Fingerprint sensor and correction method thereof

0Associated
Cases 
0Associated
Defendants 
0Accused
Products 
0Forward
Citations 
0
Petitions 
2
Assignments
First Claim
1. A fingerprint sensor, comprising a sensing part to be touched by a finger, the sensing part comprising a plurality of sensing units, wherein the fingerprint sensor further comprises a processing unit and a correction unit, wherein:
 the sensing unit is configured to sense, when correction starts, a fingerprint signal and output first data;
the processing unit is configured to calculate, according to a mathematical model, a new correction coefficient of the corresponding sensing unit by using the first data; and
the correction unit is configured to correct, according to the new correction coefficient, second data output by the corresponding sensing unit;
wherein the processing unit comprises a first calculation subunit, configured to calculate, according to the mathematical model, a new correction coefficient of the corresponding sensing unit by using the first data and the old correction coefficient of the corresponding sensing unit,
2 Assignments
0 Petitions
Accused Products
Abstract
A fingerprint sensor and a correction method thereof are provided. The fingerprint sensor includes a sensing part to be touched by a finger, including sensing units, a processing unit, and a correction unit; the sensing unit is configured to sense, when correction starts, a fingerprint signal and output first data; the processing unit is configured to calculate, according to a mathematical model, a new correction coefficient of the corresponding sensing unit by the first data; and the correction unit is configured to correct, according to the new correction coefficient, second data output by the corresponding sensing unit. The fingerprint sensor can automatically modify a correction coefficient when a physical characteristic of the fingerprint sensor changes, thereby reducing the product return rate and prolonging the product service life, omitting a step of correction before delivery, and reducing the manufacturing difficulty and the manufacturing costs.
4 Citations
No References
Automatically balanced sensing device and method for multiple capacitive sensors  
Patent #
US 7,504,833 B1
Filed 03/31/2006

Current Assignee
Monterey Research LLC

Sponsoring Entity
Cypress Semiconductor Corporation

Biometric information obtaining apparatus and biometric information verification apparatus  
Patent #
US 20050100200A1
Filed 12/22/2004

Current Assignee
Fujitsu Limited

Sponsoring Entity
Fujitsu Limited

BIOMETRIC AUTHENTICATION DEVICE AND BIOMETRIC AUTHENTICATION METHOD  
Patent #
US 20130259321A1
Filed 06/03/2013

Current Assignee
Fujitsu Limited

Sponsoring Entity
Fujitsu Limited

BIOMETRIC INFORMATION CORRECTION APPARATUS, BIOMETRIC INFORMATION CORRECTION METHOD AND COMPUTERREADABLE RECORDING MEDIUM FOR BIOMETRIC INFORMATION CORRECTION  
Patent #
US 20140133711A1
Filed 10/30/2013

