TW202034210A - Noise elimination method for fingerprint identification, control device, and information processing device eliminates interference of common-mode noises in fingerprint recognition using a software - Google Patents

Abstract

A method for eliminating noise in fingerprint recognition is characterized by comprising the following steps. In a row of sampling values comprising a plurality of sampling groups, perform an edge slope assimilation calculation and a preliminary stretching operation on a preceding sampling group and a subsequent sampling group in any two adjacent sampling groups, and then calculating a height difference between the last numerical value of the preceding sampling group and the first numerical value of the subsequent sampling group, and perform a vertical translation operation on the plurality of values of the subsequent sampling group according to the height difference so as to align the after-sampling group and the preceding sampling group.

Description

Fingerprint identification noise elimination method, control device and information processing device

The present invention relates to the technical field of fingerprint recognition, in particular to a noise elimination method and control device for fingerprint recognition.

As financial transactions such as mobile payment and electronic securities orders gradually use electronic products such as smart phones and smart wearable products as the main operating platforms, the biometric identification technology-based identity recognition system is integrated into these smart types Mobile electronic devices are also getting more and more attention. Biometric identification technology uses the inherent physiological characteristics of the human body as the identification basis for individual organisms, such as: iris (Iris), face (Face), voice print (Voice), fingerprint (Fingerprint) and other physiological characteristics .

Currently, commercially available fingerprint recognition devices are divided into optical, pressure, ultrasonic, and capacitive. Among them, the most easily integrated into a single chip through a standard CMOS process and related control circuits, capacitive fingerprint recognition devices have been It is widely used in various electronic products. FIG. 1 shows the structure diagram of a conventional capacitive fingerprint identification device. As shown in FIG. 1, the capacitive fingerprint identification device 1'is applied to a smart electronic device 2', and includes: a sensing module 11', a signal sensing unit 12', and an analog/digital The conversion unit 13', a clock unit 14', a control and processing unit 15', and a drive unit 16'.

In the operation control of fingerprint recognition, the control and processing unit 15' inputs a plurality of scanning signals to the sensing module 11' through the driving unit 16'; at this time, when a finger presses on the sensing module 11' At this time, the signal sensing unit 12' senses the different capacitance changes caused by the plural ridges and valleys of the finger dermatoglyph, and correspondingly outputs the plural sensing signals to the The signal sensing unit 12'. Further, the plurality of the sensing signals are converted into corresponding digital data by the analog/digital conversion unit 13'; and, after the signal processing of the digital data is completed by a related algorithm, the control and processing unit 15' The skin texture of the finger pressed on the sensing module 11' can be calculated and drawn. It is worth noting that since the height difference between the crests and troughs of the finger dermatoglyphs is not large, the capacitance changes caused by the crests and troughs on the sensing module 11' will not be very significant. . In this case, any external noise may cause the signal sensing unit 12' to receive a distorted sensing signal, causing the control and processing unit 15' to fail to correctly calculate and describe the pressure on the sensing module 11' The dermatoglyph on the fingers. For example, the charging circuit unit of the conventional capacitive fingerprint identification device 1'will form common-mode noise on the plurality of sensing signals received by the signal sensing unit 12'.

Figure 2a shows the waveform of the sensing signal that is not interfered by common mode noise, Figure 2b shows the waveform of the sensing signal that is interfered by common mode noise, and Figure 3a shows that the control and processing unit is not interfered by common mode noise. Figure 3b shows the finger texture calculated and depicted by the sensing signal of the control and processing unit based on the sensing signal interfered by the common mode noise. Comparing FIG. 2a and FIG. 2b, it can be found that after being interfered by common mode noise, the signal sensing unit 12' will receive a distorted sensing signal. In particular, the distorted sensing signal is no longer a complete and continuous sensing signal. FIG. 2a shows the waveform diagram of the sensing signal not interfered by common mode noise, and FIG. 2b shows the waveform diagram of the sensing signal interfered by common mode noise. It can be easily observed from Figure 3a and Figure 3b that, compared to the finger skin texture calculated and drawn based on the sensing signal that is not interfered by common mode noise (as shown in Figure 3a), according to the common mode noise The image (or image) of the finger's dermatoglyph (as shown in FIG. 3b) calculated and depicted by the interfering sensing signal appears blurry, resulting in poor fingerprint recognition.

