Screen noise reduction method and device, terminal equipment and storage medium

Abstract

The application belongs to the field of noise reduction, and particularly relates to a screen noise reduction method and device, a terminal device and a storage medium. The method comprises the following steps: acquiring sampling data and converting the sampling data into a data matrix; performing class convolution processing on the data matrix to obtain class convolution data, and obtaining a first noise reduction matrix according to the class convolution data and a preset first noise range; performing convolution processing on the first noise reduction matrix to obtain convolution data of each sampling data in the first noise reduction matrix; and obtaining a second noise reduction matrix according to the convolution data and a preset second noise range. That is, the application performs class convolution processing on the data matrix and obtains the first noise reduction matrix according to the first noise range, thereby completing the first noise reduction of the sampling data, performs convolution processing on the first noise reduction matrix and obtains the second noise reduction matrix according to the second noise range, thereby completing the second noise reduction of the sampling data, and electromagnetic interference can be effectively filtered, and the noise reduction effect can be significantly improved.

Description

TECHNICAL FIELD

[0001] The present application belongs to the field of noise reduction, and particularly relates to a screen noise reduction method and device, a terminal device, and a storage medium. BACKGROUND

[0002] The existing intelligent touch screen will generate certain electromagnetic interference when the internal circuit is complex to a certain extent.

[0003] In actual production, under the influence of factors such as screen size and cost, it is impossible to effectively filter electromagnetic interference on signals only by hardware. SUMMARY

[0004] The embodiments of the present application provide a screen noise reduction method and device, a terminal device, and a storage medium, which can effectively filter electromagnetic interference and significantly improve the noise reduction effect.

[0005] In a first aspect, the embodiments of the present application provide a screen noise reduction method, comprising:

[0006] Obtaining sampling data and converting the sampling data into a data matrix;

[0007] Performing class convolution processing on the data matrix to obtain class convolution data of each sampling data in the data matrix, and obtaining a plurality of first target data with abnormal noise in the data matrix according to each class convolution data and a preset first noise range, the plurality of first target data being non-adjacent in the data matrix;

[0008] Replacing the plurality of first target data with reference data to obtain a first noise reduction matrix, the reference data being data of the screen without being touched and without noise;

[0009] Performing convolution processing on the first noise reduction matrix to obtain convolution data of each sampling data in the first noise reduction matrix;

[0010] Obtaining a plurality of second target data with abnormal noise in the first noise reduction matrix according to each convolution data and a preset second noise range, and replacing the plurality of second target data with convolution data corresponding to each second target data to obtain a second noise reduction matrix, the second noise range being smaller than the first noise range.

[0011] In a possible implementation manner of the first aspect, the class convolution processing on the data matrix to obtain the class convolution data of each sampling data in the data matrix, and the obtaining of the plurality of first target data with abnormal noise in the data matrix according to each class convolution data and the preset first noise range, comprises:

[0012] Edge filling is performed on the data matrix based on the reference data to obtain a target matrix;

[0013] Each first convolution matrix is obtained by traversing the sampling data in the target matrix using a first convolution kernel, the first convolution matrix being a matrix corresponding to each sampling data in the target matrix;

[0014] In each first convolution matrix, a plurality of first target data with abnormal noise in the data matrix are obtained according to a noise reduction condition corresponding to the first noise range.

[0015] In each first convolution matrix, a plurality of first target data with abnormal noise in the data matrix are obtained according to a noise reduction condition corresponding to the first noise range.

[0016] In each first convolution matrix, a first processing data and a second processing data are determined, the first processing data being the sampling data corresponding to each first convolution matrix, and the second processing data being data other than the first processing data in each first convolution matrix;

[0017] If the first processing data is not within the first noise range and the second processing data is within the first noise range, a plurality of first target data with abnormal noise in the data matrix are obtained.

[0018] The convolution data of each sampling data in the first noise reduction matrix is obtained by performing convolution processing on the first noise reduction matrix.