Current Assignee
Fujitsu Limited

Sponsoring Entity
Fujitsu Limited

10 Claims
 1. A fingerprint sensor, comprising a sensing part to be touched by a finger, the sensing part comprising a plurality of sensing units, wherein the fingerprint sensor further comprises a processing unit and a correction unit, wherein:
the sensing unit is configured to sense, when correction starts, a fingerprint signal and output first data; the processing unit is configured to calculate, according to a mathematical model, a new correction coefficient of the corresponding sensing unit by using the first data; and the correction unit is configured to correct, according to the new correction coefficient, second data output by the corresponding sensing unit; wherein the processing unit comprises a first calculation subunit, configured to calculate, according to the mathematical model, a new correction coefficient of the corresponding sensing unit by using the first data and the old correction coefficient of the corresponding sensing unit,  View Dependent Claims (2, 3, 4, 5)
 6. A correction method for a fingerprint sensor, the fingerprint sensor comprising a sensing part to be touched by a finger, the sensing part comprising a plurality of sensing units, wherein the method comprises steps of:
S10;
when correction starts, sensing a fingerprint signal and obtaining first data output by each sensing unit;S20;
calculating, according to a mathematical model, a new correction coefficient of the corresponding sensing unit by using the first data; andS30;
correcting, according to the new correction coefficient, second data output by the corresponding sensing unit;wherein step S20 specifically comprises;
calculating, according to the mathematical model, a new correction coefficient of the corresponding sensing unit by using the first data and the old correction coefficient of the corresponding sensing unit, View Dependent Claims (7, 8, 9, 10)
1 Specification
The present application is a continuation of International Application No. PCT/CN2015/094441, filed on Nov. 12, 2015, which claims priority to Chinese Patent Application No. CN 201510016150.6, filed on Jan. 13, 2015, both of which are hereby incorporated by reference herein in their entireties.
The present disclosure relates to the technical field of fingerprint recognition, and in particular, to a fingerprint sensor and a correction method thereof.
Fingerprints are texture formed by uneven skin on surfaces of fingers. The fingerprints have texture characteristics of uniqueness and stability, and therefore are often used as a basis for identity recognition. A fingerprint sensor is a sensor for recognizing an identity by means of fingerprints.
As shown in
In an actual environment, differences exist among sensing units 11 in the same fingerprint sensor, in addition, unevenness of the adhesive 30 and impurities therein, as well as a flatness matter of the cover 20 or the like, also exist in the fingerprint sensor, and therefore responses from all the sensing units 11 are uneven. Thus, it may be improbable to obtain accurate fingerprint texture characteristics directly from data output by the fingerprint sensor. Instead, the output data needs to be corrected to eliminate such unevenness.
For this purpose, in practice, a correction operation is performed on the fingerprint sensor before delivery. A twopoint correction method is used in the correction operation. In the method, on the basis of a linear relationship between output data of a sensing unit 11 and an input signal of the sensing unit 11 (that is, information about a depth of fingerprint texture), the output data of the sensing unit 11 is represented by y and the input signal of the sensing unit 11 is represented by x, coefficients k and b satisfy a linear equation y=kx+b. Therefore, a linear equation shown in
The specific process of the twopoint correction method includes: firstly, a frame of data output by a sensing unit 11 is obtained upon a condition that the fingerprint sensor is nonloaded (that is, no object presses the sensing part of the sensor), and then, another frame of data output by the sensing unit 11 is obtained when a flat metallic block 50 is placed on the fingerprint sensor (as shown in

 where P1(x_{1},y_{1}) represents data output in the nonloaded condition and P2(x_{2},y_{2}) represents data output when the metallic block is placed, and it can be obtained that:
In addition, as shown in
After values of k and b are obtained, the output data can be corrected by using the following formula:

 where N and M represent rows and columns of the sensing units 11 in the fingerprint sensor, respectively, y represents data output by the sensing units 11, and x represents data after correction (that is, the input signals of the sensing units 11).
However, the abovedescribed correction method for a fingerprint sensor in the prior art has the following defects:
 1) correction needs to be performed before delivery and the correction process increases product costs.
 2) a correction environment requirement is high: a quite flat metallic block is required for correction, and it also requires no impurity such as ash exists on a sensing part of a fingerprint sensor during correction.
 3) the correction is only performed for once before delivery in the prior art, while physical characteristics of a fingerprint sensor changes over time, and therefore correction coefficients (that is, the values of k and b) obtained in the previous correction becomes invalid when the physical characteristics of the fingerprint sensor change, such that the fingerprint sensor cannot be normally used and the service life of the product is shortened.
The major objective of the present disclosure lies in providing a fingerprint sensor and a correction method thereof, intended to implement correction on the fingerprint sensor during a use process, to automatically modify a correction coefficient, so as to reduce the product return rate and prolong the product service life.
To achieve the objective, the present disclosure provides a fingerprint sensor, having a sensing part to be touched by a finger, including a sensing unit, a processing unit, and a correction unit, where
 the sensing unit is configured to sense a fingerprint signal and output first data when correction starts;
 the processing unit is configured to calculate, according to a mathematical model, a new correction coefficient of the corresponding sensing unit by the first data; and
 the correction unit is configured to correct, according to the new correction coefficient, second data output by the corresponding sensing unit.
The present disclosure also provides a correction method for a fingerprint sensor, the fingerprint sensor including a sensing part to be touched by a finger, the sensing part including a plurality of sensing units, where the method includes the steps of:
 S10: obtaining first data output by each sensing unit when correction starts;
 S20: calculating, according to a mathematical model, a new correction coefficient of the corresponding sensing unit by the first data; and
 S30: correcting, according to the new correction coefficient, second data output by the corresponding sensing unit.
The fingerprint sensor provided in the present disclosure senses an actual fingerprint signal generated when being actually pressed by a finger and obtains data output by the sensing units, to calculate a correction coefficient during use, such that a correction operation can be performed at any time during use, and that the correction coefficient can be automatically modified when a physical characteristic of the fingerprint sensor changes, thereby reducing the product return rate and prolonging the product service life. Moreover, a step of correction before delivery is simplified or omitted for the fingerprint sensor of the embodiments of the present disclosure, thereby reducing the manufacturing difficulty and the manufacturing costs.
The objective implementations, functional features, and advantages of the present disclosure are further described in the embodiments with reference to the accompanying drawings.
It should be understood that the specific embodiments described herein are used only to interpret the present disclosure, but are not intended to limit the present disclosure.
Referring to
The sensing unit is configured to sense, when correction starts, a fingerprint signal and output first data.
The processing unit is configured to calculate, according to a mathematical model, a new correction coefficient of the corresponding sensing unit by using the first data.
Specifically, the fingerprint sensor has two modes, namely, a normal mode and a correction mode. The normal mode is used for performing normal fingerprint sensing and outputting sensing data. The correction mode is used for correcting the fingerprint sensor and obtaining correction coefficients. The new correction coefficient is obtained by using a neural network, data statistics, curvefitting, a mathematical model, or the like; for example, a linear model of y=kx+b, but the present disclosure is not limited thereto. Mode switching of the fingerprint sensor may be performed by means of a mode switch unit.
The correction unit is configured to correct, according to the new correction coefficient, second data output by the corresponding sensing unit.
As shown in
Specifically, after the fingerprint sensor is switched to the correction mode, the sensing unit senses a fingerprint signal and obtains a frame of first data output when the fingerprint sensor is nonloaded (that is, no object presses the sensing part). The first data output by all the sensing units constitutes a frame of data. Then, a user is prompted to press different positions on the sensing part by a finger. Each time when the user presses the sensing part by the finger, a relevant frame of data output by the sensing units is obtained. In other words, the number of frames of data depends on how many times the sensing part is pressed by the finger.
When a fingerprint ridge part (that is, a fingerprint protrusion part) presses the sensing part, more specifically, on a sensing unit, data output by the sensing unit has a maximum value. When a fingerprint valley part (that is, a fingerprint recess part) presses the sensing part, data output by a corresponding sensing unit has a smaller value. For example, when a finger presses the sensing part as shown in