In view of this, some manufacturers of fingerprint recognition devices will add multiple analog/digital converters to the circuit architecture of the capacitive fingerprint recognition device 1', and at the same time convert the plurality of sensing signals into grayscale values. Digital data, in this way, eliminates the signal interference caused by common mode noise to the plurality of sensing signals. However, the addition of multiple analog/digital converters has greatly increased the overall hardware cost of the capacitive fingerprint identification device 1', resulting in an increase in the price of the capacitive fingerprint identification device 1'.

Therefore, a novel fingerprint noise elimination method is urgently needed in the art.

The main purpose of the present invention is to provide a noise elimination method for fingerprint recognition, which can be applied to any existing fingerprint recognition device. Unlike the existing noise elimination methods, which add multiple analog/digital converters to eliminate the interference caused by common mode noise on the sensing signal, the fingerprint recognition noise elimination method of the present invention can use a limited number of In the case of analog/digital converters, the interference of common mode noise is eliminated by a software method.

In order to achieve the above-mentioned main purpose of the present invention, the inventor of the present case provides an embodiment of the fingerprint recognition noise elimination method, which includes the following steps:

Acquire a plurality of sensing sample values from a sensing module;

Performing a common mode noise detection procedure on the plurality of sensing sample values; and

When the execution result of the common mode noise detection procedure is yes, execute a common mode noise elimination procedure, and then execute a fingerprint image generation procedure, and when the execution result of the common mode noise detection procedure is no, execute The fingerprint image generation program.

In one embodiment, the common mode noise detection procedure includes the following steps:

Divide the plurality of sensing sample values into a plurality of sample groups by a predetermined number n as a group, where n is a positive integer;

Calculate an edge slope of each of the sample groups, and perform a vertical offset calculation on any two adjacent sample groups, which includes: according to the edge slope of the previous sample group and the nth The sensing sample value generates a predicted value, and a subtraction calculation is performed on the predicted value and the first sensing sample value of a subsequent sample group to generate a vertical offset; and

When a sum of the plurality of vertical offsets is greater than a preset threshold, the execution result of the common mode noise detection program is set to Yes, and when the sum is not greater than the preset threshold, the total is set The execution result of the mode noise detection program is no.

In one embodiment, the common mode noise cancellation procedure includes the following steps:

In a row of sample values containing a plurality of the sample groups, performing an edge slope assimilation calculation on a preceding sample group and a subsequent sample group in any two adjacent sample groups, which includes The subsequent sampling group performs a longitudinal stretching operation so that the edge slope becomes the same as the edge slope of the previous sampling group, and the nth sensor of the previous sampling group is calculated A height difference between the sample value and the first value of the subsequent sample group processed by the longitudinal stretching operation; and

According to the height difference, a vertical translation operation is performed on the n values of the subsequent sampling group processed by the longitudinal stretching operation, so that the subsequent sampling group and the previous sampling group are aligned and joined.

In one embodiment, the common mode noise cancellation procedure includes the following steps:

In each of the plurality of row sampling values including a plurality of the sampling groups, each calculate a sampling mean and a standard deviation to obtain a plurality of the sampling mean and a plurality of the standard deviations; and

Select a reference row from the sample values of the plurality of rows, and perform a stretching operation on the other rows according to the ratio of the standard deviation of the other rows to the standard deviation of the reference row, and according to the reference row The sampling average performs an average uniform operation on other rows to generate a stitched fingerprint image.

In an embodiment, the common mode noise elimination procedure further includes: performing a Gaussian filtering procedure on the stitched fingerprint image.

In an embodiment, the reference line is the first line in the plurality of lines of sample values.

In one embodiment, the n is 8.

To achieve the foregoing objective, the present invention further provides a fingerprint recognition control device, which has a control and processing module to perform the aforementioned fingerprint recognition noise elimination method.

To achieve the foregoing objective, the present invention further provides an information processing device, which has a central processing unit and the aforementioned fingerprint recognition control device, and the central processing unit is used to communicate with the control and processing module.

In a possible embodiment, the information processing device can be a smart phone or a portable computer.