[0019] Each second convolution matrix is obtained by traversing the sampling data in the first noise reduction matrix using a second convolution kernel, the second convolution matrix being a matrix corresponding to each sampling data in the first noise reduction matrix;

[0020] The convolution data of each sampling data in the data matrix is obtained according to a preset weight matrix of the second convolution kernel and each second convolution matrix.

[0021] The size of the second convolution kernel is (2n+1)*(2n+1), and the weight matrix includes (2n+1)*(2n+1) weights, wherein the weight in the (n+1)th row and the (n+1)th column is in a first numerical range, the weights in the (n+1)th row and other columns are zero, the weights in the (2n+1)th row and other rows are in a second numerical range, n is an integer greater than or equal to 1, the first numerical range is greater than the second numerical range, and the sum of the (2n+1)*(2n+1) weights is 1.

[0022] The size of the second convolution kernel is (2n+1)*(2n+1), the weight matrix includes (2n+1)*(2n+1) weights, the weight of the (n+1)th row and the (n+1)th column is located in a first numerical range, the weights of the (n+1)th column in other rows except the (n+1)th row are zero, the weights of other columns except the (n+1)th column in the (2n+1) columns are located in a second numerical range, the n is an integer greater than or equal to 1, the first numerical range is greater than the second numerical range, and the sum of the (2n+1)*(2n+1) weights is 1.

[0023] The second aspect of the present application provides a screen noise reduction device, comprising:

[0024] Determine whether each convolution data is located in the second noise range;

[0025] If the convolution data is located in the second noise range, a plurality of second target data in the first noise reduction matrix with abnormal noise is obtained.

[0026] The second aspect of the present application provides a screen noise reduction device, comprising:

[0027] The acquisition module is configured to acquire sampling data and convert the sampling data into a data matrix;

[0028] The convolution module is configured to perform convolution processing on the data matrix to obtain convolution data of each sampling data in the data matrix, and obtain a plurality of first target data with abnormal noise in the data matrix according to each convolution data and a preset first noise range, wherein the plurality of first target data are not adjacent in the data matrix.

[0029] The first replacement module is configured to replace the plurality of first target data with reference data to obtain a first noise reduction matrix, wherein the reference data is data of the screen without touch and noise.

[0030] The convolution module is configured to perform convolution processing on the data matrix to obtain convolution data of each sampling data in the data matrix, and obtain a plurality of first target data with abnormal noise in the data matrix according to each convolution data and a preset first noise range, wherein the plurality of first target data are not adjacent in the data matrix.

[0031] The second replacement module is configured to obtain a plurality of second target data with abnormal noise in the first noise reduction matrix according to each convolution data and a preset second noise range, replace the plurality of second target data with convolution data corresponding to each second target data to obtain a second noise reduction matrix, and the second noise range is smaller than the first noise range.

[0032] In a third aspect, an embodiment of the present application provides a terminal device, comprising a memory, a processor, and a computer program stored in the memory and capable of running on the processor, and the processor implements the noise reduction method of the screen according to any one of the first aspect when executing the computer program.

[0033] In a fourth aspect, an embodiment of the present application provides a computer readable storage medium, which stores a computer program, and the computer program is executed by a processor to implement the noise reduction method of the screen according to any one of the first aspect.

[0034] Compared with the prior art, the embodiment of the present application has the beneficial effects that: the technical scheme of the present application obtains sampling data and converts the sampling data into a data matrix; performs a convolution-like processing on the data matrix to obtain convolution-like data of each sampling data in the data matrix, and obtains a plurality of first target data with abnormal noise in the data matrix according to each convolution-like data and a preset first noise range, the plurality of first target data are not adjacent in the data matrix; replaces the plurality of first target data with reference data to obtain a first noise reduction matrix, the reference data is data of the screen without being touched and without noise; performs a convolution processing on the first noise reduction matrix to obtain convolution data of each sampling data in the first noise reduction matrix; obtains a plurality of second target data with abnormal noise in the first noise reduction matrix according to each convolution data and a preset second noise range, and replaces the plurality of second target data with the convolution data corresponding to each second target data to obtain a second noise reduction matrix. That is, the present application can perform a convolution-like processing on the data matrix, obtain the first noise reduction matrix according to the first noise range, complete the first noise reduction of the sampling data, perform a convolution processing on the first noise reduction matrix, obtain the second noise reduction matrix according to the second noise range, and complete the second noise reduction of the sampling data, effectively filter the electromagnetic interference in the sampling data, and significantly improve the noise reduction effect. BRIEF DESCRIPTION OF DRAWINGS