The user presses the sensing part by the finger for multiple times, in such a manner that a pressing position and a pressing angle of the finger on the sensing part vary each time. Therefore, after the user presses the fingerprint sensor by the finger for a plurality of times, each sensing unit in the fingerprint sensor can be pressed by the fingerprint ridge part. The number of times for which the sensing part is pressed by the finger, that is, the number of frames obtained at load, is determined as required. Generally, the larger the sensing part is, the more frames are required, while the smaller the sensing part is, the less frames are required, on the basis that it is ensured all the sensing units are pressed by the fingerprint ridge part. For example, as shown in
The processing unit includes a selection subunit and a second calculation subunit. The selection subunit selects a maximum value corresponding to each sensing unit from the at least two frames of data. The second calculation subunit is configured to calculate a new correction coefficient for the sensing unit.
Specifically, the selection subunit selects a maximum value corresponding to each sensing unit from at least two frames of data which are obtained at load, the maximum values of all the sensing units are equivalent to constituting a frame of data output when all the sensing units are pressed by the fingerprint ridge part. The frame of data is similar to data output by the sensing units when the sensing part is pressed by a flat metallic block during correction. Correction can be performed on the fingerprint sensor by using the frame of data and a frame of data output by the sensing units in the nonloaded condition.
In the present disclosure, data obtained by a sensing unit in the nonloaded condition is represented by Dmin, a maximum value of data obtained by the sensing unit at load is represented by Dmax, the sensing part be pressed by a finger for Num times (Num is a constant and is determined according to a specific situation), and the plurality of sensing units are arranged in N rows and M columns. Dmax_{i,j}=MAX(D1_{i,j}, D2_{i,j}, D3_{i,j}, . . . , Dnum_{i,j}), where Dnum_{i,j }represents the (Num)^{th }frame of data output by the sensing unit at the i^{th }row and the j^{th }column, i ∈ [1,N], and j ∈ [1,M].
Then, the second calculation subunit calculates values of correction coefficients k and b based on a liner model y=kx+b:
If previous correction has been performed on the fingerprint sensor or primary correction before delivery has been performed, old correction coefficients are stored in a storage unit. The old correction coefficients are replaced with the correction coefficients obtained in this correction process which are used as new correction coefficients, and the new correction coefficients are stored. Alternatively, the old correction coefficients are replaced with new correction coefficients of a weighted sum of the old correction coefficients and the correction coefficients obtained in this correction process, and the new correction coefficients are stored. The correction coefficients obtained in this correction process are represented by k_{new }and b_{new}, the old correction coefficients are represented by k_{old }and b_{old}, the new correction coefficients are represented by k and b, and the weighted values, which can be set according to specific situation, are 0.8 and 0.2, it can be obtained that:
Compared to the method of direct replacement, the method of replacing the old correction coefficients with a calculated weighted sum of the old correction coefficients and the correction coefficients obtained in this correction process can accelerate a convergence speed of correction coefficient calculation and improve robustness of the system.
If no previous correction has been performed on the fingerprint sensor and this correction process is the first correction, the correction coefficients obtained in this correction process are used as new correction coefficients of the fingerprint sensor, and are stored.
When the fingerprint sensor is switched to the normal mode for sensing fingerprint information, the correction unit corrects, according to the following formula, data output by the sensing units:
where N and M represent rows and columns of the sensing units in the fingerprint sensor, respectively, y′ represents the second data output by the sensing units, and x represents data after correction by means of the new correction coefficients.
The fingerprint sensor in the first embodiment of the present disclosure obtains at least two frames of data output by the sensing units when different positions of the sensing part are pressed by a finger, and selects maximum values from the two frames of data to constitute a frame of data output when all the sensing units are pressed by the fingerprint ridge part. The frame of data is similar to data output by the sensing units when the sensing part is pressed by a flat metallic block during correction. Correction can be performed on the fingerprint sensor by using the frame of data and a frame of data output by the sensing units in a nonloaded condition, to obtain new correction coefficients. A correction operation can be performed at any time during use for the fingerprint sensor in the embodiment of the present disclosure and the correction coefficients can be automatically modified when a physical characteristic of the fingerprint sensor changes, thereby reducing the product return rate and prolonging the product service life. Moreover, a step of correction before delivery is simplified or omitted for the fingerprint sensor in the embodiment of the present disclosure, thereby reducing the manufacturing difficulty and the manufacturing costs.
Referring to