In order to enable your reviewer to further understand the structure, features and purpose of the present invention, drawings and detailed descriptions of preferred specific embodiments are attached as follows.

Please also refer to FIGS. 4-8, in which, FIG. 4 is a flowchart of an embodiment of the noise elimination method for fingerprint recognition of the present invention; FIG. 5 is an implementation of a common mode noise detection procedure of the noise elimination method of FIG. 4 Example flowchart; FIG. 6 is a flowchart of an embodiment of a common mode noise elimination procedure of the noise elimination method of FIG. 4; FIG. 7 is another embodiment of a common mode noise elimination procedure of the noise elimination method of FIG. 4 And FIG. 8 is a circuit structure diagram of an embodiment of the fingerprint recognition control device of the present invention, in which a fingerprint recognition control device 100 has a sensing module 110 and a control and processing module 120.

As shown in FIG. 4, the noise elimination method includes: the control and processing module 120 acquires a plurality of sensing sample values from the sensing module 120 (step 1); Perform a common-mode noise detection procedure by measuring the sample value (step 2); and the control and processing module 120 performs a common-mode noise elimination procedure when the execution result of the common-mode noise detection procedure is yes, and then executes a fingerprint Image generation program, and when the execution result of the common mode noise detection program is no, execute the fingerprint image generation program (step 3).

As shown in FIG. 5, the common mode noise detection procedure includes the following steps: the control and processing module 120 divides the plurality of sensing sample values into a plurality of sample groups by a predetermined number n, where n is a positive integer (Step 21); The control and processing module 120 calculates an edge slope of each of the sample groups, and performs a vertical offset calculation on any two adjacent sample groups, which includes: according to one of the preceding The edge slope of the sampling group and the n-th sensing sample value generate a predicted value, and a subtraction calculation is performed based on the predicted value and the first sensing sample value of a subsequent sampling group To generate a vertical offset (step 22); and the control and processing module 120 sets the execution result of the common mode noise detection program when a sum of the vertical offsets is greater than a predetermined threshold If yes, and when the sum is not greater than the preset threshold, set the execution result of the common mode noise detection procedure to be no (step 23).

As shown in FIG. 6, the common mode noise elimination procedure includes the following steps: the control and processing module 120 performs sampling on one of any two adjacent sampling groups in a row of sampling values including a plurality of the sampling groups. An edge slope assimilation calculation is performed on the preceding sample group and a succeeding sample group, which includes performing a longitudinal stretching operation on the succeeding sample group so that the edge slope becomes the same as the preceding sample group The edge slopes of are the same, and a height difference between the n-th sensing sample value of the previous sample group and the first value processed by the longitudinal stretching operation of the following sample group is calculated (Step 31); and the control and processing module 120 performs a vertical translation operation on the n values of the subsequent sampling group processed by the longitudinal stretching operation according to the height difference, so that the subsequent sampling group Align and join with the previous sample group (step 32).

As shown in FIG. 7, the common mode noise elimination procedure may also include the following steps: the control and processing module 120 each calculates a sample average and a standard among the sample values of the plurality of rows each including a plurality of the sample groups. Difference in order to obtain a plurality of the sampling mean values and a plurality of the standard deviations (step 33); the control and processing module 120 selects a reference row from the plurality of row sampling values, and uses the standard deviations of the other rows Perform a stretching operation on other rows based on the ratio of the standard deviation of the reference row, and perform a mean uniform operation on the other rows according to the sample average value of the reference row to generate a stitched fingerprint image (step 34 ); and the control and processing module 120 performs a Gaussian filtering procedure on the stitched fingerprint image (step 35). Wherein, the reference row may be the first row in the plurality of rows of sample values. In addition, in a preferred embodiment, n is 8.

Please also refer to FIGS. 9a-c, which show a schematic diagram of an operation of the common mode noise elimination procedure of the present invention. As shown in the figure, an edge slope assimilation is performed on a previous sampling group (having an edge slope K1) and a subsequent sampling group (having an edge slope K2) of the two adjacent sampling groups in FIG. 9a After calculation, a height difference D between the n-th sensing sample value of the previous sampling group and the first value of the following sampling group can be calculated (as shown in Fig. 9b); and After performing a vertical translation operation on the n values of the subsequent sampling group according to the height difference D, the subsequent sampling group and the previous sampling group can be aligned and joined (as shown in Figure 9c) .

