JP2004326820A - Fingerprint input device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a fingerprint input device which can appropriately judge the authenticity of a finger used for input. <P>SOLUTION: A finger placed on an image sensor 3 is irradiated with a light emitting diode 4 and the finger is substantially made to emit light by scattering the light in the finger. A swelling line part of a fingerprint where high luminance is detected owing to the contact with the image sensor 3 and a trough line part of the fingerprint where low luminance is detected owing to a gap with the image sensor 3 are identified and the fingerprint is detected. Thus, a regular light emitting diode 4 can be used instead of a laser diode and the manufacture cost of the whole device is reduced. As light emitting intensity and the oscillation frequency of the light emitting diode 4 are made to time-sequentially change, the change of the light emitting state of the light emitting diode 4 is detected from an image obtained through the image sensor 3, it is judged whether the real time input of the fingerprint is appropriately performed or not based on whether the change is correlated with the time sequential change of the light emitting state of a light source. <P>COPYRIGHT: (C)2005,JPO&NCIPI

Description

  The present invention relates to an improvement in a fingerprint input device, and more particularly, to an improvement in cost reduction of a hardware structure required for determining whether a finger used for input is a living body, and an improvement in electronic and physical characteristics. The present invention relates to an improvement for achieving enhanced security against forged fingerprint input.

  As means for determining whether the finger used for input of the fingerprint input device is a living body or a counterfeit product, the finger is transmitted and irradiated with laser light from a laser diode, and the output light is speckled. For example, Japanese Patent Application Laid-Open Publication No. H11-163873 discloses a device that performs determination based on whether or not light noise is included.

  In addition, the fingertip is pressed against the glass of the input surface to optically identify the ridges (convex) and the valleys (concave) of the fingerprint and read the image. There is also known an apparatus that determines whether a finger is a living body or a counterfeit by detecting a temporal change in a sensor output that occurs.

  However, the technique disclosed in Patent Document 1 requires the use of speckle light, and the light source for illuminating the finger is limited to an expensive laser diode for irradiating coherent light. There is.

  In addition, in the case of using a change in the degree of adhesion of the fingertip to judge the authenticity of the finger, malfunction is assumed when a forged product made of an elastic material whose degree of adhesion changes due to the pressing force is used, and Since the change in the sensor output caused by the change in the degree of adhesion due to sweating is exponential (lawful), electronic forgery of the sensor signal using an unjust device may be a concern.

Japanese Patent Application No. 63-154962 JP-A-2-5190 JP-A-7-308308 JP-A-3-87981

  Therefore, an object of the present invention is to provide a fingerprint input device that solves the above-mentioned drawbacks of the prior art, does not require a particularly expensive light source, and can accurately determine the authenticity of a finger used for input. It is in. Further, it is also possible to provide a fingerprint input device capable of ensuring necessary and sufficient security even when the forgery of the sensor signal is difficult and the sensor signal is stolen and reproduced and input. Is part of that challenge.

  The present invention provides an image sensor on which a finger to be inspected is placed, a light source that irradiates the finger placed on the image sensor and diffuses light therein to substantially emit light, and a predetermined light source. An image acquisition unit that activates an image sensor for each cycle to acquire an image of a finger; and a living body determination unit that determines the presence or absence of blood flow based on a chronological change in the image acquired by the image acquisition unit. The above object has been achieved by a configuration characterized by the following.

As described above, the finger is made to emit light by diffusing the light inside, and the ridges (fingerprints where relatively high brightness is detected by the contact with the image sensor) of the fingerprint and the valley are in close contact with the image sensor. Since it is configured to discriminate a line portion (a concave portion of a fingerprint in which relatively low luminance is detected due to the influence of a gap generated between the image sensor), there is a difference from a conventional fingerprint input device using speckle light. Therefore, there is no need to consider the wavelength of the light source. Therefore, a normal light emitting diode can be used, which is advantageous for reducing the manufacturing cost of the device.
In addition, by determining the presence or absence of blood flow by evaluating the chronological change of the image acquired at every predetermined period by the image acquisition unit by the living body determination unit, it is determined whether the finger of the subject is a living body. Since identification is performed, it is possible to prevent a mistake in authenticity determination of a forged finger with a high probability.