 where k_{new }and b_{new }represent new correction coefficients of the sensing unit; k_{old }and b_{old }represent old correction coefficients of the sensing unit; y represents the first data output by the sensing unit; Y represents an image constituted by first data output by all the sensing units; F(Y) represents data of the corresponding sensing units after performing median filtering on Y; Avg(Y) represent data of the corresponding sensing units after calculating an average value of Y; σ represents a learning factor, where a range of σ is [0,1] and
and x_{old }represents data of the sensing unit obtained after correction process by means of the old correction coefficients.
When the fingerprint sensor is switched to the normal mode for sensing fingerprint information, the correction unit corrects, according to the following formula, data output by the sensing units:

 where N and M represent rows and columns of the sensing units in the fingerprint sensor, respectively, y′ represent the second data output by the sensing units, and x represent data after correction process by means of the new correction coefficients.
The fingerprint sensor in the second embodiment of the present disclosure obtains a frame of first data by a finger pressing, and calculates a new correction coefficient by using the first data in combination with an old correction coefficient. A correction operation can be performed at any time during use for the fingerprint sensor in the embodiment of the present disclosure and the correction coefficients can be automatically modified when a physical characteristic of the fingerprint sensor changes, thereby reducing the product return rate and prolonging the product service life.
Referring to
When detecting that the fingerprint sensor starts for the first time, the determining unit determines that the correction condition is satisfied and enables the correction mode.
Optionally, a correction time may be set. Upon detecting that the preset correction time is reached, the determining unit determines that the correction condition is satisfied and enables the correction mode. The correction time may be a specific time point, for example, the correction is set to perform once at a certain time on a certain day of a certain month; alternatively, the correction time may be a period of time, for example, the correction is set to perform once for every three months.
Optionally, an effect of correction by means of the old correction coefficient can be evaluated. If the determining unit determines that the effect does not reach the designated requirement, the correction condition is satisfied and the correction mode is enabled. The effect of correction by means of the old correction coefficient is evaluated by means of an image smoothness or a rootmeansquare error of data, that is output by all the sensing units and corrected by means of the old correction coefficient.

