattern of the fingerprint to the image sensor 4.

  Examples of the plate-shaped optical member include a thin film having a transmittance at a desired wavelength having a thickness of about several to several hundreds of micrometers, a fiber, and a microlens array.

  Reference numeral 4 denotes an image sensor serving as an image reading unit, and here, a two-dimensional CMOS image sensor. Reference numeral 4a denotes a pixel portion, and 4b denotes a color filter portion which is wavelength separation means. Reference numeral 7 denotes a substrate on which an image sensor and LEDs are mounted.

  An infrared light source 1 (2a) for performing authentication is arranged around the finger, and illuminates the entire finger with infrared light that spreads from a distance slightly away. By doing so, the dependence of the position of the infrared light source on the acquired image is reduced, and the incident infrared light spreads evenly inside the finger, and a wide range of fingerprint patterns can be read.

  On the other hand, the light source 2 (2b) for discriminating whether it is a living body or a non-living body is provided at only one place immediately below the base of the finger. The light source at one location may be anything as long as it emits light of a plurality of wavelengths, but here it is a white light LED. The finger illuminated in white is imaged by the image sensor while being divided into red and green for each wavelength by the color filter (4b).

  Reference numerals 5a and 5b denote the incident directions of light from the LED 2b to the finger 1, 5a schematically showing light of a red wavelength component and 5b of light of a green wavelength component. The light beams 5a and 5b are actually mixed as white light and incident on the finger. However, the light beams 5a and 5b are scattered and transmitted in the finger depending on the wavelength. Reference numerals 6 a and 6 b denote light emission directions from the finger 1 toward the optical member 3 and the image sensor 4 after the light emitted from the LED 2 b is scattered in the finger 1. 6a schematically shows light of a red wavelength component, and 6b schematically shows light of a green wavelength component. 6a and 6b are wavelength-separated through the color filter 4b and imaged by the pixel unit 4a.

  The configuration of the CMOS type image sensor according to this embodiment will be described with reference to FIGS.

  FIG. 3 is a configuration diagram of the image sensor 104 of FIG. Here, the horizontal scanning direction in a general area sensor corresponds to the main scanning direction, and the vertical scanning direction corresponds to the sub-scanning direction. First, select one vertical row (for example, the top row), and select pixels from one horizontal end of the row to the opposite end of the same row (for example, from the left to the right). Read sequentially. Thereafter, the next vertical row is selected, and similarly, pixels are sequentially read from one end in the horizontal direction toward the opposite end of the same row. In this way, each row is read in the vertical direction to obtain the pixels of the entire screen. For this reason, the horizontal scanning is defined as main scanning, and the vertical scanning is defined as sub-scanning.

  Therefore, in the following description of the image sensor, the main scanning direction is described as the same meaning as the horizontal direction and the sub-scanning direction as the vertical direction.

  In FIG. 3, reference numerals 41a and 41b denote pixel portions constituting one pixel of the image sensor, and 41a denotes a red pixel (R) in the color filter 4b in FIG. 2 above the pixel, and 41b denotes a pixel. The green pixel (G) of the color filter 4b in FIG.

  Reference numeral 42 denotes an input terminal for a readout pulse (φS) in the pixel unit 41 (41a and 41b). Reference numeral 43 denotes an input terminal for a reset pulse (φR) in the pixel unit 41 (41a and 41b). Reference numeral 44 denotes an input terminal for a transfer pulse (φT) in the pixel unit 41 (41a and 41b). Reference numeral 45 denotes a signal readout terminal (P0) in the pixel unit 41 (41a and 41b). Reference numeral 46 denotes a signal line for sending a read pulse (φS) from the selector unit to the horizontal pixels, 47 denotes a signal line for sending a reset pulse (φR) from the selector unit to the horizontal pixels, and 48 denotes a horizontal line from the selector unit. This is a signal line for sending a transfer pulse (φT) to each pixel in the direction. Reference numeral 49 is a vertical signal line, 40 is a constant current source, and 51 is a capacitor connected to the vertical signal line 49. A transfer switch 52 has a gate connected to the horizontal shift register 56 and a source-drain connected to the vertical signal line 49 and output signal lines 53a and 53b. 54 a and 54 b are output amplifiers connected to the output signal lines 53 a and 53 b, and 55 a and 55 b are output terminals of the image sensor 6. 53a, 54a, and 55a are signal systems related to reading of the red pixel (R), and 53b, 54b, and 55b are signal systems related to reading of the green pixel (G). 55a corresponds to 110a in FIG. 1, and 55b corresponds to 110b in FIG.

  Reference numeral 56 denotes a horizontal shift register (HSR), 57 denotes an input terminal for the start pulse (HST), and 58 denotes an input terminal for the transfer clock (HCLK). 59 is a vertical shift register (VSR), 60 is an input terminal for the start pulse (VST), and 61 is an input terminal for the transfer clock (VCLK). 62 is a shift register (ESR) for an electronic shutter of a system called a rolling shutter described later, and 63 is an input terminal for the start pulse (EST). 64 is an output line of the vertical shift register (VSR), and 65 is an output line of the shift register (ESR) for the electronic shutter. Reference numeral 66 denotes a selector unit, 67 denotes an input terminal of the original signal TRS of the transfer pulse, 68 denotes an input terminal of the original signal RES of the reset pulse, and 69 denotes an input terminal of the original signal SEL of the read pulse.