  Furthermore, in addition to the above configuration, a light emission control unit that outputs a light emission control signal that changes the light emission state of the light source in time series, and detects a change in the light emission state of the light source from an image acquired by the image acquisition unit, and detects the change And real-time input determination means for determining whether or not the light emission state of the light source is correlated with a time-series change. It is also possible to adopt a configuration in which a function for removing an effect caused by a change in the state is added.

According to this configuration, the light emitting state of the light source that irradiates the finger changes in a time-series manner, and at the same time, the change in the light emitting state of the light source is reflected in the data of the image acquired at predetermined intervals by the image acquiring unit. Then, the real-time input determination unit detects a change in the light emitting state of the light source from the image obtained by the image obtaining unit, and determines whether the change is correlated with a time-series change in the light emitting state of the light source. judge. This makes it possible to determine whether or not the fingerprint image input to the fingerprint input device is an image input in real time (real time). It will be possible to prevent with a very high probability.
Also, on the side of the living body determination unit, a function of removing an effect caused by a change in the light emission state of the light source from the image acquired by the image acquisition unit is added, so that regardless of the change in the light emission state of the light source, In addition, the presence or absence of blood flow can be determined by accurately evaluating images acquired at predetermined intervals, and it can be determined whether or not the finger to be inspected is a living body.

  By changing the light emission intensity of the light source with the light emission control signal, the light emission state of the light source can be changed in time series.

  In this case, the real-time input determination unit detects a change in the light emission intensity of the light source from the image acquired by the image acquisition unit, and determines whether the change is correlated with a time-series change in the light emission intensity of the light source. Will be determined. Then, the living body determination unit evaluates the image acquired at predetermined intervals by removing an effect caused by a change in the light emission intensity of the light source from the image acquired by the image acquisition unit, and determines whether there is a blood flow. .

  Further, the light emission state of the light source can be changed in time series by changing the oscillation frequency of the light source with the light emission control signal.

  In this case, the real-time input determination unit detects a change in the oscillation frequency of the light source from the image acquired by the image acquisition unit, and determines whether the variation is correlated with a time-series change in the oscillation frequency of the light source. Will be determined. In addition, the living body determination unit evaluates images acquired at predetermined intervals by removing an effect caused by a change in the oscillation frequency of the light source from the image acquired by the image acquisition unit, and determines the presence or absence of blood flow. .

  Further, the light emission state of the light source may be changed in time series by changing the light emission intensity and the oscillation frequency of the light source with the light emission control signal.

  In this case, the real-time input determination unit detects a change in the light emission intensity and the oscillation frequency of the light source from the image acquired by the image acquisition unit, and the change corresponds to a time-series change in the light emission intensity and the oscillation frequency of the light source. It is determined whether or not there is a correlation. In addition, the living body determination unit evaluates images acquired at predetermined intervals by removing an influence caused by a change in the light emission intensity and the oscillation frequency of the light source from the image acquired by the image acquisition unit, and evaluates the presence or absence of blood flow. Is determined.

  Further, the light emission control unit may be provided with a signal generation function of randomly generating a light emission control signal.

When this configuration is applied, it is extremely difficult to externally predict how the light emission state of the light source changes in a time series, so that a sensor signal generated by a legitimate method is plagiarized and used later. It is possible to sufficiently cope with fraud such as reproduction input.
For example, even if the stolen sensor signal was reproduced and input again later when the fingerprint input device was properly used in the past, in most cases, the light emitting state of the light source recorded in the reproduced and input signal was often changed. Since the chronological change is different from the chronological change of the light emitting state of the light source due to the newly generated light emission control signal at the time of this input operation, it is real time that the current fingerprint input operation is incorrect. It is accurately detected by the input determining means.

  Preferably, the light emission control unit is provided on a main device side connected to the fingerprint input device main body.

  By providing the light emission control unit for generating the light emission control signal on the main device side, not on the fingerprint input device main body, it becomes difficult for a third party to steal the light emission control signal, and the security against unauthorized operation is further enhanced. Will be.

The fingerprint input device of the present invention uses a light source that irradiates a finger placed on an image sensor and diffuses light therein to emit light substantially, and is relatively high due to contact with the image sensor. Since the fingerprint is detected by identifying the ridge portion of the fingerprint where the luminance is detected and the valley portion of the fingerprint where the relatively low luminance is detected due to the gap between the image sensor and the light source, There is no need to consider wavelengths. Therefore, it is not necessary to use a coherent light laser diode as in the conventional fingerprint input device using speckle light, and a normal light emitting diode can be used, and the manufacturing cost of the entire device can be easily reduced. can do.
In addition, the image sensor is activated at predetermined intervals to acquire a fingerprint image, and the living body determination unit evaluates a time-series change of the image, and determines whether or not the finger to be inspected is a living body based on the presence or absence of blood flow. Therefore, it is possible to prevent a mistake in authenticity determination of a forged finger with a high probability.