 where N and M represents rows and columns of the sensing units in the fingerprint sensor, respectively, y represents the data output by the sensing units, x_{old }represents the data of the sensing units after being corrected by means of the old correction coefficient, f represents an image of data of all the sensing units after correction, and h_{1}=[1, −1]. If the image smoothness or the rootmeansquare error is greater than a preset threshold, it is determined that the correction effect does not reach the designated requirement.
In addition, a user may also start the correction mode manually. For example, the user sends a correction instruction by means of a menu option, a functional or virtual button, a touch control instruction, a preset gesture action, or the like. If the determining unit receives the correction instruction, the determining unit determines that the correction condition is satisfied and enables the correction mode.
Optionally, the determining unit may also determine, based on a current mode, whether the correction condition is satisfied. Upon detecting that the fingerprint sensor is currently in the correction mode, the determining unit determines that the correction condition is satisfied.
It should be understood that the correction mode can also be enabled by setting other correction conditions apart from the foregoing listed examples, which is not described in detail herein.
Referring to
In the S100: a correction mode is switched to.
In this embodiment, the fingerprint sensor has two modes, namely, a normal mode and a correction mode. The normal mode is used for performing normal fingerprint sensing and outputting sensing data. The correction mode is used for correcting the fingerprint sensor and obtaining correction coefficients. If the fingerprint sensor is switched to the correction mode, a correction operation starts.
In the S110: a frame of data that is output by the sensing units when the fingerprint sensor is nonloaded, is obtained.
The fingerprint sensor includes a sensing part to be pressed by a finger. The sensing part is usually covered by a cover shown in
In the S120: at least two frames of data that are output by the sensing units when different positions on the sensing part of the fingerprint sensor are pressed by a finger, is obtained.
The fingerprint sensor may prompt a user to press different positions on the sensing part by a finger. Each time when the user presses the sensing part by the finger, a relevant frame of data output by the sensing units is obtained. In other words, the number of frames of second data depends on how many times the sensing part is pressed by the finger
In an actual use process, the user presses the sensing part by the finger in such a manner that a pressing position and a pressing angle of the finger on the sensing part vary each time. Therefore, after the user presses the fingerprint sensor by the finger for a plurality of times, each sensing unit in the fingerprint sensor can be pressed by the fingerprint ridge part. The number of times for which the sensing part is pressed by the finger, that is, the number of frames obtained data, is determined as required. Generally, the larger the sensing part is, the more frames are required, while the smaller the sensing part is, the less frames are required, on the basis that it is ensured all the sensing units are pressed by the fingerprint ridge part. For example, as shown in
In the S130: a maximum value corresponding to each sensing unit is selected from the at least two frames of data.
The fingerprint sensor selects a maximum value from at least two frames of data output by each sensing unit, the maximum values of all the sensing units are equivalent to constituting a frame of data output when all the sensing units are pressed by the fingerprint ridge part. The frame of data is similar to data output by the sensing units when the sensing part is pressed by a flat metallic block during correction. Correction can be performed on the fingerprint sensor by using the frame of data and a frame of data output by the sensing units in a nonloaded condition.
Data in a nonloaded condition is represented by Dmin, a maximum value during press by a finger is represented by Dmax, the sensing part be pressed by a finger for Num times (Num is a constant and is determined according to a specific situation), and the plurality of sensing units are arranged in N rows and M columns, Dmax_{i,j}=MAX(D1_{i,j}, D2_{i,j}, D3_{i,j}, . . . Dnum_{i,j}), where Dnum_{i,j }represents the (Num)^{th }frame of data output by the sensing unit at the i^{th }row and the j^{th }column, i ∈ [1, N], and j ∈ [1, M].
In the S140: The fingerprint sensor calculates a new correction coefficient of the sensing unit: k_{new}=corresponding maximum value−data output in a nonloaded condition, and b_{new}=data output in a nonloaded condition.
Calculation is performed by using the selected maximum value Dmax and Dmin according to the linear model y=kx+b, to obtain values of the correction coefficients k and b:
In the S150: whether there is an old correction coefficient is determined.
If a previous correction has been performed on the fingerprint sensor or primary correction before delivery has been performed, the old correction coefficient is stored and the method turns to S160. If no previous correction has been performed and this correction process is the correction for the first time, there is not an old correction coefficient and the method turns to S170.
In the S160: the old correction coefficients are replaced with the correction coefficients obtained in this correction process which are used as new correction coefficients.
If there is an old correction coefficient, the fingerprint sensor replaces the old correction coefficient by using a correction coefficient obtained in this correction process which are used as a new correction coefficient, and stores the new correction coefficients.
In some embodiments, a new correction coefficient and an old correction coefficient may also be weighted summed, and the calculation result thereof may be used as a new correction coefficient to replace the old correction coefficient and the new correction coefficient is stored. The correction coefficients obtained in this correction are represented by k_{new }and b_{new}, old correction coefficients are represented by k_{old }and b_{old}, new correction coefficients are represented by k and b, and the weighted values, which can be set according to specific situation, are 0.8 and 0.2, it can be obtained that:
Compared to the method of direct replacement, the method of replacing the old correction coefficients with a calculated weighted sum of the old correction coefficients and the correction coefficients obtained in this correction process can accelerate a convergence speed of correction coefficient calculation and improve robustness of the system.
In the S170: the correction coefficient obtained in this correction process is used directly as a correction coefficient of the fingerprint sensor.
If there is not an old correction coefficient, the fingerprint sensor directly uses the correction coefficient obtained in this correction process as a correction coefficient of the fingerprint sensor and stores the correction coefficient obtained in this correction process.
Up to now, the correction is completed. When the fingerprint sensor is switched to the normal mode for sensing fingerprint information, data output by the sensing units is corrected according to the following formula:

 where N and M represent rows and columns of the sensing units in the fingerprint sensor, y′ represents the second data output by the sensing units, and x represents data corrected by means of the new correction coefficient.
In some embodiments, the S110 of obtaining data in the nonloaded condition may also be performed after the S130 of selecting the maximum value.
In the embodiment of the present disclosure, at least two frames of data output by the sensing units when different positions of the sensing part are pressed by a finger, are obtained, and maximum values are selected from the two frames of data, to constitute a frame of data output when all the sensing units are pressed by the fingerprint ridge part. The frame of data is similar to data output by the sensing units when the sensing part is pressed by a flat metallic block during correction. Correction can be performed on the fingerprint sensor by using the frame of data and a frame of data output by the sensing units in a nonloaded condition, to obtain new correction coefficients. A correction operation can be performed at any time during use for the fingerprint sensor in the embodiment of the present disclosure and the correction coefficients can be automatically modified when a physical characteristic of the fingerprint sensor changes, thereby reducing the product return rate and prolonging the product service life. Moreover, a step of correction before delivery is simplified or omitted for the fingerprint sensor in the embodiment of the present disclosure, thereby reducing the manufacturing difficulty and the manufacturing costs.
Referring to
In the S200: t a correction mode is switched to.
In the S210: when correction starts, a fingerprint signal is sensed and first data output by each sensing unit is obtained.
In the S220: a new correction coefficient of the corresponding sensing unit is calculated, according to a mathematical model based on the first data and an old correction coefficient of the corresponding sensing unit.