[0035] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed in the embodiments or the prior art description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0036] Figure 1 is an application scenario diagram of a noise reduction method of a screen provided by an embodiment of the present application;

[0037] Figure 2a is a schematic flowchart of a noise reduction method of a screen provided by an embodiment of the present application;

[0038] Figure 2bis an example diagram of the obtained sampling data provided by an embodiment of the present application;

[0039] Figure 2c is an example diagram of the first noise reduction matrix provided by an embodiment of the present application;

[0040] Figure 3a is a schematic flow chart of a method for obtaining the first target data provided by an embodiment of the present application;

[0041] Figure 3b is an example diagram of the target matrix provided by an embodiment of the present application;

[0042] Figure 3c is an example diagram of each first convolution matrix provided by an embodiment of the present application;

[0043] Figure 4 is a schematic flow chart of a specific method of S303 provided by an embodiment of the present application;

[0044] Figure 5a is a schematic flow chart of a method for performing convolution processing on the first noise reduction matrix provided by an embodiment of the present application;

[0045] Figure 5b is an example diagram of each second convolution matrix provided by an embodiment of the present application;

[0046] Figure 5c is an example diagram of the convolution data of each sampling data in the data matrix provided by an embodiment of the present application;

[0047] Figure 5d is a schematic flow chart of another method for performing convolution processing on the noise reduction matrix provided by an embodiment of the present application;

[0048] Figure 5e is an example diagram of the determination of the deviation value provided by an embodiment of the present application.

[0049] Figure 6a is a schematic flow chart of a method for obtaining the second target data provided by an embodiment of the present application;

[0050] Figure 6b is an example diagram of the second noise reduction matrix provided by an embodiment of the present application;

[0051] Figure 6c is a grayscale diagram before and after noise reduction in a five-finger touch operation provided by an embodiment of the present application;

[0052] Figure 7 is a structural schematic diagram of a noise reduction device of a screen provided by an embodiment of the present application;

[0053] Figure 8is a structural schematic diagram of a terminal device provided by an embodiment of the present application. DETAILED DESCRIPTION

[0054] In the following description, for the purposes of explanation and not limitation, specific details are set forth, such as particular system configurations, techniques, etc., in order to provide a thorough understanding of the embodiments of the present application. However, it will be apparent to those skilled in the art that the present application can be practiced in other embodiments that depart from these specific details. In other instances, detailed descriptions of well-known systems, devices, circuits, and methods are omitted so as not to obscure the description of the present application with unnecessary detail. In other instances, well-known structures are not described in detail in order to avoid obscuring the present application.

[0055] It is to be understood that the terminology "includes", "has", "holds", "contains" and / or "comprising", "including", "containing", "having" and / or "comprises" used in the specification and in the following claims indicates the presence of the stated features, integers, steps, operations, elements, and / or components but does not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and / or groups thereof.

[0056] It is also to be understood that the terminology "and / or" used in the present specification and in the claims pages hereto refers to one or more of the associated listed items, in any combination and all possible combinations, and includes these combinations.

[0057] Reference in the specification to "an embodiment" or "some embodiments" of the present application means that a particular feature, structure, or characteristic described in connection with the embodiment is included in at least one embodiment of the present application. The appearances of the phrase "in other embodiments" in various places in the specification are not necessarily all referring to the same embodiment, however, and the appearances of the phrases "in at least one embodiment" and "in some embodiments" in various places in the specification are not necessarily all referring to the same embodiment, unless otherwise indicated. The terms "including", "containing", "comprising", "having" and variations thereof in the specification are meant to encompass the item listed thereafter, but do not exclude the presence of one or more other items.

[0058] In addition, in the description of the specification and the appended claims, the terms "first", "second", and the like are used only to distinguish descriptions, and cannot be understood as indicating or implying relative importance.