Based on the above description, the present invention further provides an information processing device. Please refer to FIG. 10, which is a block diagram of an embodiment of the information processing apparatus of the present invention. As shown in FIG. 10, an information processing device 200 includes a fingerprint recognition control device 210 and a central processing unit 220, wherein the fingerprint recognition control device 210 has the sensing module 110 of the fingerprint recognition control device 100 shown in FIG. And the control and processing module 120, and the central processing unit 220 is used to communicate with the control and processing module 120.

In this way, the above has completely and clearly explained the composition and technical features of the fingerprint recognition noise elimination method and control device of the present invention; and from the above, it can be seen that the present invention has the following advantages:

1. The fingerprint recognition noise elimination method proposed by the present invention can be applied to any existing fingerprint recognition device.

2. Different from the existing noise elimination methods, which add multiple analog/digital converters to eliminate the signal interference caused by common-mode noise on the complex sensing signal, the fingerprint recognition noise elimination method of the present invention can be used in When a limited number of analog/digital converters are used, the interference of common mode noise can be eliminated by a software method.

3. The application method of the present invention is simple, fast to execute, and only occupies a small part of hardware resources in the fingerprint identification control device.

It must be emphasized that the foregoing disclosures in this case are preferred embodiments, and any partial changes or modifications that are derived from the technical ideas of this case and are easily inferred by those who are familiar with the art will not deviate from the patent of this case. Right category.

In summary, regardless of the purpose, means and effects of this case, it is shown that it is very different from the conventional technology, and its first invention is suitable for practicality, and it does meet the patent requirements of the invention. Please check it out and grant the patent as soon as possible. Society is for the best prayer.

<The present invention> Step 1: Acquire multiple sensing sample values from a sensing module Step 2: Perform a common mode noise detection procedure on the plurality of sensing sample values Step 3: When the execution result of the common mode noise detection procedure is yes, execute a common mode noise elimination procedure, and then execute a fingerprint image generation procedure, and if the execution result of the common mode noise detection procedure is no , Execute the fingerprint image generation program Step 21: Divide the plurality of sensing sample values into a plurality of sample groups by a predetermined number n as a group, where n is a positive integer Step 22: Calculate an edge slope of each of the sample groups, and generate a predicted value for any two adjacent n-th sensing sample values, and based on the predicted value and a subsequent sample Perform a subtraction calculation on the first said sensing sample value of the group to generate a vertical offset. The sampling group performs a vertical offset calculation Step 23: Determine whether a sum of the plurality of vertical offsets is greater than a preset threshold Step 31: Perform an edge slope assimilation calculation and a height difference calculation Step 32 Step 33: Perform a vertical translation operation. Calculate a sampling mean and a standard deviation for each row of sampling values Step 34: Perform a stretching operation and a mean uniform operation on each row Step 35: Perform a Gaussian filtering procedure on the stitched fingerprint image 100: Fingerprint recognition control device 110: Sensing module 120: control and processing module 200: Information processing device 210: Fingerprint recognition control device 220: Central Processing Unit

＜Acquaintances＞ 1’: Capacitive fingerprint recognition device 2’: Smart electronic device 11’: Sensing module 12’: Signal sensing unit 13’: Analog/digital conversion unit 14’: Clock unit 15’: Control and Processing Unit 16’: drive unit

FIG. 1 is a structural diagram of a conventional capacitive fingerprint identification device. Figure 2a is a waveform diagram of a sensing signal that is not interfered by common mode noise. Figure 2b is a waveform diagram of a sensing signal interfered by common mode noise. Figure 3a shows the finger skin texture calculated and depicted by the control and processing unit based on the sensing signal not interfered by common mode noise. Figure 3b shows the finger skin texture calculated and drawn by the control and processing unit based on the sensing signal interfered by common mode noise. FIG. 4 is a flowchart of an embodiment of the fingerprint recognition noise elimination method of the present invention. 5 is a flowchart of an embodiment of a common mode noise detection procedure of the noise elimination method of FIG. 4. FIG. 6 is a flowchart of an embodiment of a common mode noise elimination procedure of the noise elimination method of FIG. 4. FIG. 7 is a flowchart of another embodiment of a common mode noise elimination procedure of the noise elimination method of FIG. 4. FIG. 8 is a circuit structure diagram of an embodiment of the fingerprint recognition control device of the present invention. Figures 9a-c show a schematic diagram of an operation of the common mode noise elimination procedure of the present invention. FIG. 10 is a block diagram of an embodiment of the information processing device of the present invention.

