JP4269654B2 - Fingerprint authentication apparatus and method, and authentication apparatus and method - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention uses personal biological characteristics and behavioral characteristics to authenticate a person who authenticates the person and rejects a person other than the person (or substance). It is related with the authentication apparatus which performs.
[0002]
[Prior art]
2. Description of the Related Art There is a fingerprint authentication device that performs personal authentication by identifying a spiral pattern (fingerprint) formed on the surface of an inner portion on the tip side of a first joint of a human finger. The fingerprint authentication apparatus is provided with a fingerprint sensor, and the fingerprint is detected by the fingerprint sensor. As the fingerprint sensor, for example, a capacitance detection type fingerprint sensor that detects the fingerprint by sensing the capacitance of each point on the detection surface when the inner part of the tip of the finger is placed on the detection surface (for example, Patent Documents) 1), a pressure detection type fingerprint sensor that detects the fingerprint by sensing the pressure of each point on the detection surface when the inner portion of the finger tip is placed on the detection surface, or the inner portion of the finger tip There is a photodetection type fingerprint sensor that detects a fingerprint by sensing reflected light from each point on the detection surface when placed on the detection surface. The fingerprint authentication device performs personal authentication by collating a fingerprint detected by these sensors with a pre-registered fingerprint.
[0003]
[Patent Document 1]
JP-A-11-197135
[0004]
[Problems to be solved by the invention]
By the way, in the fingerprint authentication device, the detection accuracy of the fingerprint sensor is limited, and the detection output of the fingerprint sensor varies from time to time even for fingerprints of the same person due to differences in environment, situation, or condition of the person. For this reason, it is difficult to perform identity authentication with 100% accuracy.
[0005]
However, the fingerprint authentication apparatus is always required to improve accuracy in order to bring the authenticity of the personal authentication closer to 100% due to the essential request for its function. That is, in the fingerprint authentication device, it is always required to improve the accuracy of authenticating the person's fingerprint and rejecting the person's fingerprint (or substance).
[0006]
The present invention has been proposed in view of such a situation, and can be used together with fingerprint authentication to improve the accuracy of personal authentication, an authentication method, and a fingerprint to which the authentication device is applied. An object is to provide an authentication device and a fingerprint authentication method.
[0007]
[Means for Solving the Problems]
In the present invention Affect The fingerprint authentication device For detecting the fingerprint formed on the surface of the tip of the finger that is the detection target For detection surface Fingertip image data detected by multiple sensing elements provided on the detection surface when the entire surface of the fingertip is placed and the fingertip after a part of the fingertip surface touches the detection surface When at least a part of the surface of the tip of the finger comes into contact with the detection surface until the entire surface of the surface comes into contact with the detection surface, for each time detected by each sensing element provided on the detection surface Fingerprint motion data consisting of fingerprint data and Detect fingerprint A sensor, Fingerprint sensor On the detection surface Finger Place Was sometimes The fingerprint motion data detected by the fingerprint sensor between the time when a part of the surface of the tip of the finger touches the detection surface and the time when the entire surface of the finger touches the detection surface. , Detected by the fingerprint sensor when the entire surface of the finger tip touches the detection surface fingerprint Image data and But for each user Memory Registered information storage means, Finger to be detected But For detection surface Be placed When the entire surface of the tip of the finger touches the detection surface, the fingerprint is detected from the fingerprint image data detected by the fingerprint sensor and the image data stored in the registration information storage means. Detected by sensor A degree of coincidence indicating whether or not there is a correlation between the image data and the image data stored in the registered information storage unit is detected for each of a plurality of regions included in the image data. Data relating to the positions of all areas larger than a predetermined threshold is extracted, and as a result of extraction, data relating to the interrelationship between the areas of the obtained degree of coincidence distribution and image data stored in the registration information storage means The number of regions having high similarity between the image data detected by the fingerprint sensor and the image data stored in the registered information storage means based on the data relating to the correlation between the regions of the distribution of coincidence generated from The fingerprint detected from the finger placed on the detection surface of the fingerprint sensor is detected depending on whether the detected verification rate is equal to or greater than a predetermined threshold. The first authentication process for determining whether the fingerprint is the fingerprint, and when the finger to be detected is placed on the detection surface, the finger is touched after a part of the front surface of the finger contacts the detection surface. Provided in the detection surface calculated from the output level of the fingerprint detected for each of the plurality of sensing areas provided in the detection surface of the fingerprint sensor until the entire surface of the tip of the finger contacts the detection surface. When the average value of the output level in all sensing regions is monitored in time series, the detection surface of the fingerprint sensor from the first time when the average value of all sensing regions provided in the detection surface of the fingerprint sensor starts to change. Data obtained as an average value from the first time to the second time until the second time when the average value of all the sensing areas provided in the inside does not change, data of the elapsed time from the first time to the second time, Arbitrary time between the first time and the second time At least the correlation between the sensing areas of the average value of the output level in all sensing areas provided in the detection surface and the average value of all sensing areas provided in the detection surface at the second time The fingerprint detected by the finger placed on the detection surface of the fingerprint sensor from the motion data having feature information including one and the motion data stored in the registration information storage means is the user's fingerprint registered in advance. The fingerprint detected from the finger placed on the detection surface of the fingerprint sensor is pre-registered with the second authentication process for determining whether the second authentication process, the first authentication process result, and the second authentication process result Only when it shows that it is a user's fingerprint, it executes an output process that outputs the fact to the outside And authentication means.