[0059] The existing intelligent touch screen will produce certain electromagnetic interference when the internal circuit is complex to a certain extent.

[0060] In actual production, under the influence of factors such as screen size, cost, etc., it is impossible to effectively filter the interference of electromagnetic signals by hardware alone.

[0061] In order to solve the above-mentioned defects, the inventive concept of the present application is:

[0062] The application can convert the obtained sampling data into a data matrix, perform convolution processing on the data matrix, obtain a first noise reduction matrix according to a first noise range, complete the first noise reduction of the sampling data, perform convolution processing on the first noise reduction matrix, obtain a second noise reduction matrix according to a second noise range, and complete the second noise reduction of the sampling data. That is, the application provides a software noise reduction method, which can solve the problem that the hardware noise reduction method cannot effectively filter electromagnetic interference, effectively filter electromagnetic interference in the sampling data, and significantly improve the beneficial effect of noise reduction effect.

[0063] In order to illustrate the technical solutions of the application, specific embodiments are described below.

[0064] Please refer to Figure 1 , Figure 1 is an application scenario diagram of a noise reduction method of a screen provided by an embodiment of the application. For the convenience of description, only the parts related to the application are shown. The application scenario includes, but is not limited to, a capacitive sensing array 10, a driving circuit 20, a detection circuit 30, and a processing unit 40. The output end of the driving circuit 20 is electrically connected to the input end of the driving circuit 20, the output end of the driving circuit 20 is electrically connected to the input end of the detection circuit 30, and the output end of the detection circuit 30 is electrically connected to the input end of the processing unit 40.

[0065] The capacitive sensing array 10 includes a plurality of sensing units 11 arranged in rows and columns. Each sensing unit 11 includes a first electrode (for example, a driving electrode) and a second electrode (for example, a receiving electrode). When a voltage signal is provided to the first electrode, an electric field is generated between the first electrode and the second electrode and a coupling capacitance is formed. The first electrode and the second electrode in the embodiment of the application can be appropriately configured and are not specifically limited, as long as a specific coupling capacitance can be formed. In the embodiment of the application, whether an object (for example, but not limited to, a finger, a water droplet, or a metal, etc.) is close to the sensing unit 11 can be detected by judging the charge change of the sensing unit 11.

[0066] The driving circuit 20 is a signal generator that can send a driving signal to the first electrode of the sensing unit 11. The driving signal in the embodiment of the application can be a time-varying signal, for example, a periodic signal. In other embodiments, the driving signal can be a pulse signal, for example, a square wave, a triangular wave, etc. The type of the pulse signal is not limited in the embodiment of the application. The driving signal can couple a detection signal to the second electrode of the sensing unit 11 through the coupling capacitance.

[0067] The driving circuit 20 in the embodiment of the application can be multiple, each of which provides a driving signal for each row of sensing units 11 of the capacitive sensing array 10. The multiple driving circuits 20 can sequentially or in parallel drive the sensing units 11.

[0068] The detection circuit 30 coupled to the capacitive sensing array 10 is configured to modulate the detection signals generated by the plurality of sensing units 11 in each row to generate modulated detection signals. The modulation is configured to change the amplitude, frequency or phase of the detection signals generated by the plurality of sensing units 11.

[0069] The processing unit 40 can be a central processing unit (CPU), and can also be other general-purpose processors, digital signal processors (DSP), application specific integrated circuits (ASIC), field programmable gate arrays (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor.

[0070] The processing unit 40 is configured to obtain the modulated detection signals (which can also be referred to as sampling data in the embodiments of the present application), and convert the sampling data into a data matrix; perform a quasi-convolution operation on the data matrix to obtain quasi-convolution data of each sampling data in the data matrix, and obtain a plurality of first target data in the data matrix having abnormal noise according to the quasi-convolution data and a preset first noise range, the plurality of first target data being non-adjacent in the data matrix; replace the plurality of first target data with reference data to obtain a first de-noising matrix; perform a convolution operation on the first de-noising matrix to obtain convolution data of each sampling data in the first de-noising matrix; obtain a plurality of second target data in the first de-noising matrix having abnormal noise according to the convolution data and a preset second noise range, and replace the plurality of second target data with the convolution data corresponding to the plurality of second target data to obtain a second de-noising matrix.