  Furthermore, the light emission state of the light source, that is, the light emission intensity and the oscillation frequency are changed in a time series according to the light emission control signal from the light emission control unit, and the light emission state of the light source is also changed in a time series from the image acquired by the image acquisition means. And whether the change is correlated with a time-series change in the light emission state of the light source to determine whether or not a real-time fingerprint input using a valid fingerprint input device is being performed. With this configuration, it is possible to prevent a mistake in authenticity determination for electronic forgery of the sensor signal with an extremely high probability.

  In addition, the light-emitting intensity of the light source and the time-series change of the oscillation frequency are randomly generated each time by the signal generation function of the fingerprint input device. It is extremely difficult to electronically falsify input data by knowing in advance the time-series change of the oscillation frequency. Further, even if a sensor signal generated by a legitimate method is stolen and reproduced and input later, this can be accurately detected as an illegal input operation.

  In addition, since the light emission control unit is not provided in the fingerprint input device main body but provided on the side of the main device connected to the fingerprint input device main body, it is extremely difficult for a third party to steal the light emission control signal itself. Therefore, security against unauthorized operations is further enhanced.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings. FIGS. 1 and 2 are functional block diagrams showing an embodiment in which the present invention is applied to a fingerprint input device used for restricting the use of a workstation, a personal computer, or the like. 2 shows a portion of the fingerprint input device main body 1 connected to a workstation, a personal computer, or the like. FIG. 2 shows a configuration of a main device 2 such as a workstation or a personal computer related to the functions of the fingerprint input device main body 1. The figure includes a part.

  As shown in FIG. 1, an image sensor 3 for placing a finger 15 to be inspected is provided on the upper surface of a housing of the fingerprint input device main body 1. And a light emitting diode 4 serving as a light source for irradiating a finger 15 placed on the image sensor 3. The image sensor 3 is a two-dimensional image sensor having a large number of light receiving elements 5 arranged two-dimensionally, and the surface that comes into contact with the finger 15 is mechanically and electrically protected by a light-transmitting protective film 6. The buffer 7 is a part that captures an image from the image sensor 3 and functions as a part of an image acquisition unit.

  The driver 8 receives an external command, that is, a light emission control signal from the main device 2 in this case, and controls the light emission intensity of the light emitting diode 4, the oscillation frequency, and the like. That is, in this embodiment, the light emission control unit for outputting the light emission control signal is provided not on the fingerprint input device main body 1 but on the main device 2 side.

As shown in FIG. 2, the fingerprint input device main body 1 is connected to the main device 2 via an interface 9, and the CPU 10 and the like of the main device 2 constitute a light emission control unit of the fingerprint input device. . The CPU 10 also functions as a living body determination unit and a real-time input determination unit of the fingerprint input device.
Each of the ROM 11, the RAM 12, the external storage device 13, and the manual data input device 14 is a component on the main device 2 side. Among them, the ROM 11 is also a storage unit for storing a plurality of light emission control signals, and emits light. It also serves as a part of the signal generation function realizing means in the control unit.

  In this embodiment, both the light emission intensity and the oscillation frequency of the light emitting diode 4 are changed in time series by the light emission control signal. The ROM 11 stores, for example, (P1, T1), (P2, T2), .., (Pi, Ti),..., (Pn, Tn). Here, Pi is a value indicating the light emission intensity, and Ti is a time corresponding to the time during which the light emission intensity Pi is continued, that is, a value corresponding to the reciprocal of the oscillation frequency.

  The CPU 10 as a light emission control unit receives a start signal from the fingerprint input device main body 1, randomly selects one light emission control signal from a plurality of light emission control signals, and configures this light emission control signal (P1, T1), (P2, T2),..., (Pi, Ti),..., (Pn, Tn) are read out in the order of i = 1 to n, and the emission intensity of Pi is sequentially reduced to Ti seconds. By maintaining only this, the light emission intensity and the oscillation frequency of the light emitting diode 4 are changed in time series.