[0008]
Also, in the present invention Affect The fingerprint authentication method is Fingerprints detected by a plurality of sensing elements provided on the detection surface when the entire surface of the fingertip is placed on the detection surface for detecting the fingerprint formed on the surface of the fingertip that is the detection target Image data and at least part of the finger tip surface is detected between when the part of the finger tip surface touches the detection surface and when the entire surface of the finger tip touches the detection surface. A fingerprint detection step of detecting, by a fingerprint sensor provided with a detection surface, fingerprint motion data consisting of fingerprint data for each time detected by each sensing element provided on the detection surface when touching the surface; When the finger to be detected is placed on the detection surface and the entire surface of the tip of the finger touches the detection surface, the fingerprint image data detected by the fingerprint sensor and the detection surface of the fingerprint sensor When the finger is placed The fingerprint motion data detected by the fingerprint sensor between the time when a part of the edge surface touches the detection surface and the time when the entire tip of the finger touches the detection surface, and the finger tip on the detection surface. The image data and registration information detected by the fingerprint sensor from the image data stored in the registration information storage means stored for each user and the fingerprint image data detected by the fingerprint sensor when the entire surface is in contact A degree of coincidence indicating whether there is a correlation with the image data stored in the storage means is detected for each of a plurality of regions included in the image data, and as a result of the detection, the degree of coincidence is lower than a predetermined threshold value. Data relating to the positions of all large areas is extracted, and the data obtained as a result of the extraction is obtained from data relating to the interrelationship between the areas of the obtained distribution of coincidence and the image data stored in the registration information storage means. The number of regions having high similarity between the image data detected by the fingerprint sensor and the image data stored in the registered information storage means based on the data relating to the correlation between the regions of the distribution of coincidence generated from Is detected, and it is determined whether the fingerprint detected from the finger placed on the detection surface of the fingerprint sensor is a pre-registered user's fingerprint based on whether the detected verification rate is equal to or greater than a predetermined threshold. A first authentication processing step for executing the first authentication processing, and when the finger to be detected is placed on the detection surface, the finger is placed after a part of the surface of the tip of the finger contacts the detection surface. Provided in the detection surface calculated from the output level of the fingerprint detected for each of the plurality of sensing areas provided in the detection surface of the fingerprint sensor until the entire surface of the tip of the finger contacts the detection surface. In all sensing areas When the average value of the output level is monitored in time series, it is provided in the detection surface of the fingerprint sensor from the first time when the average value of all the sensing areas provided in the detection surface of the fingerprint sensor starts to change. Data obtained as an average value until the second time when the average value of all the sensing areas does not change, data of elapsed time from the first time to the second time, from the first time The data of the average value of the output level in all the sensing areas provided in the detection surface at an arbitrary time until the second time, and all the sensing areas provided in the detection surface in the second time Detected from a finger placed on the detection surface of the fingerprint sensor from motion data having feature information including at least one of the correlations between the sensing areas of the average value and motion data stored in the registered information storage means Fingerprints The second authentication processing step for executing the second authentication processing for determining whether the user's fingerprint is registered, the processing result of the first authentication processing, and the processing result of the second authentication processing are fingerprints. Only when the fingerprint detected from the finger placed on the detection surface of the sensor indicates that the fingerprint of the user is registered in advance, an output step of executing an output process for outputting the fact to the outside is included. .
[0009]
In the above-described fingerprint authentication apparatus and method of the present invention, when a detection object such as a finger is placed on the detection surface of the sensor, the operation characteristic of the detection object is compared with the registered operation characteristic. At the same time, the detected uneven surface pattern of the detected object is collated with a registered fingerprint to authenticate the user who placed the detected object.
[0013]
DETAILED DESCRIPTION OF THE INVENTION
First, before describing the embodiment of the present invention, the operation of a human finger during fingerprint authentication will be described.
[0014]
As shown in FIG. 1, the fingerprint authentication device 1 includes a fingerprint sensor 2 that detects a spiral-shaped uneven pattern (fingerprint) formed on the surface of the inner portion on the tip side of the first joint of a human finger. Yes. The fingerprint sensor 2 has a detection surface 3 on which the inner part on the tip side from the first joint of the finger of a person who performs personal authentication is placed. The detection surface 3 is provided with a plurality of sensing elements arranged, for example, in a matrix at a pitch sufficiently finer than the pitch of the concave and convex portions of the fingerprint. Each sensing element senses, for example, capacitance, pressure, or amount of reflected light, and outputs a sensing output level corresponding to the distance from the detection surface 3 to the substance placed on the detection surface 3. The fingerprint sensor 2 generates a two-dimensional image of the uneven pattern on the surface of the substance placed on the detection surface 3 based on the sensing output level obtained from each sensing element, and outputs the two-dimensional image as fingerprint data. To do.
[0015]
In such a fingerprint authentication device 1, when a finger is placed on the detection surface 3 and a certain pressure is applied to the detection surface 3, the fingerprint sensor 2 is applied to the detection surface 3. The fingerprint data of the placed finger is detected. The fingerprint authentication device 1 compares the fingerprint data detected by the fingerprint sensor 2 with the fingerprint data registered in advance, and performs personal authentication.
[0016]
By the way, when fingerprint authentication is performed by the fingerprint authentication device 1, the operation of the finger until the detection of the fingerprint data by the fingerprint sensor 2 is started, that is, until the detection surface 3 is pressed with a constant pressure by the tip of the finger. Is as follows.
[0017]
The tip of the finger is gradually brought closer to the detection surface 3 from a distance from the fingerprint authentication device 1, and a part of the inner part of the tip of the finger contacts the detection surface 3. Subsequently, after a part of the inner part of the tip of the finger comes into contact with the detection surface 3, the other part gradually comes into contact with the detection surface 3. Pressed. And the detection surface 3 will be in the state pressed by fixed pressure with the finger | toe.
[0018]
As described above, after the finger presses the detection surface 3 with a constant pressure, the fingerprint authentication device 1 detects the fingerprint.
[0019]
Here, when performing fingerprint authentication by the fingerprint authentication device 1, the present inventor detects until the detection of the fingerprint data is started, that is, until the detection surface 3 is pressed with a constant pressure by the tip of the finger. The sensing output level of the sensing element at each position on the surface 3 was observed. Then, the sensing output level of each sensing element gradually increased from the initial level (the level when nothing was placed on the detection surface 3 and not in proximity). When the detection surface 3 is pressed with a constant pressure by the inner part of the tip of the finger, the sensing output level of each sensing element does not vary at a constant value and is saturated. Further, the sensing output level and the amount of change over time differed for each sensing element.