[0071] In other embodiments, more or fewer components than those shown in the example can be included, or certain components can be combined, or different components can be included. Figure 1 The above description is merely exemplary and should not be construed as a specific limitation of the present application. For example, an analog-to-digital converter, an encoder, a decoder, etc. can also be included. Figure 1 The above description is merely exemplary and should not be construed as a specific limitation of the present application. For example, an analog-to-digital converter, an encoder, a decoder, etc. can also be included.

[0072] Please refer to Figure 2a , Figure 2a is a schematic flowchart of a screen de-noising method provided by the embodiments of the present application. Figure 2a The execution subject of the method in Figure 1 may be the processing unit. As shown in Figure 2a , the method includes S201 to S205.

[0073] S201, the processing unit obtains sampling data and converts the sampling data into a data matrix.

[0074] Specifically, the sampling data refers to the modulated detection signal generated by the detection circuit.

[0075] The detection circuit inputs the modulated detection signal to the processing unit, and the processing unit can obtain the sampling data.

[0076] In the embodiment of the application, the obtained sampling data is stored in one-dimensional form, that is, the sampling data is stored in a simple ordering structure. For example, please refer to Figure 2b , Figure 2b is an example of the obtained sampling data provided by the embodiment of the application, in which Figure 2b the capacitive sensing array has 3 rows and 3 columns in total, the sampling data of the first row is N1, N2, N3 from left to right, the sampling data of the second row is N4, M, N5 from left to right, and the sampling data of the third row is N6, N7, N8 from left to right.

[0077] The obtained sampling data is stored in one-dimensional form, which can be represented as [N1, N2, N3], [N4, M, N5] and [N6, N7, N8].

[0078] The embodiment of the application converts the one-dimensional sampling data into a two-dimensional data matrix. The data matrix is composed of multiple one-dimensional sampling data, also known as table data. For example, the data matrix can be represented as:

[0079]

[0080] The embodiment of the application can use a matrix transformation function to convert the sampling data into a data matrix, for example, the matrix transformation function includes a reshape function, a resize function, etc. The embodiment of the application does not limit the method of converting the sampling data into a data matrix.

[0081] The embodiment of the application converts the sampling data into a data matrix in order to facilitate calculation and improve the calculation rate.

[0082] It should be noted that if the processing unit directly obtains a two-dimensional data matrix, the data matrix does not need to be converted.

[0083] S202, the processing unit performs convolution-like processing on the data matrix to obtain convolution-like data of each sampling data in the data matrix, and obtains multiple first target data with abnormal noise in the data matrix according to each convolution-like data and a preset first noise range.

[0084] Specifically, the multiple first target data are not adjacent in the data matrix.

[0085] The first noise range preset in the embodiment of the application is set according to the reference data, and the reference data is data of a screen that is not touched and has no noise. The reference data of different types of devices may have certain differences, and the reference data can be any value in the range of 0 bits to 255 bits, for example, the reference data is 128 bits.

[0086] The sampling data obtained in S201 in the embodiment of the application can be represented by the reference data, and the reference data can be denoted as S in the embodiment of the application, for example, if N1, N2, and N3 are values greater than S, then N1=S+H1, N2=S+H2, and N1=S+H3, and if N1, N2, and N3 are values less than S, then N1=S-H1, N2=S-H2, and N1=S-H3. Wherein, |+H1|, |+H2|, |+H3|, |-H1|, |-H2|, and |-H3| are referred to as deviation values of the sampling data and the reference data.

[0087] When the first sensing unit is touched, the sampling data will deviate from the reference data by more than 10 bits, and the processing unit identifies the first sensing unit corresponding to the sampling data with a deviation value greater than 10 as a touch point, and a deviation value within 10 will not affect the identification, so the first noise range of the embodiment of the application is [S-10, S+10].