  Here, only the data strings (P1, T1), (P2, T2), ..., (Pi, Ti), ..., (Pn, Tn) constituting one light emission control signal are shown in a simplified manner. In practice, however, a plurality of data strings having the same data arrangement and having different data item values, that is, different light emission intensities and light emission durations, are stored in the ROM 11 by the number of light emission control signals. ing.

  It is desirable that the minimum value of the light emission intensity Pi be set at least to such a level as to enable fingerprint identification processing by the image sensor 3, and the value of the oscillation frequency Ti is determined by the brightness of the image due to blood pulsation. In order to prevent erroneous recognition due to the change, a value different from the frequency of blood pulsation should be set.

  Next, a fingerprint input process using the fingerprint input device body 1 will be described with reference to the flowcharts of FIGS.

  First, the CPU 10 reads an output from the image sensor 3 via the interface 9 and the buffer 7 (step s1), and determines whether a deviation between the value and a value detected in the previous processing cycle exceeds a predetermined value. That is, it is determined whether the finger 15 is placed on the image sensor 3 and the output from the image sensor 3 fluctuates (step s2).

  If the result of this determination is false, it means that the output from the image sensor 3 has not fluctuated, that is, the finger 15 has not been placed, so that the CPU 10 outputs the output value read this time. After being temporarily stored in the register as the comparison target value to be used in the next cycle (step s3), the process returns to step s1. Unless the finger 15 is placed on the image sensor 3 and the output does not significantly fluctuate, the processes of steps s1 to s3 are repeatedly executed in the same manner as described above.

  Then, when it is confirmed by the determination processing in step s2 that the finger 15 has been placed on the image sensor 3 and the output has fluctuated, the CPU 10 as a signal generation function realizing means uses a random number function or the like to perform a predetermined operation. A random number is randomly generated within the range of the upper limit and the lower limit of the above, and a data string of the light emission control signal corresponding to the random number is selected from the ROM 11 (step s4). For example, if there are ten sets of light emission control signals, the above-described upper and lower limits are divided into ten equal parts, and one set of light emission control signals is associated with each of them, and the generated random number is It is possible to randomly select only one set of data strings of the light emission control signal depending on which divided range belongs.

  Next, the CPU 10 as the light emission control unit initializes the value of the index i to 1 and reads the first data (Pi, Ti) corresponding to the index i = 1 from the data sequence of the light emission control signal of the ROM 11 selected in the process of step s4. ) Is read, and a light emission control signal having an output intensity Pi is output to the driver 8 to start lighting the light emitting diode 4 (step s5). Then, the CPU 10 resets and restarts the timer T1 for measuring the elapsed time after the start of light emission and the timer T2 for measuring the light emission time based on each data of the data string of the light emission control signal, and temporarily stores the timer. Is initialized (step s6).

  Next, the CPU 10 constituting a part of the image acquisition means reads the image of the image sensor 3 via the buffer 7 (step s7), calculates an average brightness P of the image (step s8), and obtains the value P Is additionally stored in the data file for temporary storage in the form of (P, T1) together with the elapsed time T1 after the start of light emission (step s9).

Then, the CPU 10 as a living body determination unit removes a change in the light emission state of the light emitting diode 4, that is, an effect caused by the intensity of the light emission intensity, from the image read in the process of step s7, and obtains the ridge of the fingerprint from the relative contrast ratio. The part and the valley part are identified and detected (step s10). As means for removing the influence caused by the intensity of the light emission of the light emitting diode 4 from the read image, for example, the average brightness of the whole image is obtained, and the brightness (contrast) of each part of the fingerprint is determined based on the brightness of the background. Concentration).
Since the finger 15 placed on the image sensor 3 emits light substantially by scattering light from the light emitting diode 4 therein, as shown in FIG. The ridge portion directly in contact with is photographed brightly, and the valley line portion formed with the image sensor 3 by the depression of the fingerprint is photographed relatively dark.

Next, the CPU 10 as the living body determination means compares the density of each part of the fingerprint thus obtained with the fingerprint stored in the previous cycle data storage register, that is, the density of each part of the fingerprint read in the processing of the previous cycle (step s11). Then, it is determined whether or not there is a blood flow, that is, whether the finger 15 on the image sensor 3 is a genuine finger with blood or a counterfeit (step s12).
If there is a blood flow, the movement of red blood cells and hemoglobin changes the light and darkness of the light propagation path at each position inside the finger, so as described above, after removing the effect caused by the intensity of the light emission intensity from the image, By comparing the change in the brightness of the minute part of each part of the image with the data of the corresponding part photographed in the previous cycle, it is possible to determine the presence or absence of blood flow, that is, whether or not a living body.