  FIG. 4 is a configuration diagram of the pixel unit 41 (41a and 41b) in FIG. In FIG. 4, 71 is a power supply voltage (VCC), 72 is a reset voltage (VR), 73 is a photodiode, 74 to 77 are switches made of MOS transistors, 78 is a parasitic capacitance (FD), and 79 is a ground.

  Here, the operation of the image sensor 104 will be described with reference to FIGS. First, charges are accumulated by incident light in the photodiode 73 with the reset switch 74 and the switch 75 connected to the photodiode 73 turned off.

  Thereafter, the parasitic capacitance 78 is reset by turning on the switch 74 while the switch 76 is turned off. Next, the switch 74 is turned off and the switch 76 is turned on, whereby the charge in the reset state is read out to the signal readout terminal 45.

  Next, by turning on the switch 75 with the switch 76 turned off, the charge accumulated in the photodiode 73 is transferred to the parasitic capacitance 78. Next, the signal charge is read out to the signal readout terminal 45 by turning on the switch 76 with the switch 75 turned off.

  The drive pulses φS, φR, and φT of each MOS transistor are generated by vertical shift registers 59 and 62 and a selector unit 66 as will be described later, and are supplied to the input terminals 42 to 44 of the pixels through the signal lines 46 to 48, respectively. The For each pulse of the clock signal input from the input terminal 60, one pulse of the signals TRS, RES, and SEL is input to the input terminals 67 to 69, respectively. Therefore, drive pulses φS, φR, and φT are output in synchronization with the signals TRS, RES, and SEL, respectively. As a result, the drive pulses φS, φR, and φT are supplied to the input terminals 42 to 44.

  The signal readout terminal 45 is connected to the constant current source 40 by a vertical signal line 49 and is connected to the vertical signal line capacitor 51 and the transfer switch 52. The charge signal is transferred to the vertical signal line capacitor 51 via the vertical signal line 49. Thereafter, according to the output of the horizontal shift register 56, the transfer switch 52 is sequentially scanned, and the signal of the vertical signal line capacitor 51 is sequentially read out to the output signal lines 53 (53a and 53b). The read signal is output from the output terminal 55 (55a and 55b) via the output amplifier 54 (54a and 54b). Here, scanning of the vertical shift register (VSR) 59 is started by a start pulse (VST) 60. The transfer clock (VCLK) 61 is sequentially transferred to VS1, VS2,... VSn via the output line 64. The electronic shutter vertical shift register (ESR) 62 starts scanning with a start pulse (EST) input from the input terminal 63, and a transfer clock (VCLK) input from the input terminal 61 is sequentially transferred to the output line 65. It will be done.

  The readout order of each pixel unit 41 (41a and 41b) is as follows. First, the first row in the vertical direction is selected, and pixel units 41 (41a and 41b) connected to each column from left to right as the horizontal shift register 56 scans. ) Is selected and output. When the output of the first row is completed, the second row is selected, and the pixel portions 41 (41a and 41b) connected to the respective columns are selected and output from the left to the right as the horizontal shift register 56 scans again.

  Thereafter, similarly, according to the sequential scanning of the vertical shift register 59, scanning is performed from the first, second, third, fourth, fifth,.

  By the way, the exposure period of the image sensor is determined by the accumulation period in which the imaging pixels accumulate light charges and the period in which light from the subject is incident on the imaging pixels.

  Here, unlike an IT (interline transfer) type or FIT (frame-interline transfer) type CCD element, a CMOS type imaging element does not include a light-shielded buffer memory unit. For this reason, even during the period in which the signals obtained from the pixel portions 41 (41a and 41b) are sequentially read, the pixel portions 41 (41a and 41b) that have not yet been read continue to be exposed. Therefore, when the screen output is continuously read, the exposure time becomes substantially equal to the screen read time.

  However, when an LED is used as the light source and external light is not incident on the light shielding member or the like, it is possible to consider only the lighting period as the exposure period.

  As another method for controlling another exposure time, in a CMOS type image sensor, a driving method called an electronic shutter (focal plane shutter) is called a rolling shutter that performs vertical start and end vertical scanning in parallel. Can be taken. Thereby, the exposure time can be set in units of the number of vertical scanning lines at the start and end of accumulation. In FIG. 3, the ESR 62 is a vertical scanning shift register that resets pixels and starts accumulation, and the VSR 59 is a vertical scanning shift register that transfers charges and ends accumulation. When the electronic shutter function is used, the ESR 62 is scanned prior to the VSR 59, and a period corresponding to the interval is an exposure period.

  Next, the operation in the present embodiment will be described with reference to FIGS. FIG. 5 is a flowchart illustrating an image acquisition routine executed by the control unit 123 in the biometric authentication device of FIG. FIG. 6 is a diagram for explaining steps for removing external light. 7 to 9 are explanatory diagrams for performing living body detection by dividing luminance shading into red and green wavelength bands as the scattered light distribution inside the subject and comparing the distributions.