[0088] The convolution processing of the data matrix in the embodiment of the application is to traverse each sampling data in the data matrix by using a convolution kernel, and the convolution kernel performs convolution calculation to obtain convolution data after traversing each sampling data. The convolution processing of the data matrix only traverses each sampling data in the data matrix by using a convolution kernel, and outputs the convolution data corresponding to each sampling data, without the need for convolution calculation.

[0089] After obtaining the convolution data, a plurality of first target data with abnormal noise in the data matrix is obtained according to the convolution data and the preset first noise range, so that the plurality of first target data with abnormal noise can be screened out, and then the plurality of first target data is replaced to remove the abnormal noise in the sampling data, thereby significantly improving the noise reduction effect.

[0090] Please refer to Figure 3a , Figure 3a is a schematic flowchart of a method for obtaining first target data provided by the embodiment of the application. Figure 3a The execution subject of the method in Figure 1 may be the processing unit. As shown in FIG. 3a, the method includes S301 to S303.

[0091] S301, the processing unit performs edge padding on the data matrix based on the reference data to obtain a target matrix.

[0092] Specifically, if the first convolution kernel is directly used to traverse the sampling data in the data matrix, the size of the obtained convolution matrix is different from that of the data matrix.

[0093] For example, if the size of the data matrix is n*n and the size of the first convolution kernel is f*f, the size of the obtained convolution matrix is (n-f+1)*(n-f+1) if the first convolution kernel is directly used to traverse the sampling data in the data matrix. For example, if the size of the data matrix is 6*6 and the size of the first convolution kernel is 3*3, the size of the obtained convolution matrix is 4*4 if the first convolution kernel is directly used to traverse the sampling data in the data matrix.

[0094] In order to make the size of the convolution matrix the same as that of the data matrix, the edge padding of the data matrix is performed by using the reference data in the embodiment of the present application. The data matrix after padding is referred to as a target matrix.

[0095] For example, the size of the data matrix is X*Y, that is, X rows and Y columns of data. The X*Y data matrix is padded to a matrix of (X+2) rows and (Y+2) columns, that is, a new value is added around the original data matrix to expand the size of the data matrix. The new value added by padding is the reference value S. Please refer to Figure 3b , Figure 3b is an example diagram of a target matrix provided by the embodiment of the present application. Figure 3b The size of the data matrix in the above table is 3*3, and the size of the target matrix is 5*5.

[0096] It should be noted that, in order to facilitate calculation, the new value added by padding is stored separately in the embodiment of the present application.

[0097] S302, the processing unit traverses the sampling data in the target matrix by using the first convolution kernel to obtain each first convolution matrix.

[0098] Specifically, the first convolution matrix is a matrix corresponding to each sampling data in the target matrix.

[0099] In the embodiment of the present application, the size of the first convolution kernel is at least 3*3. The size of 3*3 is the minimum convolution kernel size required to extract the complete features of the sampling data in the target matrix. If the size is less than 3*3, the extraction will be inaccurate when extracting the features of the sampling data, and the noise resistance of the sampling data will be reduced.

[0100] The size of the first convolution kernel is 3*3 in the embodiment of the present application.

[0101] Please refer to Figure 3c , Figure 3c is an example diagram of each first convolution matrix provided by the embodiment of the present application,Figure 3c In the diagram, A represents a 3x3 convolution kernel, B represents the target matrix, and C represents each of the first convolution matrices. The 3x3 convolution kernel is used to iterate through the sampled data N1, N2, N3, N4, M, N5, N6, N7, and N8 in the target matrix B to obtain the first convolution matrices C1, C2, C3, C4, C5, C6, C7, C8, and C9.

[0102] S303. The processing unit obtains multiple first target data containing abnormal noise in the data matrix according to the noise reduction conditions corresponding to the first noise range in each first convolution matrix.

[0103] Please refer to Figure 4 , Figure 4 This is a schematic flowchart illustrating a specific method of S303 provided in an embodiment of this application. Figure 4 The execution entity of the method in the middle can be Figure 1 The processing unit within. For example... Figure 4 As shown, the method includes: S401 to S402.

[0104] S401. The processing unit determines the first processing data and the second processing data in each of the first convolution matrices.