[0020]
Furthermore, the inventor has observed the characteristics of the sensing output level of each sensing element as described above for a large number of unspecified persons. As a result, the characteristics of the sensing output level of each sensing element are different for each individual. That is, the present inventor has an individual operation characteristic in the output characteristics of each sensing element until the detection surface 3 is pressed with a constant pressure by the tip of the finger when placing the finger on the detection surface. I found.
[0021]
The inventor of the present invention has made use of the above-described individual operation characteristics.
[0022]
Hereinafter, as an embodiment of the present invention, a fingerprint authentication device to which the present invention is applied will be described in detail with reference to the drawings.
In FIG. 2, the external view of the fingerprint authentication apparatus 10 of embodiment of this invention is shown.
[0023]
As shown in FIG. 2, the fingerprint authentication device 10 includes a fingerprint sensor 12 provided on a housing 11 so that a planar detection surface 13 is exposed. The fingerprint sensor 12 is a sensor that detects a fingerprint of a finger placed on the detection surface 13.
[0024]
The fingerprint sensor 12 applied to the fingerprint authentication device 10 is a sensor that detects a two-dimensional image of a fingerprint by, for example, a capacitance method. As shown in FIG. 3, the capacitive fingerprint sensor 12 is provided with a plurality of capacitors 14 arranged in a matrix at a pitch sufficiently finer than the pitch of the fingerprint irregularities on the detection surface 13. . Each capacitor 14 serves as a sensing element of the fingerprint sensor 12. Each capacitor 14 changes its capacitance value according to the distance to the substance when the substance is close. Therefore, when the inner portion on the tip side from the first joint of the finger is placed on the detection surface 13, the capacitance value of the capacitor 14 located in the opposite portion changes according to the unevenness of the fingerprint of the finger.
[0025]
The fingerprint sensor 12 extracts the capacitance values of the plurality of capacitors 14 arranged in a matrix form as a sensing output level, and generates two-dimensional image data corresponding to the position of each capacitor 14 based on the extracted sensing output level. . This two-dimensional image data becomes image data corresponding to the uneven pattern of the fingerprint. The fingerprint sensor 12 outputs this image data as fingerprint image data.
[0026]
The fingerprint sensor 12 also generates motion data indicating operation characteristics when the finger is placed on the detection surface 13 together with the fingerprint image data. Details of the motion data will be described later.
[0027]
Here, the electrostatic capacity detection type fingerprint sensor has been described as an example, but the fingerprint sensor 12 may be any detection method as long as it can detect a two-dimensional fingerprint image. May be. For example, a pressure detection type fingerprint sensor that detects the fingerprint by sensing the pressure at each point on the detection surface when the inner part of the finger tip is placed on the detection surface, or the inner part of the finger tip is placed on the detection surface A photodetection type fingerprint sensor that detects a fingerprint by sensing reflected light from each point on the detection surface at the time may be applied as the fingerprint sensor 12.
[0028]
Next, the internal configuration of the fingerprint authentication device 10 will be described. FIG. 4 shows an internal block diagram of the fingerprint authentication device 10.
[0029]
As shown in FIG. 4, the fingerprint authentication device 10 includes a fingerprint sensor 12, a registered fingerprint storage unit 21, a registered motion storage unit 22, a collation unit 23, and an output unit 24.
[0030]
The fingerprint sensor 12 detects fingerprint image data indicating a fingerprint pattern of a finger placed on the detection surface 13 and motion data indicating operation characteristics when the finger is placed on the detection surface 13 and outputs the detected motion data to the outside. .
[0031]
The registered fingerprint storage unit 21 stores fingerprint image data of a user's registered fingerprint.
[0032]
The registered motion storage unit 22 stores motion data of user registered motion.
[0033]
The collation unit 23 collates the fingerprint image data detected by the fingerprint sensor 12 with the registered fingerprint image data. At the same time, the collation unit 23 collates the motion data detected by the fingerprint sensor 12 with the registered motion data, and authenticates the user who placed the finger on the detection surface 13 by using both. The collation unit 13 outputs OK if the user is authenticated, and outputs NG if the user is not authenticated.
[0034]
The output unit 24 notifies the OK / NG determination result output from the verification unit 23 to, for example, a security device provided outside.
[0035]
Further, when the fingerprint authentication device 10 having the above block configuration is specifically configured by hardware, a configuration as shown in FIG. 5 is obtained.
[0036]
The fingerprint authentication device 10 includes a fingerprint sensor 12, a USB terminal 32 to which a USB interface cable is connected, a USB controller 33, a CPU 34, a program RAM or ROM 35, a flash memory 36, and a fingerprint verification LSI 37. ing. The USB controller 33, CPU 34, program RAM or ROM 35, flash memory 36, and fingerprint verification LSI 37 are connected to a data transfer bus 38, respectively.
[0037]
The USB controller 33 is a circuit that transfers data according to the USB protocol. The function of the output unit 24 shown in FIG. 4 is realized. The USB controller 33 is an interface circuit that transfers data to / from an external personal computer or security device according to the USB protocol, for example. Note that the fingerprint authentication apparatus 10 may transfer data with an external personal computer or the like using another interface such as RS232C instead of the USB interface.
[0038]
The CPU 34 is a circuit that controls the operation of the entire apparatus based on a program stored in the program RAM or ROM 35. For example, the operation control of the matching unit 23 shown in FIG. 4 is performed.
[0039]
The flash memory 36 is a nonvolatile memory. The flash memory 36 realizes the functions of the registered fingerprint storage unit 21 and the registered motion storage unit 22 shown in FIG. 4, and stores registered fingerprint image data and registered motion data.
[0040]
The fingerprint verification LSI 37 authenticates the fingerprint image data and motion data detected by the fingerprint sensor 12 based on the fingerprint image data and motion data stored in the flash memory 36. The function of the collation unit 23 shown in FIG. 4 is realized.