Step 1: Acquire multiple sensing sample values from a sensing module

Step 2: Perform a common mode noise detection procedure on the plurality of sensing sample values

Step 3: When the execution result of the common mode noise detection procedure is yes, execute a common mode noise elimination procedure, and then execute a fingerprint image generation procedure, and if the execution result of the common mode noise detection procedure is no , Execute the fingerprint image generation program

Claims (10)

A fingerprint recognition noise elimination method is applied to a fingerprint recognition control device. The method includes the following steps: (1) Retrieving a plurality of sensing sample values from a sensing module; (2) To the said Perform a common mode noise detection procedure on a plurality of sensing sample values; and (3) When the execution result of the common mode noise detection procedure is yes, execute a common mode noise elimination procedure, and then execute a fingerprint image generation procedure , And when the execution result of the common mode noise detection program is no, execute the fingerprint image generation program. For the fingerprint recognition noise elimination method described in item 1 of the scope of patent application, the common mode noise detection procedure includes the following steps: (21) Divide the plurality of sensing sample values into a group of a predetermined number n A plurality of sampling groups, n is a positive integer; (22) Calculate an edge slope of each of the sampling groups, and perform a vertical offset calculation on any two adjacent sampling groups, which includes: The edge slope of the sampling group and the n-th sensing sample value generate a predicted value, and a prediction value is performed based on the predicted value and the first sensing sample value of a subsequent sampling group Subtracting calculation to generate a vertical offset; and (23) when a sum of a plurality of the vertical offsets is greater than a preset threshold, set the execution result of the common mode noise detection program to be yes, and When the sum is not greater than the preset threshold, the execution result of the common mode noise detection program is set to be no. According to the fingerprint recognition noise elimination method described in the scope of the patent application, the common mode noise elimination procedure includes the following steps: (31) In a row of sample values containing a plurality of the sample groups, Perform an edge slope assimilation calculation for a previous sampling group and a subsequent sampling group in any two adjacent sampling groups, which includes performing a longitudinal stretching operation on the subsequent sampling group to make the The edge slope becomes the same as the edge slope of the previous sampling group, and the n-th sensing sample value of the previous sampling group is calculated and the subsequent sampling group undergoes the longitudinal stretching operation A height difference of the first processed value; and (32) perform a vertical translation operation on the n values of the subsequent sample group processed by the longitudinal stretching operation according to the height difference, so that the The subsequent sampling group is aligned with the previous sampling group. According to the fingerprint identification noise elimination method described in the scope of the patent application, the common mode noise elimination procedure includes the following steps: (33) In each of the plural rows of sample values including a plurality of the sample groups , Each calculating a sampling mean and a standard deviation to obtain a plurality of the sampling mean and a plurality of the standard deviations; and (34) selecting a reference row from the plurality of row sampling values, and according to the other rows The ratio of the standard deviation to the standard deviation of the reference row performs a stretching operation on other rows, and performs a mean uniform operation on the other rows according to the sampling mean value of the reference row to generate a spliced fingerprint image Like. According to the fingerprint recognition noise elimination method described in the scope of patent application, the common mode noise elimination procedure further includes: (35) performing a Gaussian filtering procedure on the stitched fingerprint image. The noise elimination method for fingerprint recognition described in item 4 of the scope of patent application, wherein the reference line is the first line of the plurality of line sample values. The fingerprint recognition noise elimination method described in item 2 of the scope of patent application, wherein the n is 8. A fingerprint identification control device, which has a control and processing module to implement the fingerprint identification noise elimination method as described in any one of the 1-7 patent applications. An information processing device has a central processing unit and the fingerprint identification control device as described in item 8 of the scope of patent application, and the central processing unit is used to communicate with the control and processing module. For example, the information processing device described in item 9 of the scope of patent application is a smart phone or a portable computer.

Images