[0105] Specifically, the first processed data is the sampled data corresponding to each first convolution matrix. For example: N1 in C1, N2 in C2, N3 in C3, N4 in C4, M in C5, N5 in C6, N6 in C7, N7 in C8, and N8 in C9.

[0106] The second processing data consists of the data in each first convolution matrix excluding the first processing data. For example: S, S, S, S, N2, S, N4, M in C1; S, S, S, N1, N3, N4, M, N5 in C2; S, S, S, N2, S, M, N5, S in C3; S, N1, N2, S, M, S, N7 in C4; and N1, N2, N3, N4, N5, N6, N7, N8 in C5, etc.

[0107] S402. If the first processed data is not within the first noise range and the second processed data is within the first noise range, the processing unit obtains multiple first target data with abnormal noise in the data matrix.

[0108] Specifically, if the first processed data is not within [S-10, S+10], and the second processed data is within [S-10, S+10], then the first processed data is determined as the first target data.

[0109] For example, if N1 in C1 is not within [S-10, S+10], and S, S, S, S, N2, S, N4, and M in C1 are within [S-10, S+10], then N1 in C1 is determined as the first target data.

[0110] For example: if M in C5 is not within [S-10, S+10], and N1, N2, N3, N4, N5, N6, N7, N8 in C5 are within [S-10, S+10], then M in C5 is determined as the first target data.

[0111] By following this method, multiple first target data points containing abnormal noise in the data matrix can be obtained.

[0112] It should be noted that the above method determines whether there is abnormal noise in the sampled data by judging whether the first processed data is not within the first noise range and whether the second processed data is not within the first noise range in each first convolution matrix. However, the above method can only determine the sampled data with abnormal noise when the sampled data with abnormal noise are not adjacent. It cannot determine the sampled data with abnormal noise when two or more sampled data with abnormal noise are adjacent.

[0113] For example, suppose N1 and N2 are sampled data containing abnormal noise, according to Figure 4 The method for determining the presence of abnormal noise is as follows: N1 in the C1 convolution matrix should not be within the range [S-10, S+10], and S, S, S, S, N2, S, N4, and M in C1 should also be within the range [S-10, S+10]. N2 in the C2 convolution matrix should not be within the range [S-10, S+10], and S, S, S, N1, N3, N4, M, and N5 in C1 should also be within the range [S-10, S+10]. If the condition that N2 in C1 is within the range [S-10, S+10] conflicts with the condition that N2 in the C2 convolution matrix is ​​not within the range [S-10, S+10], the processing unit will proceed according to... Figure 4 If we use the same judgment method to judge N2, which is adjacent to N1, we will conclude that N2 is not abnormal noise. Therefore, according to... Figure 4 The judgment method can only be used to determine the presence of abnormal noise when the sampled data are not adjacent; it cannot be used when two or more sampled data with abnormal noise are adjacent.

[0114] S203, The processing unit replaces multiple first target data with reference data to obtain the first noise reduction matrix.

[0115] Specifically, the processing unit replaces multiple first target data with reference data S to obtain the first noise reduction matrix.

[0116] For example, please refer to Figure 2c , Figure 2cis an example diagram of a first denoising matrix provided by an embodiment of the present application. Assuming that N1, N3 and N6 are determined as the first target data by the method of Figure 4 , the first denoising matrix obtained is D.

[0117] S204, the processing unit performs convolution processing on the first denoising matrix to obtain convolution data of each sampling data in the first denoising matrix.

[0118] Please refer to Figure 5a , Figure 5a is a schematic flowchart of a method of performing convolution processing on a first denoising matrix provided by an embodiment of the present application. Figure 5a The execution subject of the method in Figure 1 may be the processing unit. As shown in Figure 5a , the method includes S501-S502.

[0119] S501, the processing unit traverses the sampling data in the first denoising matrix by using a second convolution kernel to obtain each second convolution matrix.

[0120] Specifically, the second convolution matrix is a matrix corresponding to each sampling data in the first denoising matrix.

[0121] In the embodiment of the present application, the size of the second convolution kernel is (2n+1)*(2n+1), and n is an integer greater than or equal to 1. The size of the second convolution kernel is taken as an example in the embodiment of the present application.