[0041]
Next, the operation of the fingerprint authentication device 10 will be described.
[0042]
First, the fingerprint authentication device 10 performs a user's fingerprint registration operation as an initial setting operation when performing fingerprint authentication.
[0043]
When registering a fingerprint, first, the user places a predetermined finger (for example, an index finger) on the detection surface 13. The fingerprint authentication device 10 captures the fingerprint image data of the finger and the motion data when the finger is placed by the fingerprint sensor 12. The fingerprint authentication device 10 stores the captured fingerprint image data and motion data in the registered fingerprint storage unit 21 and the registered motion storage unit 22 as registration data. By performing the operation as described above, registration of the user's fingerprint is completed. When registering fingerprints of a plurality of users, for example, an ID for each user may be provided, and fingerprint image data and motion data may be registered corresponding to the ID.
[0044]
When the above fingerprint registration is completed, the fingerprint authentication device 10 can authenticate the user.
[0045]
Next, the operation during fingerprint authentication is as follows.
[0046]
First, the fingerprint authentication device 10 starts an authentication operation, for example, by detecting that any object is placed on the detection surface 13 or by detecting that a user ID or the like has been input. When the personal authentication operation is started, the fingerprint sensor 12 detects fingerprint image data and motion data of an object placed on the detection surface 13 and supplies it to the collation unit 23.
[0047]
The collation unit 23 determines whether or not the fingerprint image data detected by the fingerprint sensor 13 matches the registered fingerprint image data by performing pattern matching or the like, for example. If both fingerprint image data match, the collation unit 23 determines that the fingerprint is the registered user's own fingerprint. If the two fingerprint image data do not match, the collation unit 23 determines that the fingerprint (or substance) is other than the registered user. The fingerprint verification may be performed on the entire fingerprint, or only an arbitrary part of the fingerprint may be extracted.
[0048]
Furthermore, the collation unit 23 determines whether there is sufficient correlation between the motion data detected by the fingerprint sensor 13 and the registered motion data, for example, by detecting the degree of correlation. The collation unit 23 determines that the registered user himself / herself has placed his / her finger if the two pieces of motion data are sufficiently correlated. The collation unit 23 determines that a user (or substance) other than the registered user has placed a finger if the motion data of both is not sufficiently correlated.
[0049]
When the collation unit 23 obtains a collation result indicating that both the fingerprint image data and the motion data are registered users, the collation unit 23 outputs a personal authentication determination result (OK) indicating that the user is a registered user. . When the collation unit 23 obtains a collation result indicating that either the fingerprint image data or the motion data is not a registered user, the verification result (NG) indicating that the user is not a registered user. ) Is output.
[0050]
The fingerprint authentication device 10 supplies the determination result obtained by performing the above operation to a computer and various security devices, thereby realizing access management, privacy protection, and the like.
[0051]
Next, a specific procedure for fingerprint collation processing will be described.
[0052]
The fingerprint authentication device 10 performs fingerprint authentication processing by the pattern matching method described below.
[0053]
In the pattern matching method, when a fingerprint is registered, a plurality of linear images D1 are extracted from a part of the fingerprint image to be registered, and the linear images D1 and their extraction positions are stored therein as registration data. At the time of fingerprint collation, pattern matching is performed on the fingerprint image D2 to be collated while shifting the position of each linear image D1, and a position with a high degree of coincidence is detected. This is a method for determining that the fingerprints are the same when the mutual relationship between the linear images at positions with a high degree of coincidence and the mutual relationship between the extracted positions of the registered linear images are equal.
[0054]
For example, in this pattern matching method, as shown in FIG. 6, nine horizontal registration data D1H0 to D1H8, vertical registration data D1V0 to D1V8, and registration data D1H0 to D1H8, as a plurality of linear images D1, Position information (X0, Y0) to (X8, Y0) and (X0, Y0) to (X0, Y8) of D1V0 to D1V8 are registered for each user ID.
[0055]
As shown in FIG. 7, the horizontal direction registration data D1H0 to D1H8 are image data obtained by cutting out a part of a fingerprint image to be registered for 64 pixels in the horizontal direction. Specifically, a total of nine regions of 1 pixel in the vertical direction and 64 pixels in the horizontal direction are cut out from the fingerprint image. As shown in FIG. 8, the vertical direction registration data D1V0 to D1V8 are image data obtained by cutting out a part of a fingerprint image to be registered for 64 pixels in the vertical direction. Specifically, a total of nine areas of one pixel in the horizontal direction and 64 pixels in the vertical direction are cut out from the fingerprint image.
[0056]
The position information (X0, Y0) to (X8, Y0) and (X0, Y0) to (X0, Y8) are the reference positions (for example, the pixel positions at the left end and the upper end) of each linear image cut out as described above. Is a value that identifies
[0057]
In this pattern matching method, collation is performed as follows based on the registration data as described above.
[0058]
First, the collation unit 23 sequentially reads the image data D1H0 to D1H8 and D1V0 to D1V8 of each linear image, and sequentially moves the read linear image on the fingerprint image D2 to be collated. The collation unit 23 detects the degree of coincidence indicating the degree of correlation between the linear image D1 and the fingerprint image D2 to be collated at each movement position.
[0059]
For example, in the case of fingerprint data D1H0 to D1H8 of linear images that are continuous in the horizontal direction, the degree of coincidence is detected by raster scanning the fingerprint image D2 to be collated in the horizontal direction as shown in FIG. This is done by determining whether each bit matches or not. For example, in the case of fingerprint data D1V0 to D1V8 of linear images that are continuous in the vertical direction, as shown in FIG. 10, the fingerprint image D2 to be collated is raster scanned in the vertical direction, and each bit at each scanning position. This is done by determining the match / mismatch.
[0060]
Based on this detection result, the collation unit 23 extracts a collation position where the degree of coincidence exceeds a predetermined threshold for each linear image.
[0061]
Then, the collation unit 23 determines that the fingerprints are the same based on the mutual relationship between the linear images of the obtained distribution of coincidence and the mutual relationship between the extracted positions of the registered linear images.