  Here, when the determination result of step s12 is false, that is, when the density of the image of each part of the fingerprint is the same as that of the previous cycle and no blood pulsation is detected, the image sensor 3 Since the placed finger 15 means a counterfeit product, the CPU 10 as the living body determination means rejects the input operation with the finger 15 and returns to the process of step s1, where the proper placement of the finger is performed. Return to the initial standby state.

  On the other hand, if the result of the determination in step s12 is true, and at least this finger 15 is confirmed to be a living body finger, the CPU 10 determines whether the finger 15 has been removed this time from the influence of the intensity of the light emission intensity. Is updated and stored in the previous cycle data storage register as comparison data used in the comparison process of step s11 in the next cycle (step s13).

  Next, the CPU 10 as the light emission control unit determines whether or not the measured time of the timer T2 for measuring the light emission time has reached the light emission time Ti determined by the light emission control signal (Pi, Ti) (step s14). ).

  If the measurement time of the timer T2 has not reached the light emission time Ti, the CPU 10 as the light emission control unit retains the current light emission intensity Pi and shifts to the process of step S7 again. The processing of s7 to step s14 is repeatedly executed, and data of the relationship (P, T1) between the average value P of the brightness of the image and the elapsed time T1 after the start of light emission is additionally stored in a data file for temporary storage. In parallel with this, the fingerprint detection processing and the living body detection processing are repeatedly executed.

  When it is confirmed in the determination processing in step s14 that the measurement time of the timer T2 has reached the light emission time Ti while repeatedly performing such processing, the CPU 10 as the light emission control unit increments the value of the index i by one. Then, it is determined whether or not the current value of the index i exceeds the number n of the components of the data of the currently selected light emission control signal (step s16).

  Here, if the current value of the index i does not exceed the number n of data, it means that it is necessary to read the next data of the data string constituting the light emission control signal and keep the light emitting diode 4 lit. Therefore, in this case, the CPU 10 as the light emission control unit resets and restarts the value of the timer T2 again to start measuring the light emission time based on the next data (step s17), and the value is incremented in the processing of step s15. The next data (Pi, Ti) is read from the currently selected light emission control signal in accordance with the value of the index i, and the magnitude of the output intensity Pi is switched based on the value Pi to start lighting the light emitting diode 4 ( Step s18), the process returns to step S7, and the same process as above is repeatedly executed.

  By this processing operation, both the light emission intensity and the oscillation frequency of the light emitting diode 4 are chronologically (P1, T1), (P2, T2), ..., (Pi, Ti), ..., (Pn, Tn). An example of a time-series change in the light emitting state of the light emitting diode 4 is shown by a thick solid line in FIG.

  Then, while such a switching process is repeated, finally, the current value of the index i exceeds the number n of data, and the light emission control by the last data (Pn, Tn) of the selected light emission control signal is completed. When it is detected in the determination process in step s16 that the light emission control signals (P1, T1), ( (P2, T2),..., (Pi, Ti),..., (Pn, Tn), the change in the light emission intensity Pi of the light emitting diode 4 (the thick solid line diagram in FIG. The correlation with the change in the average value (P, T1) of the brightness of the image stored in the data file (the thin solid line diagram in FIG. 5) is calculated (step s19), and the correlation is set in advance. Greater than the reference value Squid whether, that is, following the change of the emission intensity of the light emitting diodes 4 to determine whether the average value of the brightness of the image is changed (step s20).

  Here, the change in the average brightness P of the image caused by the pulsation of blood is sufficiently smaller than the change in the average brightness P of the image caused by the change in the light emission intensity Pi. Since the values of the light emission times Ti of 1 to n are also set to values different from the blood pulsation, even if the average value P of the brightness of the image slightly changes due to the blood pulsation, the image is not changed. Whether the light input to the sensor 3 accurately responds to the change in the light emission control signal, that is, between the change in the light emission intensity of the light emitting diode 4 and the change in the average value of the image brightness. It is possible to accurately determine whether or not there is a sufficient correlation.