  In FIG. 5, when the routine is started in step 80, all light sources are turned off in step 81. In step 82, the signal switching unit 1006 in FIG. 1 is selected so as to obtain the pixel output of the red color filter (R). In step 83, one screen is acquired via the analog / digital converter (ADC) 107. The acquired image is temporarily stored in the frame memory 117 via the difference generation unit 114. At this time, since the light source is not attached, an image using mainly the red wavelength component of the external light component is obtained. Which band of the external light component is acquired depends on the characteristics of the color filter, and only red may be selected or may include infrared light. Here, infrared light is included.

  Next, the control unit 123 turns on the white light source in step 84 and acquires one screen via an analog / digital converter (ADC) in step 85. The acquired image is also temporarily stored in the frame memory 117 via the difference generation unit. At this time, since the white LED light source is attached, an image of a component in which the red wavelength component in the white LED light source is scattered inside the finger in addition to the mainly red wavelength component of the external light component is obtained.

  In step 86, the difference generation unit 114 obtains the image in step 83 and the difference image in step 85.

  Thereby, since the external light component is canceled, only the component in which the red wavelength component in the white LED light source is scattered inside the finger is obtained.

  Next, in step 87, the signal switching unit 106 in FIG. 1 is selected so as to obtain the pixel output of the green color filter (G). In step 88, an image is acquired. The acquired image is temporarily stored in the frame memory 117 via the difference generation unit 114. At this time, since the white LED light source is attached, an image of a component in which the green wavelength component in the white LED light source is scattered inside the finger in addition to the mainly green wavelength component among the external light components is obtained. Next, in Step 89, the white light source is turned off. In step 90, one screen is acquired. At this time, since no light source is attached, an image mainly composed of a green wavelength component among external light components can be obtained. The acquired image is also temporarily stored in the frame memory 117 via the difference generation unit 114. In step 91, the difference generation unit 114 obtains the image in step 88 and the difference image in step 90. Thereby, since the external light component is canceled, only the component in which the green wavelength component in the white LED light source is scattered inside the finger is obtained.

  Next, in step 92, a difference image between the image in step 86 and the image in step 91 is generated in the difference generation unit 114 in FIG. 1, and the living body detection unit 115 analyzes the difference to detect whether it is a living body.

  In step 93, when the control unit 123 determines that the living body detection unit 115 has detected the living body, the living body determination flag is output in step 94 and the process proceeds to step 95. In step 95, the light source 1 (103a) which is the infrared light source of FIG. 1 is turned on. In step 96, an authentication image is acquired. Although the color filter remains green, there is no problem because the green filter also transmits infrared light. Next, in step 97, the light source 1 (103a) which is an infrared light source is turned off. In step 98, by taking the difference between the image obtained in step 96 and the image obtained in step 90, an image of only the infrared light component of the light source 1 is obtained. Thereafter, the routine ends at step 100.

  In step 93, when the control unit 123 receives a detection result signal from the living body detection unit 115 and determines that it is a fake living body (not a living body), a false living body determination flag is output in step 99, and the routine is executed in step 100. finish.

  FIGS. 6A and 6B are schematic views for explaining this external light removal. The horizontal axis represents the position, and A and A 'correspond to A-A' in FIG. The vertical axis indicates the output level.

  Reference numeral 601 in FIG. 6A denotes an output level between A and A ′ of the image based only on the external light component acquired in step 83. Here, the fine image change component resulting from the ridge pattern of the fingerprint is omitted for explanation. Even in processing such as external light removal and biometric discrimination described here, such high frequency component image changes are not necessary as information, so in practice, before these processing, high frequency components such as LPF (low pass filter) are used. There are cases where processing is performed after image changes have been removed. In FIG. 6B, reference numeral 602 denotes an output level between A and A ′ of an image acquired by turning on the light source in step 85. Reference numeral 603 denotes a signal level between A and A ′ obtained by subtracting 601 from 602. By this calculation, the external light component is removed, and only the component 603 caused by the light source is acquired.

  7A to 9C are schematic diagrams for explaining the principle of distinguishing between a living body and an outside body (pseudo-living body) using a difference in luminance shading of an acquired image by this wavelength. As in FIG. 6, the horizontal axis represents the position, and A and A ′ correspond to A-A ′ in FIG. 1. The vertical axis indicates the output level.

  FIG. 7 is a diagram for explaining the case where the subject (finger) is a living body. Reference numeral 701 in FIG. 7A denotes an AA of an image based only on a component in which the red wavelength component acquired in step 86 is scattered inside the finger. 'Is the output level between. In FIG. 7B, reference numeral 702 denotes an output level between A and A ′ of an image by only the component in which the green wavelength component acquired in step 91 is scattered inside the finger. Reference numeral 703 denotes a value obtained by normalizing the peak value of 702 to the peak value of 701. A hatched portion indicated by 704 is a scattered light distribution difference due to a wavelength component created from the difference between 701 and 703 for the difference analysis in step 92.

  FIG. 8 is a diagram for explaining a case where a fake finger made of resin, plastic, or the like is used as a subject. Reference numeral 801 in FIG. 8A indicates that the red wavelength component acquired in step 86 is scattered inside the finger. This is the output level between AA ′ of the image by only the component. In FIG. 8B, reference numeral 802 denotes an output level between A and A ′ of an image by only the component in which the green wavelength component acquired in step 91 is scattered inside the finger. Reference numeral 803 denotes a value obtained by normalizing the peak value of 802 to the peak value of 801. A horizontal line portion indicated by 804 is a scattered light distribution difference due to a wavelength component created from the difference between 801 and 803 for the difference analysis in step 92.