 where k_{new }and b_{new }represent new correction coefficients of the sensing unit; k_{old }and b_{old }represent old correction coefficients of the sensing unit; y is the first data output by the sensing unit; Y represents an image constituted by first data output by all the sensing units; F(Y) is data of the corresponding sensing units after median filtering is performed on Y; Avg(Y) represents data of the corresponding sensing units after an average of Y is calculated; σ represents a learning factor, where a range of σ is [0,1] and
and x_{old }represents data of the sensing unit obtained after correction by means of the old correction coefficient.
In the S230: whether there is an old correction coefficient is determined.
If a previous correction has been performed on the fingerprint sensor or primary correction before delivery has been performed, an old correction coefficient is stored and the method turns to S240. If no previous correction has been performed and this correction is the correction for the first time, there is not an old correction coefficient and the method turns to S250.
In the S240: the old correction coefficients are replaced with the correction coefficients obtained in this correction process which are used as new correction coefficients.
If there is an old correction coefficient, the fingerprint sensor replaces the old correction coefficient with a correction coefficient obtained in this correction process which is used as a new correction coefficient, and stores the new correction coefficient.
In some embodiments, a new correction coefficient and an old correction coefficient may also be weighted summed, and the calculation result thereof may be used as a new correction coefficient to replace the old correction coefficient and the new correction coefficient is stored. The correction coefficients obtained in this correction are represented by k_{new }and b_{new}, old correction coefficients are represented by k_{old }and b_{old}, new correction coefficients are represented by k and b, and weighted values, which can be set according to specific situation, are 0.8 and 0.2, it can be obtained that:
Compared to the method of direct replacement, the method of replacing the old correction coefficients with a calculated weighted sum of the old correction coefficients and the correction coefficients obtained in this correction process can accelerate a convergence speed of correction coefficient calculation and improve robustness of the system.
In the S250: the correction coefficient obtained in this correction process is used directly as a correction coefficient of the fingerprint sensor.
In the 260: according second data output by the corresponding sensing unit is corrected according to the new correction coefficient.
When the fingerprint sensor is switched to the normal mode for sensing fingerprint information, data output by the sensing units is corrected according to the following formula:

 where N and M represent rows and columns of the sensing units in the fingerprint sensor, y′ represents the second data output by the sensing units, and x represents data corrected by means of the new correction coefficient.
S230, S240, and S250 correspond to and are the same as S150, S160, and S170, respectively, which will not be described in detail again herein.
In this embodiment, a new correction coefficient is calculated by using a frame of first data obtained through one press by a finger in combination with an old correction coefficient. A correction operation can be performed by the fingerprint sensor at any time during use in the embodiment of the present disclosure and the correction coefficient can be automatically modified when a physical characteristic of the fingerprint sensor changes, thereby reducing the product return rate and prolonging the product service life.
Referring to
In the S300: whether a correction condition is satisfied is determined.
In this embodiment, a correction mode is enabled by setting a correction condition. When it is detected that the fingerprint sensor is enabled for the first time, a preset correction time is reached, a correction instruction is received, or an effect of correction by means of an old correction coefficient does not reach a designated requirement, it is determined that the correction condition is satisfied, the correction mode is enabled, the method turns to S310, and a correction operation starts. When it is determined that the correction condition is not satisfied, the determination are continuously performed.
The fingerprint sensor may be set to be in a correction modeenabled state by default when it is enabled for the first time. Optionally, a correction time may be set. The correction time may be a specific time point, for example, the correction is set to be performed once at a certain time on a certain day of a certain month, or may be set to be performed at a time interval, for example, the correction may set to be performed once for every three days. Furthermore, the user sends a correction instruction by means of a menu option, a functional or virtual button, a touch control instruction, a preset gesture action, or the like, to manually enable the correction mode.
Optionally, an effect of correction by means of the old correction coefficient can be evaluated. If the determining unit determines that the effect does not reach the designated requirement, the correction condition is satisfied and the correction mode is enabled. The effect of correction by means of the old correction coefficient is evaluated by means of an image smoothness or a rootmeansquare error of data that is output by all the sensing units and corrected by means of the old correction coefficient.