[0062]
The above fingerprint collation processing is specifically performed according to the procedure shown in the flowchart of FIG. The following fingerprint collation processing is performed by, for example, the CPU 34 shown in FIG.
[0063]
First, the linear image data D2 detected from the fingerprint sensor 12 is input (step S11).
[0064]
Subsequently, the variable m is set to 0 (step S12). The variable m is a variable that specifies the registered 9 × 2 linear image data (registered data).
[0065]
Subsequently, for example, based on the user ID input by the key, the m-th registration data D in the horizontal direction for the corresponding user ID is read from the flash memory 36, and this registration data is loaded into the fingerprint verification LSI 37 (step 37) S13).
[0066]
Subsequently, the fingerprint data D2 detected from the fingerprint sensor 3 is output to the fingerprint verification LSI 37 (step S14). As a result, while the horizontal linear image is raster-scanned on the image based on the fingerprint data D2, the degree of coincidence at each position is detected, and a collation position distribution is obtained so that the degree of coincidence exceeds a predetermined threshold.
[0067]
Subsequently, the variable m is incremented (step S15).
[0068]
Subsequently, it is determined whether or not the variable m exceeds the value 9 (step S16). In this step S16, it can be determined whether or not a coincidence distribution has been detected for each of the nine linear images. If a negative result is obtained in step S16, the process returns to step S13. As a result, the processing procedure of steps S13-S14-S15-S16 is repeated, and the matching position distribution having a high degree of coincidence can be detected for all nine linear images cut out in the horizontal direction (or vertical direction). .
[0069]
If an affirmative result is obtained in step S16 and a matching position distribution having a high degree of coincidence is detected for each of the nine linear images, then a matching position distribution having a high degree of coincidence has been detected for the registered data in the vertical direction. It is determined whether or not (step S17). If a negative result is obtained, the processing target is switched to vertical registration data (step S18), and the process returns to step S13. Thereby, the processing procedure of steps S13-S14-S15-S16 is repeated, and the distribution of the collation positions having a high degree of coincidence can be detected for each of the nine linear images cut out in the vertical direction.
[0070]
If an affirmative result is obtained in step S17 and a matching position distribution with a high degree of matching is detected, then a matching rate detection process is executed, and each of the 9 × 2 linear images has a high degree of matching on D2. A combination of a relative position of each linear image in the registered image and a relative position with a high similarity is detected from the distribution of the verification positions, and the number of relative positions with a high similarity is set as a verification rate N. The matching rate N is detected from the distribution of coincidence detected for the 9 × 2 line images (step S19).
[0071]
Subsequently, it is determined whether or not the verification rate N is equal to or greater than a certain value (step S20). If the collation rate N is equal to or greater than a certain value, a match determination result is output (step S21), and the process ends. If the verification rate N is less than or equal to a certain value, a mismatch determination result is output (step S22), and the process ends.
[0072]
In the fingerprint authentication device 10, the fingerprint collation process is performed as described above.
[0073]
Next, the motion data detected by the fingerprint sensor 12 will be described in detail.
[0074]
In the fingerprint sensor 12, motion data detection areas are set at arbitrary positions on the detection surface 13. On the detection surface 13, for example, a plurality of sensing elements (capacitors in the case of a capacitive fingerprint sensor) formed in a matrix are provided. Each detection area has an arbitrary number of sensing elements. Is set to include.
[0075]
An example of setting the detection area is shown in FIG. In the example shown in FIG. 12, the entire detection surface 13 is composed of horizontal 128 dots × vertical 192 dots sensing elements, and the detection surface 13 has a rectangular detection of horizontal 16 dots × vertical 16 dots. Five areas are set. Each detection area (A to E) shown in FIG. 12 is set to a position that surrounds the periphery when the inner part of the tip of the finger is placed on the detection surface 13.
[0076]
When the finger is placed on the detection surface 13 by the user, the fingerprint sensor 12 determines the average value X of the sensing output levels of all the sensing elements in each detection area from each detection area set as described above. Is calculated and monitored in time series for a predetermined time. The start timing of the monitor is a timing at which the sensing output level from the sensing element starts to fluctuate due to the detection target approaching the detection surface 13. The monitor end timing is a timing at which a constant pressure is applied to the detection surface 13 by the object to be detected and the sense output level from the sensing element does not fluctuate, that is, the sense output level is saturated.
[0077]
The fingerprint sensor 12 outputs the above monitor results as motion data.
[0078]
Specifically, a method for generating motion data in the case of a capacitive fingerprint sensor will be described.
[0079]
In the case of the capacitive fingerprint sensor 12, the sensing output level from each sensing element is extracted as an analog signal (Rf) for each line as shown in FIG. Further, in the case of the capacitive fingerprint sensor 12, the unevenness of the fingerprint is represented by a difference from the surrounding sensing output level. Therefore, in the case of the capacitive fingerprint sensor 12, the average value X of the sensing output levels of all the sensing elements in each detection area is calculated as follows.
[0080]
First, an average RF level including a predetermined number of surrounding sensing elements is obtained for the sensing element of interest. For example, as shown in FIG. 14, the average RF level (AVE_Rf) of the 7 × 7 sensing elements around the sensing element of interest is obtained. Subsequently, the output level (Rf) of the sensing element of interest is subtracted from this average RF level (AVE_Rf), and an absolute value | (RF−AVE_Rf) | of the subtraction result is obtained. Subsequently, the absolute value | (RF−AVE_Rf) | of the subtraction result is averaged in each detection area to obtain an average value X.
[0081]
In the case of a capacitive fingerprint sensor, the average value X of the sensing output levels of all sensing elements in each detection area can be obtained as described above.
[0082]
Subsequently, the capacitive fingerprint sensor 12 obtains the above average value X for each detection area, and monitors the obtained average value X in time series. The fingerprint sensor 12 applies a certain pressure to the detection surface 13 with the finger from the time when the average value X of the sensed output level starts changing at least because the tip of the finger approaches the detection surface 13. Thus, the time-series average value X obtained from each detection area until the point when the average value X of the sensed output level does not change is output as motion data.