  FIG. 9 is a diagram illustrating a case where a film on which a color fingerprint image in which light transmittance is adjusted to a living body is used as a subject as a fake finger. Reference numeral 901 in FIG. 9A denotes an output level between A and A ′ of an image that is obtained by only the component in which the red wavelength component acquired in step 86 is scattered inside the finger. In FIG. 9B, reference numeral 902 denotes an output level between A and A ′ of an image by only the component in which the green wavelength component acquired in step 91 is scattered inside the finger. Reference numeral 903 denotes a value obtained by normalizing the peak value of 902 to the peak value of 901. A hatched portion indicated by 904 is a scattered light distribution difference due to a wavelength component created from the difference between 901 and 903 for the difference analysis in step 92.

  In this way, in the case of a living body, the internal scattered light distribution depends on the wavelength due to the internal structure of the living body, so that the difference between images having different wavelengths is positioned as indicated by the hatched (horizontal line) portion in the range from A to A ′. It varies greatly depending on.

  This is because the central part of the finger is structurally different from the surroundings of the finger, such as the presence of bones, and the difference in the thickness of the living body between the central part of the finger and the surroundings is a factor that determines the transmittance of pigments and the like. This is because a difference in internal scattering distribution for each wavelength of incident light peculiar to a living body is given. As a result, the internal scattered light spreads over the entire finger especially in the wavelength band longer than 610 nm, such as red, so the output level difference between the finger center (M point) and the finger periphery (P point) is compared. Less. In the wavelength band where the wavelength such as green is shorter than 570 nm, the internal scattered light does not spread at the center of the finger, so the output level difference between the center of the finger (M ′ point) and the periphery of the finger (P ′ point) is large. The change curve of the output level depending on the position is defined as luminance shading. Luminance shading between finger center (M point) and finger periphery (P point) at long wavelength, finger center (M 'point) and finger periphery (P' point) at short wavelength In the luminance shading during (), the amount and curve shape vary greatly depending on the wavelength.

  On the other hand, in the case of a fake finger made of resin or plastic, the wavelength dependence of the change between the central part and the peripheral part is small. In other words, the amount of luminance shading and the shading curve shape after standardization do not change much with wavelength. This is because the influence of the thickness component is the main factor because it is made of the same component material as a living body.

  Further, when a film on which a color fingerprint image is printed is used as a fake finger, the amount of change itself between the central portion and the peripheral portion is small. In addition, the wavelength dependence of the change between the central part and the peripheral part is small. In other words, the amount of luminance shading and the shading curve shape after standardization do not change much with wavelength. This is because, unlike a living body, since it is printed on a film, the influence of the thickness component is small, and the amount of transmitted light is determined depending on the printed color component.

  In this manner, the living body and the false living body are identified by determining the difference in the luminance shading change amount according to the wavelength with the threshold set in advance in the living body detecting unit 115.

  In this way, by illuminating with light from only one place, dependence on how the incident light is scattered inside the finger occurs. From the captured image, this dependency in the acquired image plane of the finger can be read as a luminance distribution (luminance shading). Since the scattering state inside the subject changes depending on the wavelength, whether or not the subject is a living body can be identified by looking at the difference in the luminance distribution (luminance shading) depending on the wavelength.

  The light source for acquiring an image for fingerprint authentication needs to be illuminated as uniformly as possible from a plurality of locations. However, the light source for detecting the living body of the present invention may be “illuminated at a different wavelength from almost one location”. is important. The white LED is suitable for such a configuration, but not only white but also a filter having a plurality of separable wavelength bands can be used in combination with filter means.

  By acquiring the scattered light distribution information in the subject plane, three-dimensional information can be acquired from the wavelength dependence difference of the internal scattered light according to the subject shape. Provided is a highly accurate biological identification method in which the output obtained by the color of the skin and the surface state of the finger is not easily influenced. Moreover, even if the light transmittance is a thing matched to a living body like a film on which a color fingerprint image is printed, it can be detected that it is a fake finger (solution of problem α).

  The illumination means for biometric authentication selects the optimum infrared light to spread scattered light uniformly within the subject, and the second illumination means provided for identifying the living body is a scatter within the subject surface. The identification can be performed efficiently by using visible light that tends to have a difference in light distribution.

  By using visible light, an identification method with a simple configuration and method can be provided (solution of problems α and β).

  In particular, by comparing the difference between the red band, which is visible light, and the green-blue band, the visible light is efficiently separated, and by utilizing the two wavelength bands of visible light that are easy to separate, It is possible to provide an identification method with a simple configuration and method (solution of problem β).

  In order to separate visible light, a biometric authentication device that uses scattered light, which has conventionally used infrared light, is also provided with a color filter, which makes it possible to distribute scattered light in the subject surface by visible light in a simple and low-cost manner. The difference in information can be acquired (solution of problem β).