 where N and M represent rows and columns of the sensing units in the fingerprint sensor, respectively, y represents the data output by the sensing units, x_{old }represents the data of the sensing units after being corrected by means of the old correction coefficient, f represents an image of data of all the sensing units after correction, and h_{1}=[1, −1]. If the image smoothness or the rootmeansquare error is greater than a preset threshold, it is determined that the correction effect does not reach the designated requirement.
It should be understood that the correction function can also be enabled by setting other correction conditions apart from the foregoing listed examples, which is not described in detail herein.
Optionally, the fingerprint sensor may also determine, by determining a current mode, whether the correction condition is satisfied. For example, if it is detected that the fingerprint sensor is currently in the correction mode, the fingerprint sensor determines that the correction condition is satisfied.
In the S310: when correction starts, a fingerprint signal is sensed and first data output by each sensing unit is obtained.
In the S320: a new correction coefficient of the corresponding sensing unit is calculated by the first data according to a mathematical model.
In the S330: whether there is an old correction coefficient is determined.
If there is an old correction coefficient, S340 is processed; and otherwise, S350 is processed.
In the S340, the old correction coefficients are replaced with the correction coefficients obtained in this correction process which are used as new correction coefficients.
If there is an old correction coefficient, the fingerprint sensor replaces the old correction coefficient with a correction coefficient obtained in this correction process which is used as a new correction coefficient, and stores the new correction coefficient.
In some embodiments, a new correction coefficient and an old correction coefficient may also be weighted summed, and the calculation result thereof may be used as a new correction coefficient to replace the old correction coefficient and the new correction coefficient is stored. The correction coefficients obtained in this correction are represented by k_{new }and b_{new}, old correction coefficients are represented by k_{old }and b_{old }new correction coefficients are represented by k and b, and the weighted values, which can be set according to specific situation, are 0.8 and 0.2, it can be obtained that:
Compared to the method of direct replacement, the method of replacing the old correction coefficients with a calculated weighted sum of the old correction coefficients and the correction coefficients obtained in this correction process can accelerate a convergence speed of correction coefficient calculation and improve robustness of the system.
In the S350: the correction coefficient obtained in this correction process is used directly as a correction coefficient of the fingerprint sensor.
In the S360: according to the new correction coefficient, second data output by the corresponding sensing unit is corrected according to the new correction coefficient.
It should be noted that: the present disclosure is described by examples based on division of all the foregoing functional modules when the fingerprint sensor provided in the embodiment performs correction, and in actual application, the functions can be allocated to be implemented by different functional modules according to needs. In addition, the fingerprint sensor provided in the embodiments and the correction method embodiments for a fingerprint sensor belong to the same inventive concept, and the technical features in the method embodiments are accordingly applicable in the embodiments of the fingerprint sensor.
A person of ordinary skill in the art may understand that all or some of the steps of the foregoing embodiments may be implemented by using hardware, or may be implemented by a program instructing relevant hardware. The program may be stored in a computer readable storage medium. The storage medium may be a ROM/RAM, a magnetic disk, an optical disc, or the like.
The foregoing descriptions are merely preferred embodiments of the present disclosure but are not intended to limit the patent scope of the present disclosure. A person skilled in the art may implement the present disclosure by using various variant solutions without departing from the scope and essence of the present disclosure, for example, further another embodiment is obtained if a feature of one embodiment is applied to another embodiment. Any alteration, equivalent change or modification made within the technical concept of the present disclosure shall fall within the scope of the claims of the present disclosure.