[0083]
The motion data generated in this way is parameter time-series data representing how clearly the crest and trough portions of the fingerprint are output. That is, Rf = “output value of the peak (or valley)” and AVE_Rf = “average value in the vicinity”, so | Rf−AVE_Rf | = “the height of the peak (or valley) as viewed from the vicinity (or Therefore, the motion data is a time-series change of the absolute value accumulated in a certain region. For this reason, for example, a fingerprint that has almost no difference in height between peaks and valleys has Rf−AVE_Rf = 0 even if the total output from the fingerprint sensor 12 is large. 0.
[0084]
FIG. 15 shows an example of motion data when a certain user places a finger on the detection surface 13. In FIG. 15, the horizontal axis represents time, and the vertical axis represents the average value X of the sensing output level. Note that the motion data shown in FIG. 15 is composed of a sensing element of 128 horizontal pixels × vertical 192 dots as shown in FIG. It is the motion data acquired from the shape detection areas A to E. In FIG. 15, the average value of the sensing output level obtained from the detection surface A is shown by a graph XA, the average value of the sensing output level obtained from the detection surface B is shown by a graph XB, and the sensing value obtained from the detection surface C is shown. The average value of the output level is indicated by a graph XC, the average value of the sensed output level obtained from the detection surface D is indicated by a graph XD, and the average value of the sensed output level obtained from the detection surface E is indicated by a graph XE. .
[0085]
In the case of the graph shown in FIG. 15, the time when the average value X of the sensed output level starts changing due to the finger tip approaching the detection surface 13 is about 100 ms (T1). Further, the time point when the average value X of the sensed output level does not change due to the constant pressure applied to the detection surface 13 by the finger is the time point (T2) of about 550 ms. The sensed output level after the time 550 ms (T2) is a stable value. In this example, the fingerprint sensor 12 outputs the average values XA to XE of the detection output levels of the detection areas A to X as motion data from 100 ms (time T1) to 550 ms (time T2).
[0086]
The motion data generated in this way is unique and unique to each individual due to, for example, differences in the shape of individual fingers, the shape of the inside of the finger, or differences in movement between individuals.
[0087]
The motion data output from the fingerprint sensor 12 in this way is supplied to the collation unit 23, and the correlation with the motion data registered in advance by the supply unit 23 is collated, and personal authentication is performed.
[0088]
In the fingerprint sensor 12, the motion data is not all data from the time when the average value X of the sensing output level starts to change (T1) until the time when the average value X of the sensing output level does not change (T2). Alternatively, only feature information may be extracted and used as motion data. Examples of the characteristic information include the following (1) to (3).
[0089]
(1) Time (T2−T1) from the time point (T1) when the average value X of the sensing output level starts to change for a certain detection area to the time point (T2) when the average value X of the sensing output level does not change ). In the case of the example shown in FIG. 15, 550 ms-100 ms = 450 ms. In addition, when there are a plurality of detection areas, the time for each detection area.
[0090]
(2) The value of the average value X of the sensing output level in each detection area at the time point (T2) when the average value X of the sensing output level no longer changes. In addition, the sensing output of each detection area at an arbitrary time from the time when the average value X of the sensing output level starts to change (T1) to the time when the average value X of the sensing output level does not change (T2). The average value X of the level.
[0091]
(3) Correlation of the average value X of the sensing output level of each detection area at the time when the average value X of the sensing output level no longer changes (for example, at the time of 400 ms in FIG. 15). For example, in the case shown in FIG. 15, the relationship is such that the level of the region A is the highest, the level of the detection area B is the next highest, and the level of the detection area C is the next highest. In addition, the sensing output of each detection area at an arbitrary time from the time when the average value X of the sensing output level starts to change (T1) to the time when the average value X of the sensing output level does not change (T2). Correlation of the average value X of the level.
[0092]
In addition, the fingerprint sensor 12 may use the combination information as motion data instead of any one of the above-described feature information (1) to (3).
[0093]
That is, the fingerprint sensor 12 senses at a predetermined time from the time point (T1) when the average value X of the sensed output level starts changing (T1) to the time point (T2) when the average value X of the sensed output level stops changing. A change amount or level value of the average value X of the output level may be extracted as a feature point and used as motion data.
[0094]
Thus, by using only feature points as motion data, the amount of processing for generating motion data can be reduced, and the amount of computation at the time of collation processing can be reduced.
[0095]
As described above, the fingerprint authentication device 10 performs fingerprint collation, detects an operation characteristic at the time of fingerprint collation, and authenticates the operation characteristic. For this reason, the fingerprint authentication device 10 can more reliably authenticate the person.
[0096]
Further, the fingerprint authentication apparatus 10 detects the fine stripe pattern characteristic of the fingerprint by obtaining the degree of change of the output level between the elements having a fine positional relationship and measuring the time series change. Further, by measuring, storing, and collating the detection process, fingerprint collation with higher accuracy can be performed.
[0097]
Note that the motion data detected by the fingerprint sensor 12 may have different values depending on the position where the finger is placed on the detection surface 13. For example, depending on how the detection area is set, for example, when the finger is placed at the center of the detection surface 13 and when the finger is placed at the end of the detection surface 13, the motion data differs. There is. In consideration of this, for example, as shown in FIG. 16, the detection area may be formed in a strip shape, and a detection region having a horizontal stripe shape may be formed over the entire detection surface 13. Alternatively, when the authentication process based on the fingerprint image data is performed first, the fingerprint image data calculates the position where the finger is placed on the detection surface 13, and the motion data is corrected based on the calculated position. Good. Alternatively, a detection area may be set for the entire detection surface 13 so that the motion data can be reliably detected regardless of the position of the finger on the detection surface 13.
[0098]
Further, the detection surface 13 may be supported by an elastic mechanism so that motion data can be detected stably. The configuration of the fingerprint sensor 12 in which the detection surface 13 is supported by an elastic mechanism is shown in FIGS.