  By recognizing luminance shading components according to the finger cross-sectional shape at low frequency, one-dimensional or two-dimensional information can be acquired, and it is simple but three-dimensional of the subject that cannot be obtained with a single point of transmission difference Information is obtained, and the accuracy of living body detection is improved (solution of problem α).

  By entering light from other than the imaging surface that is closely attached to acquire a biological image, it is possible to acquire only the internal scattered light distribution information except for the reflected image from the subject surface. Thereby, living body detection accuracy improves (solution of subject alpha).

  It is arranged on the same surface as the imaging surface with respect to the living body and irradiates light from a lower part that is substantially in contact with the living body. As a result, it is possible to efficiently scatter visible light that is difficult to transmit inside the subject, and to easily obtain a luminance shading shape of the scattered light distribution of the subject that is easy to detect a living body, thereby realizing a simple and low-cost detection means. (Solution of issue β).

  External light removal means for acquiring a differential image in which the illumination light quantity of the subject illumination means is changed is provided. Thereby, although it is a simple structure, the component resulting from external light can be excluded from the scattered light inside the subject, and only the wavelength component of the light source to be used can be extracted to perform highly accurate biometric authentication and biometric detection. (Solution of issues β and γ)

  Conventionally, a biometric authentication device that uses scattered light that has used infrared light also provides a color filter to obtain a difference due to visible light components of scattered light distribution information in a subject plane by a simple and low-cost method. . Using this difference in wavelength components, it is possible to easily identify whether the living body is a pseudo-living body (solution of problems α and β).

  As a result, after removing the component caused by the external light from the scattered light inside the subject, the difference between the scattered light distribution information in the subject surface due to the visible light component is obtained, A device that can easily identify whether it is a pseudo-living body can be realized with a simple configuration (solution of problem γ).

  In addition, there are various methods for creating false living organisms that can be used to counter individual pseudo-biometric biometric detection methods. Therefore, there are several different approaches for pseudo-biometric biometric detection methods that can be used to counter this. It is desirable to combine them. In this sense, the combined use with the conventional living body detection method is one of the useful forms of the present invention.

(Second Embodiment)
Next, the configuration of the second embodiment to which the present invention is applied will be described with reference to FIGS.

  Here, as the biometric authentication device, an authentication device using an optical fingerprint sensor that reads a fingerprint by giving scattered light to the inside of the finger and bringing the finger into close contact with the imaging unit is shown.

  In the first embodiment, a white diode is used and red light and green light are separated by a filter. However, in the second embodiment, a red diode and a green diode are used, and a color filter is unnecessary.

  FIG. 10 is a block diagram showing a schematic configuration of a fingerprint authentication device as Embodiment 2 of the present invention. FIG. 11 is a schematic diagram showing an external configuration.

  The fingerprint authentication apparatus according to the present embodiment includes an image acquisition unit 101 and an authentication unit 102 as an image acquisition apparatus.

  In the image acquisition unit 101 in FIG. 10, reference numerals 103a, 103b, and 103c denote LED light sources as illumination means. Reference numeral 103a denotes a light source 1 as a first illumination unit having an infrared wavelength for acquiring a subject image for authentication. Reference numerals 103b and 103c denote a light source 2 and a light source 3 as second illumination means having a visible light wavelength for acquiring a subject image for discriminating whether the subject is a living body or a non-living body or a fake finger. Here, 103b uses a red wavelength LED, and 103c uses a green wavelength LED. Reference numeral 108 denotes a light source control unit that controls the luminance and lighting timing of each LED, and 112a, 112b, and 112c are control signal lines thereof. In the present embodiment, the light source control unit 108 serves as a wavelength separation unit, and lights each of the 103b red wavelength LED and the 103c green wavelength LED in a time-sharing manner. Acquire an image for the wavelength.

  Reference numeral 104 denotes an image sensor such as a CMOS type or a CCD type as image reading means. In this embodiment, a CMOS image sensor is exemplified. Reference numeral 105 denotes a drive pulse generator (TG unit) that controls the sampling timing of the image sensor and the ADC 107. Reference numeral 107 denotes an analog / digital converter (ADC). Reference numeral 110 denotes an analog image data signal line. Reference numeral 113 denotes a digital image data signal line. 112d and e are drive pulse signal lines sent from the drive pulse generator to the image sensor and the analog / digital converter (ADC).

  In the authentication unit 102, 114 is a difference generation unit as a comparison unit that compares red and green visible light component images acquired separately by switching the light source control unit 108. While temporarily storing a plurality of images in the frame memory 117, a difference between the images is generated. Reference numeral 114 also serves as an external light component removing unit to be described later. Reference numeral 115 denotes a living body detection unit serving as a living body detection unit for identifying whether a living body or a living body other than a living body such as a pseudo living body. The living body detection is performed by using a difference in luminance distribution in the object plane between red and green images.

  Reference numeral 116 denotes a pre-processing unit that performs image processing such as edge enhancement in order to perform feature extraction at a later stage. Reference numeral 118 denotes a feature extraction unit, and reference numeral 119 denotes a registration collation unit that registers the individual features extracted in 118 in the database or compares and collates with registered data. A database 120 stores personal data. A control unit 123 controls the operation of each unit.