  Then, when a sufficient correlation is recognized between the change in the light emission intensity of the light emitting diode 4 and the change in the average value of the brightness of the image, the CPU 10 as the real time input determination means is selected this time. It is determined that an appropriate fingerprint input operation has been performed under the control of the light emission control signal, and an appropriate signal indicating that an appropriate input operation has been performed in addition to the detected fingerprint data is provided as a fingerprint collation processing routine. (Step s21), all the processes related to one fingerprint input are completed, and the process again enters a standby state waiting for a fingerprint input operation. Since the process related to the fingerprint verification itself is already known, the description is omitted here.

  If a sufficient correlation is not found in the discrimination processing in step s20, the CPU 10 as the real-time input determination means performs an unauthorized operation such as inputting a forged signal using an unauthorized data reproducing device. In addition to the data of the detected fingerprint, an improper signal indicating that an unauthorized operation has been performed is passed to the fingerprint collation processing routine (step s22), and processing relating to one fingerprint input is performed. Is completed, and the process again enters a standby state waiting for a fingerprint input operation.

  As described above, a plurality of light emission control signals having different light emission intensities and oscillation frequencies of the light emitting diode 4 are stored in the ROM 11, and the light emission control signal is randomly selected from the plurality of light emission control signals to change the light emission state of the light emitting diode 4 in time series. Although an example of the process of changing the light emission intensity has been described, a plurality of light emission control signals in which only the light emission intensity of the light emitting diode 4 is changed without changing the oscillation frequency (the value of each Ti), or the light emission intensity of the light emitting diode 4 is changed. A plurality of light emission control signals which differ only in the oscillation frequency (the value of each Ti) without changing them may be stored in the ROM 11, and the light emission control signal may be randomly selected from the stored light emission control signals.

  Even when only one light emission control signal is stored in the ROM 11, if the light emission intensity and the oscillation frequency (the value of each Ti) of the light emitting diode 4 are set in an intricate manner, an unauthorized data reproducing apparatus may be used. Thus, it is possible to sufficiently cope with the input of a forged signal.

  In the above-described embodiment, the light emission control unit of the fingerprint input device is configured by the CPU 10 and the ROM 11 of the main device 2 and the real-time input determination means is configured by the CPU 10 of the main device 2 to obtain illegal data reproduction. Although strong security is provided for the input of a forged signal using a device or the like, technically, a dedicated CPU is provided on the fingerprint input device main body 1 side to serve as a real-time input determination unit. Alternatively, a light emission control unit and a real-time input determination unit may be used.

FIG. 3 is a functional block diagram illustrating a fingerprint input device main body. FIG. 3 is a functional block diagram showing a part of the configuration of the main device in relation to the function of the fingerprint input device. It is the flowchart which showed the outline of the fingerprint input process using the fingerprint input device main body. It is a continuation of the flowchart showing the outline of the fingerprint input process. FIG. 4 is a conceptual diagram showing an example of a time-series change of a light-emitting state Pi of a light-emitting diode and an example of a change of an average brightness P of an image corresponding to the time-series change.

Explanation of reference numerals

DESCRIPTION OF SYMBOLS 1 Fingerprint input device main body 2 Main device 3 Image sensor 4 Light emitting diode (light source)
5 Light-receiving element 6 Light-transmitting protective film 7 Buffer (part of image acquisition means)
8 driver 9 interface 10 CPU (light emission control unit, living body judgment unit, real time input judgment unit, signal generation function realization unit)
11 ROM
12 RAM
13 External storage device 14 Manual data input device (CRT / MDI)
15 fingers

Claims (5)

An image sensor that is in close contact with the finger at the time of detection and captures light emitted by the finger itself,
From the image captured by the image sensor, based on a relative brightness ratio, identification means for identifying the ridge and valley of the fingerprint of the finger,
A fingerprint input device comprising: a living body determination unit configured to detect presence / absence of blood flow in the finger based on a time-series change in brightness of an image captured by the image sensor.   2. The fingerprint input device according to claim 1, wherein the living body determination unit detects a change in brightness of a minute portion of an image captured by the image sensor.   The identification means, of the image captured by the image sensor, a portion where relatively high brightness is detected as the ridge portion of the fingerprint, and a portion where relatively low brightness is detected is the portion of the fingerprint. 2. The fingerprint input device according to claim 1, wherein each of the valley lines is detected.   The fingerprint input device according to claim 1, wherein the image sensor has a light receiving element that is in close contact with the finger at the time of detection.   2. The fingerprint input device according to claim 1, wherein a protective film is provided on a surface of the image sensor that contacts the finger.

Images