[0099]
As shown in FIGS. 17 and 18, the fingerprint sensor 12 includes a support substrate 31 fixed to the casing of the fingerprint authentication device 10. Four springs 32 a to 32 d are provided on the main surface of the support substrate 31. The springs 32 a to 32 d are configured such that the direction perpendicular to the main surface of the support substrate 31 is the direction in which the spring 32 a to 32 d is elastically deformed. The fingerprint sensor 12 includes a rectangular thin plate-shaped movable substrate 33 having a detection surface 13 attached to the main surface. The movable substrate 33 is supported by the support substrate 31 via springs 32a to 32d at the four corners on the back surface side to which the detection surface 13 is not attached. Thus, in the fingerprint sensor 12, the detection surface 13 is supported by the elastic member so as to be movable in a direction perpendicular to the main surface.
[0100]
The fingerprint sensor 12 is provided with a finger placement detection switch 34 between the support substrate 31 and the movable substrate 33. The finger placement detection switch 34 is a switch that is turned on when the movable substrate 33 moves in the direction of the support substrate 31 by a certain distance Z.
[0101]
In the fingerprint sensor 12 having such a configuration, since the detection surface 13 is supported by the elastic member from the back surface side, when a finger or the like is placed on the detection surface 13, the constant is determined by the elastic force of the elastic member. Only the pressure is applied to the detection surface 13.
[0102]
Therefore, with the fingerprint sensor 12 having such a configuration, it is possible to accurately detect the time point (T2) at which the average value X of the sensed output level does not change, and to detect motion data stably. In addition, since the finger placement detection switch 34 is provided in the fingerprint sensor 12, it can be detected reliably and in a short time that a certain pressure determined by the elastic force is applied to the detection surface 13.
[0103]
Furthermore, in the fingerprint sensor 12 having such a configuration, fingerprint image data is not detected unless a certain pressure determined by the elastic force of the elastic member is applied. Therefore, for example, even when a pseudo-fingerprint or the like formed of a soft member is placed, it is required to press with a constant pressure, so that the shape of the pseudo-fingerprint can be destroyed.
[0104]
【The invention's effect】
In the fingerprint authentication device and method according to the present invention, when an object to be detected such as a finger is placed on the detection surface of the sensor, the operation characteristic of the object to be detected is compared with the registered operation characteristic. Then, the detected unevenness of the surface of the detected object is collated with a registered fingerprint to authenticate the user who placed the detected object.
[0105]
For this reason, in this fingerprint authentication apparatus and method, the accuracy of personal authentication can be improved.
[0106]
In the authentication apparatus and method according to the present invention, when a detection object such as a finger is placed on the detection surface of the sensor, the operation characteristic of the detection object is compared with the registered operation characteristic, The identity of the user who placed the detected object is authenticated.
[0107]
For this reason, in this authentication apparatus and method, the accuracy of personal authentication can be improved by using it together with fingerprint authentication.
[Brief description of the drawings]
FIG. 1 is a configuration diagram of a fingerprint authentication device of the present invention.
FIG. 2 is a schematic perspective view of the fingerprint authentication device according to the embodiment of the present invention.
FIG. 3 is a diagram for explaining the arrangement of sensing elements provided in a fingerprint sensor.
FIG. 4 is a block configuration diagram of the fingerprint authentication device according to the embodiment of the present invention.
FIG. 5 is a block diagram of a fingerprint authentication device configured with hardware.
FIG. 6 is a diagram for explaining the contents of registration data.
FIG. 7 is a diagram for explaining a linear image cut out in the horizontal direction;
FIG. 8 is a diagram for explaining a linear image cut out in the vertical direction;
9 is a diagram for explaining the scanning order of the linear image in FIG. 7; FIG.
10 is a diagram for explaining the scanning order of the linear image in FIG. 8; FIG.
FIG. 11 is a flowchart showing a fingerprint collating operation.
FIG. 12 is a diagram for explaining an example of arrangement of detection areas;
FIG. 13 is a diagram for explaining a sensed output detected from a detection area.
FIG. 14 is a diagram for explaining detection of an averaged Rf level.
FIG. 15 is a diagram for describing motion data.
FIG. 16 is a diagram for explaining another arrangement example of detection areas;
FIG. 17 is a side view of a fingerprint sensor provided with an elastic mechanism.
FIG. 18 is a plan view of a fingerprint sensor provided with an elastic mechanism.