  Reference numerals 124a, b, c, d, and e denote data lines for transmitting image data. Reference numeral 125 denotes a data line and a control line between the database and the registration / verification unit. Reference numeral 126 denotes a signal line that controls the operation of the light source control unit 108, and reference numeral 127 denotes a signal line that controls the operation of the drive pulse unit (TG unit) 105. Reference numerals 129a to 129e denote control lines to each unit of the authentication unit 102. Reference numeral 130 denotes a signal line for transmitting a living body detection result.

  In the schematic view of FIG. 11, FIG. 11 (a) is a cross-sectional view, and FIG. 11 (b) is a top view.

  Reference numeral 1 denotes a finger to be subjected to fingerprint authentication as a subject, which is placed in close contact with the imaging surface. Reference numerals 2a, 2b and 2c denote LED light sources as illumination means for irradiating the finger part. Among these, 2a is 103a of FIG. 10, and is the light source 1 as the 1st illumination means of the infrared wavelength for acquiring the to-be-photographed image for authentication.

  Reference numerals 2b and 2c denote 103b and 103c in FIG. 10, which are a light source 2 and a light source 2 as a second illuminating means having a visible light wavelength for acquiring a subject image for discriminating whether a living body or a non-living body is a false finger or the like. 3. Here, 2b is a red wavelength LED, and 2c is a green wavelength LED.

  Reference numeral 3 denotes an optical member having a plate shape that guides the optical difference of the uneven pattern of the fingerprint to the image sensor 4.

  Examples of the plate-shaped optical member include a thin film having a transmittance at a desired wavelength having a thickness of about several to several hundreds of micrometers, a fiber, and a microlens array.

  Reference numeral 4 denotes an image sensor serving as an image reading unit, and here, a two-dimensional CMOS image sensor. 4a indicates a pixel portion. Reference numeral 7 denotes a substrate on which an image sensor and LEDs are mounted.

  An infrared light source 1 (2a) for performing authentication is arranged around the finger, and illuminates the entire finger with infrared light that spreads from a distance slightly away. In this way, the dependence of the infrared light source on the acquired image is less dependent, and the incident infrared light spreads evenly inside the finger, and a wide range of fingerprint patterns can be read.

  On the other hand, the light source 2 (2b) and the light source 3 (2c) for discriminating whether the subject is a living body or a non-living body are arranged close to only one place directly below the finger base side. The light sources arranged close to each other may be anything as long as they emit light of a plurality of wavelengths, but here, the red LED and the green LED are arranged side by side. An image captured by illuminating the red LED and an image captured by illuminating the green LED are acquired by the image sensor at different timings. By illuminating with light from only one location, there is a dependency on how the incident light is scattered inside the finger. From the captured image, this in-plane dependence can be read as a luminance distribution (luminance shading). Since the scattering state inside the subject changes depending on the wavelength, whether or not the subject is a living body can be identified by looking at the difference in the luminance distribution (luminance shading) depending on the wavelength. Reference numerals 5a and 5b denote the incident directions of light from the LED 2b and the LED 2c to the finger 1, 5a schematically showing light of a red wavelength component and 5b of light of a green wavelength component. The lights 5a and 5b are actually incident on the finger at different timings.

  Depending on the wavelength, different scattering and transmission occurs in the finger. Reference numerals 6a and 6b denote light emission directions from the finger 1 toward the optical member 3 and the image sensor 4 after the light emitted from the LEDs 2b and 2c is scattered in the finger 1, respectively. 6a schematically shows light of a red wavelength component, and 6b schematically shows light of a green wavelength component. 6a and 6b are imaged by the pixel unit 4a.

  Next, the operation in this embodiment will be described with reference to FIG. FIG. 12 is a flowchart illustrating an image acquisition routine executed by the control unit 123 in the biometric authentication device of FIG. The basic idea of the step of removing external light and the step of performing living body detection by dividing luminance shading into red and green wavelength bands as the scattered light distribution inside the subject and comparing the distributions is the same as in the first embodiment. is there.

  In FIG. 12, when the routine is started in step 1200, all light sources are turned off in step 1201, and then the R light source (red LED light source) is turned on. In step 1202, a screen is acquired via an analog / digital converter (ADC). The acquired image is temporarily stored in the frame memory 117 via the difference generation unit 114. At this time, since the red LED light source is attached, an image of a component in which the wavelength component of the red LED is scattered inside the finger in addition to the external light component is obtained.

  Next, in step 1203, the R light source is turned off. In step 1204, one screen is acquired via the analog / digital converter (ADC) 107. The acquired image is temporarily stored in the frame memory 117 via the difference generation unit 114. At this time, since the light source is not attached, an image using only the external light component can be obtained. In step 1205, the difference generation unit 114 obtains the image in step 1202 and the difference image in step 1204.

  Thereby, since the external light component is canceled, an image of only the component in which the wavelength component of the red LED light source is scattered inside the finger is obtained.

  In step 1206, the G light source (green LED light source) is turned on. In step 1207, one screen is acquired via the analog / digital converter (ADC) 107. The acquired image is temporarily stored in the frame memory 117 via the difference generation unit 114. At this time, since the green LED light source is attached, an image of a component in which the wavelength component of the green LED is scattered inside the finger in addition to the external light component is obtained.