[Explanation of symbols]
1,10 fingerprint detection device, 2,12 fingerprint sensor, 3,13 detection surface, 21 registered fingerprint storage unit, 22 registered motion storage unit, 23 collation unit

Claims (8)

When the entire surface of the finger tip is placed on the detection surface for detecting the fingerprint formed on the surface of the finger tip that is the detection target, the detection is made by a plurality of sensing elements provided on the detection surface. The fingerprint image data and the surface of the tip of the finger between the part of the surface of the finger tip touching the detection surface and the time when the entire surface of the finger tip touches the detection surface. A fingerprint sensor that detects motion data of a fingerprint consisting of fingerprint data for each time detected by each of the sensing elements provided on the detection surface when at least a part of the sensor touches the detection surface ;
During a portion of the surface of the tip of the finger when the finger is put it in any detection surface of the fingerprint sensor from the contact with the detection surface to the entire surface of the tip of the finger is in contact with the detection surface, Registered information stored for each user includes fingerprint motion data detected by the fingerprint sensor and fingerprint image data detected by the fingerprint sensor when the entire tip of the finger contacts the detection surface. Storage means;
When the finger to be detected is placed on the detection surface, when the entire tip of the finger touches the detection surface, the fingerprint image data detected by the fingerprint sensor and the registration information storage A degree of coincidence indicating whether there is a correlation between the image data detected by the fingerprint sensor and the image data stored in the registered information storage means from the image data stored in the means in the image data Detection is performed for each of a plurality of regions included, and as a result of the detection, data relating to the positions of all regions having a degree of coincidence larger than a predetermined threshold is extracted. From the data relating to the interrelationship between the areas and the data relating to the interrelationships between the areas of the distribution of coincidence generated from the image data stored in the registered information storage means, Whether a collation rate indicating the number of regions having a high degree of similarity between the image data detected by the sensor and the image data stored in the registration information storage means is detected, and whether the detected collation rate is equal to or greater than a predetermined threshold value The first authentication process for determining whether the fingerprint detected from the finger placed on the detection surface of the fingerprint sensor is a pre-registered user's fingerprint and the detection target finger against the detection surface When a part of the surface of the tip of the finger comes into contact with the detection surface until the entire surface of the tip of the finger comes into contact with the detection surface. When the average value of the output level in all the sensing areas provided in the detection surface calculated from the output level of the fingerprint detected for each of the plurality of sensing areas provided is time-series monitored, the fingerprint sensor detection From the first time when the average value of all the sensing areas provided in the region starts to change to the second time when the average value of all the sensing areas provided in the detection surface of the fingerprint sensor does not change Data obtained as an average value between, the data of the elapsed time from the first time to the second time, the above at any time from the first time to the second time The data of the average value of the output level in all the sensing areas provided in the detection surface, and the correlation between each sensing area of the average value of all the sensing areas provided in the detection surface at the second time From the motion data having feature information including at least one and the motion data stored in the registration information storage means, the fingerprint detected by the finger placed on the detection surface of the fingerprint sensor is registered in advance. Is a fingerprint The fingerprint detected from the finger placed on the detection surface of the fingerprint sensor in advance is a second authentication process for determining whether or not the result of the first authentication process and the result of the second authentication process are A fingerprint authentication apparatus comprising: an authentication unit that executes an output process of outputting the fact to the outside only when indicating that the fingerprint is a registered user . The sensing region, the fingerprint authentication device of the rectangular der Ru請 Motomeko 1 wherein. The above fingerprint sensor, the finger to be positioned at the periphery of the finger when placed on the detection surface, the fingerprint authentication device 請 Motomeko 2 wherein rectangular said sensing region that has been set. The sensing region, the fingerprint authentication device of the belt-like Der Ru請 Motomeko 1 wherein. Above the fingerprint sensor, so as to be arranged at predetermined intervals over the entire surface of the detection surface, the fingerprint authentication device of a strip-shaped 請 Motomeko 4, wherein the sensing area is it is configured. The fingerprint sensor is above detects the capacitance to a plurality of points on the detection surface, the fingerprint authentication 請 Motomeko 1, wherein you detect fingerprint is a pattern unevenness of the surface of the finger placed on the detection surface apparatus. The fingerprint sensor is a fingerprint authentication device 請 Motomeko 1, wherein the back surface side of the detection surface that is supported by an elastic member. When the entire surface of the finger tip is placed on the detection surface for detecting the fingerprint formed on the surface of the finger tip that is the detection target, the detection is made by a plurality of sensing elements provided on the detection surface. The fingerprint image data and the surface of the tip of the finger between the part of the surface of the finger tip touching the detection surface and the time when the entire surface of the finger tip touches the detection surface. When at least a part of the sensor touches the detection surface, the fingerprint motion data including the fingerprint data for each time detected by each of the sensing elements provided on the detection surface and the fingerprint provided with the detection surface are provided. A fingerprint detection step detected by a sensor;
When the finger to be detected is placed on the detection surface, when the entire tip of the finger touches the detection surface, the fingerprint image data detected by the fingerprint sensor and the fingerprint sensor When a finger is placed on the detection surface, a part of the surface of the tip of the finger touches the detection surface until the entire surface of the finger tip contacts the detection surface. The motion data of the detected fingerprint and the image data of the fingerprint detected by the fingerprint sensor when the entire surface of the finger touches the detection surface are stored in the registration information storage means stored for each user. The image data includes a degree of coincidence indicating whether there is a correlation between the image data detected by the fingerprint sensor and the image data stored in the registered information storage means. Each region of the distribution of coincidence obtained as a result of the extraction, extracting data related to the positions of all the regions where the degree of coincidence is greater than a predetermined threshold. Image data detected by the fingerprint sensor, and data related to the interrelationship between the regions of the distribution of coincidence generated from the image data stored in the registration information storage means A collation rate indicating the number of regions having a high degree of similarity with the image data stored in the registration information storage means is detected. Depending on whether the detected collation rate is equal to or greater than a predetermined threshold, the detection surface of the fingerprint sensor is A first authentication processing step for executing a first authentication processing for determining whether a fingerprint detected from a placed finger is a fingerprint of a user registered in advance;
When a finger to be detected is placed on the detection surface, a part of the surface of the tip of the finger contacts the detection surface until the entire surface of the finger tip contacts the detection surface. In the meantime, the average value of the output levels in all the sensing areas provided in the detection surface calculated from the output level of the fingerprint detected for each of the plurality of sensing areas provided in the detection surface of the fingerprint sensor In the case of time series monitoring, all the sensing areas provided in the detection surface of the fingerprint sensor from the first time when the average value of all the sensing areas provided in the detection surface of the fingerprint sensor starts to change. Data obtained as an average value from the first time to the second time, data obtained from the first time to the second time, and from the first time to the above Any time between the second time The data of the average value of the output level in all the sensing areas provided in the detection surface, and the correlation between each sensing area of the average value of all the sensing areas provided in the detection surface at the second time A user who has previously registered a fingerprint detected from a finger placed on the detection surface of the fingerprint sensor from motion data having feature information including at least one of the above and motion data stored in the registration information storage means A second authentication processing step for executing a second authentication processing for determining whether the fingerprint is a fingerprint of
The processing result of the first authentication processing and the processing result of the second authentication processing are determined by the fingerprint sensor. An output step for executing an output process for outputting the fact to the outside only when the fingerprint detected from the finger placed on the detection surface indicates that the fingerprint of the user is registered in advance
Fingerprint authentication method including.

Images