  Next, in step 1208, the G light source is turned off. In step 1209, the difference generation unit 114 obtains the image in step 1207 and the difference image in step 1204.

  Thereby, since the external light component is canceled, an image of only the component in which the wavelength component of the green LED light source is scattered inside the finger is obtained.

  Next, in step 1210, the difference generation unit 114 in FIG. 10 generates the image in step 1205 and the difference image in step 1209. The living body detection unit 115 analyzes the difference and detects whether it is a living body.

  In step 1211, when the control unit 123 determines that the living body detection unit receives a detection result signal, the living body determination flag is output in step 1212, and the process proceeds to step 1215. In step 1215, the light source 1 (103a) which is the infrared light source of FIG. 10 is turned on. In step 1216, an authentication image is acquired. Next, in step 1217, the light source 1 (103a) which is an infrared light source is turned off. In step 1218, an image of only the infrared light component of the light source 1 is obtained by taking the difference between the image obtained in step 1216 and the image obtained in step 1204. Thereafter, the routine ends at step 1219.

  In step 1211, when the control unit determines that the living body is detected as a fake living body (not a living body) in response to the detection result signal from the living body detecting unit, a false living body determination flag is output in step 1220, and the routine is terminated in step 1219. .

  In this manner, the living body detection unit 115 determines a difference in luminance shading caused by the scattered light distribution inside the finger by the living body detection unit 115, thereby distinguishing the living body from the false living body.

  In this way, by illuminating with light from only one place, dependence on how the incident light is scattered inside the finger occurs. From the captured image, this dependency in the acquired image plane of the finger can be read as a luminance distribution (luminance shading). Since the scattering state inside the subject changes depending on the wavelength, whether or not the subject is a living body can be identified by looking at the difference in the luminance distribution (luminance shading) depending on the wavelength.

  The light source for acquiring an image for fingerprint authentication needs to be illuminated as uniformly as possible from a plurality of locations. However, the light source for detecting the living body of the present invention may be “illuminated at a different wavelength from almost one location”. is important. A package in which LED elements of two colors, such as green and red, are contained in one package is suitable for such a configuration, but it is the same even if two green LEDs and red LEDs in different packages are arranged close to each other. The effect is obtained.

  By obtaining the scattered light distribution information in the subject plane by the comparison unit, three-dimensional information can be obtained from the wavelength dependence difference of the internal scattered light according to the subject shape. Provided is a highly accurate biological identification method in which the output obtained by the color of the skin and the surface state of the finger is not easily influenced. Moreover, even if the light transmittance is a thing matched to a living body like a film on which a color fingerprint image is printed, it can be detected that it is a fake finger (solution of problem α).

  The illumination means for biometric authentication selects the optimum infrared light to spread scattered light uniformly within the subject, and the second illumination means provided for identifying the living body is a scatter within the subject surface. The identification can be performed efficiently by using visible light that tends to have a difference in light distribution. By using visible light, an identification method with a simple configuration and method can be provided (solution of problems α and β).

  A plurality of light sources having different wavelengths are arranged in the vicinity. Thereby, it is possible to obtain a plurality of scattered light distributions in the subject surface when light is incident from the substantially the same position toward the inside of the subject while eliminating the difference in the light source position when the wavelength is changed. Therefore, the living body detection accuracy is improved with a simple configuration (solution of problems α and β).

  In particular, the visible light is efficiently separated by comparing the difference between the red band, which is visible light, and the green to blue band (solution of problem β).

  Recognizing luminance shading components according to the finger cross-sectional shape at a low frequency, one-dimensional or two-dimensional information (obtained by collecting one-dimensional information on a plurality of cross-sections) can be easily acquired. However, it is possible to obtain three-dimensional information of a subject that cannot be obtained with a single point of transmittance difference, and improve the accuracy of living body detection with a simple configuration and method (solution of problems α and β).

  By entering light from other than the imaging surface that is closely attached to obtain a biological image, it is possible to obtain only the internal scattered light distribution information with a simple configuration and method, excluding the reflected image from the subject surface (problems α and β Solution).

  It is arranged on the same surface as the imaging surface with respect to the living body and irradiates light from a lower part that is substantially in contact with the living body. This makes it possible to efficiently scatter visible light that is difficult to transmit inside the subject, and to obtain a luminance shading shape of the scattered light distribution of the subject that is easy to detect a living body with a simple configuration and method (problems α and β Solution).

  External light removal means for acquiring a differential image in which the illumination light quantity of the subject illumination means is changed is provided. Thereby, although it is a simple structure, the component resulting from external light can be excluded from the scattered light inside the subject, and only the wavelength component of the light source to be used can be extracted to perform highly accurate biometric authentication and biometric detection. (Solution of issues β and γ)

It is a block diagram which shows the typical structure of the fingerprint authentication apparatus in Embodiment 1 of this invention. It is the schematic which shows the typical structure of the fingerprint authentication apparatus in Embodiment 1 of this invention. It is a circuit diagram of an image sensor in an embodiment of the present invention. It is a circuit diagram of a pixel part of an image sensor in an embodiment of the present invention. It is a flowchart explaining the image acquisition routine in Embodiment 1 of this invention. It is a figure explaining the process of the external light removal in embodiment of this invention. It is the 1st figure explaining the process of living body detection in the embodiment of the present invention. It is a 2nd figure explaining the process of the biological body detection in embodiment of this invention. It is a 3rd figure explaining the process of the biological body detection in embodiment of this invention. It is a block diagram which shows the typical structure of the fingerprint authentication apparatus in Embodiment 2 of this invention. It is the schematic which shows the typical structure of the fingerprint authentication apparatus in Embodiment 2 of this invention. It is a flowchart explaining the image acquisition routine in Embodiment 2 of this invention.

Explanation of symbols

DESCRIPTION OF SYMBOLS 101 Image acquisition part 102 Authentication part 103a, 103b, 103c Light source 104 Image pick-up element 105 Drive pulse generation part (TG part)
106 signal switching unit 107 ADC (analog / digital converter)
108 Light source control unit 110a, 110b, 110c Analog image data signal line 112a, 112b, 112c Control signal line 113 Digital image data signal line 114 Difference generation unit 115 Living body detection unit 116 Preprocessing unit 117 Frame memory 118 Feature extraction unit 119 Registration / collation unit 120 Database 123 Control unit 124a, 124b, 124c, 124d, 124e Data line 125 Data line and control line 126, 127 Control signal line 128 Switching signal line 129a, 129b, 129c, 129d, 129e Control line 130 Signal line

Claims (14)

A biometric authentication device that obtains an image for biometric authentication using scattered light inside a subject generated by the illumination unit and the illumination unit,
Comparing means for acquiring scattered light distribution information in the subject at different wavelengths and comparing the scattered light distribution information;
A living body detecting means for performing living body detection according to the result of the comparison;
A biometric authentication device comprising: The biometric authentication device according to claim 1,
The illumination means includes
First illumination means having an infrared wavelength for biometric authentication;
A second illumination means having a visible light wavelength for comparing scattered light distribution information in the subject;
A biometric authentication device comprising: The biometric authentication device according to claim 1 or 2,
The second illumination means has two light sources having different wavelengths arranged in the vicinity of each other,
The biometric authentication apparatus further includes means for sequentially switching and emitting the light sources. The biometric authentication device according to any one of claims 1 to 3,
The second illuminating means emits at least light in a wavelength band longer than 610 nm (about 610 to 800 nm) and light in a wavelength band shorter than 570 nm (about 400 to 570 nm). Biometric authentication device. The biometric authentication device according to any one of claims 1 to 3,
The biometric authentication device, wherein the second illumination unit emits red light and green light. The biometric authentication device according to claim 1, 2, 4, or 5,
The biometric authentication apparatus further comprising an image reading unit, and the image reading unit includes a wavelength separation unit. The biometric authentication device according to any one of claims 1 to 6,
A biometric authentication device comprising means for detecting luminance shading caused by the shape of the subject as scattered light distribution information in the subject. The biometric authentication device according to any one of claims 1 to 7,
The biometric authentication apparatus includes an image reading unit including an optical member that is in close contact with the subject and an image sensor that is in close contact with the optical member,
The biometric authentication apparatus, wherein the second illuminating unit is disposed at a position where the subject is illuminated from another region of the imaging surface of the optical member to be closely attached to acquire the subject image.   9. The biometric authentication apparatus according to claim 8, wherein the second illumination unit is arranged on the same surface as the imaging surface of the optical member with respect to the subject, and emits light from a lower part that is substantially in contact with the subject. A biometric authentication device arranged at a position. A biometric authentication apparatus comprising: an illuminating unit; and an image reading unit that acquires an image for biometric authentication using scattered light inside the subject generated by the illuminating unit,
A biometric authentication device, comprising: an external light removing unit that acquires a difference image between an image when the illumination unit emits light and an image when the light does not emit light. The biometric authentication device according to claim 10,
The image reading means has a pixel group provided with means for separating visible light,
The biometric authentication device includes a switching unit that switches a wavelength band of an image acquired by switching the pixel group, and a unit that compares the acquired images with each other. A biometric authentication apparatus comprising: an illuminating unit; and an image reading unit that acquires an image for biometric authentication using scattered light inside the subject generated by the illuminating unit,
The image reading means has a pixel group provided with wavelength separation means,
The biometric authentication device includes a switching unit that switches a wavelength band of an image acquired by switching the pixel group, and a unit that compares the acquired images with each other. Obtaining a subject image of a first wavelength component;
Obtaining a subject image of a second wavelength component;
A step of acquiring a subject image having a first wavelength component and a step of comparing scattered light distribution information in a plurality of subject planes having different wavelengths from the subject image obtained in the step of acquiring a subject image having a second wavelength component; ,
Determining whether or not the subject is a living body according to a result of comparison;
A living body detection method characterized by comprising: The living body detection method according to claim 13,
Illuminating the subject with a first light quantity;
Illuminating the subject with a second amount of light;
Illuminating the subject with a first light amount and removing an external light component by obtaining a difference from an image obtained by illuminating the subject with a second light amount;
Using the image obtained from the step of illuminating the subject with the first light amount and the step of removing the external light component by obtaining a difference from the image obtained in the step of illuminating the subject with the second light amount. And detecting a living body of the subject.

Images