[0122] Please refer to Figure 5b , Figure 5b is an example diagram of each second convolution matrix provided by an embodiment of the present application, Figure 5b In the diagram, F represents a 3*3 convolution kernel, D represents the first denoising matrix, and E represents each second convolution matrix. The sampling data S, N2, S, N4, M, N5, S, N7 and N8 in the first denoising matrix D are traversed by using the 3*3 convolution kernel to obtain each second convolution matrix E1, E2, E3, E4, E5, E6, E7, E8 and E9.

[0123] S502, the processing unit obtains convolution data of each sampling data in the data matrix according to the weight matrix of the second convolution kernel and each second convolution matrix.

[0124] Specifically, the weight matrix of the second convolution kernel is pre-set in the processing unit, and the pre-set weight matrix is set according to the sampling characteristics of the second convolution kernel. For example, if the convolution matrix obtained by sampling through the second convolution kernel is Figure 5b E1-E9 in

[0125] In the analysis results, as in C5, if the deviation value of M from the reference data is the largest, among the eight data other than M, the deviation value of the data horizontally adjacent to M from the reference data is the largest, for example, N4 or N5, and most of the data in other convolution matrices also conform to this sampling feature.

[0126] Therefore, in order to avoid the deviation value of the data horizontally adjacent to the sampling data from the reference data being too large, and thus causing the convolution result to be too close to the reference data and to cover the characteristics of the sampling data, the weight of all the data horizontally adjacent to the sampling data is set to 0 (the weight of the column other than the (n+1)th column in the (n+1)th row is 0), the weight of the sampling data is set to any value in the first value range (the weight of the (n+1)th column in the (n+1)th row is in the first value range), in order to preserve the characteristics of the sampling data, the weight of the data other than the sampling data and the data horizontally adjacent to the sampling data in the convolution matrix is set to any value in the second value range (the weight of the row other than the (n+1)th row in the (2n+1)th row is in the second value range), in order to prevent the weight of the data other than the sampling data and the data horizontally adjacent to the sampling data in the convolution matrix from being too high and thus reducing the noise resistance, and the sum of all the weights in the convolution matrix is 1.

[0127] Based on the above method of setting the weight matrix, a preset weight matrix of a second convolution kernel in an embodiment of the present application includes (2n+1)*(2n+1) weights, the weight of the (n+1)th column in the (n+1)th row is in the first value range, the weight of the column other than the (n+1)th column in the (n+1)th row is 0, the weight of the row other than the (n+1)th row in the (2n+1)th row is in the second value range, the first value range is larger than the second value range, and the sum of the (2n+1)*(2n+1) weights is 1.

[0128] In the embodiment of the present application, the first value range is [0.4, 0.6], and the second value range is [0.05, 0.2].

[0129] For example, the weight matrix includes 3*3 weights, the weight of the 2th column in the 2th row is in the first value range, the weight of the column other than the 2th column in the 2th row is 0, the weight of the row other than the 2th row in the 3th row is in the second value range, and the sum of the 3*3 weights is 1.

[0130] In C5, if the deviation value of M from the reference data is the largest, among the eight data other than M, the deviation value of the data vertically adjacent to M from the reference data is the largest, for example, N2 or N7, and most of the data in other convolution matrices also conform to this sampling feature.

[0131] In order to avoid the situation that the deviation values of the data vertically adjacent to the sampling data from the reference data are too large, and the convolution result is too close to the reference data and the characteristics of the sampling data are submerged, the weight of all data vertically adjacent to the sampling data is set to 0 (the weight of the other rows except the (n+1) row in the (n+1) column is zero), the weight of the sampling data is set to any value in the first value range (the weight of the (n+1) column in the (n+1) row is in the first value range), in order to prevent the weight of the data except the sampling data and all data vertically adjacent to the sampling data in the convolution matrix from being too high and reduce the noise resistance, the weight of the data is set to any value in the second value range (the weight of the other columns except the (n+1) column in the (2n+1) column is in the second value range), and the sum of all weights in the convolution matrix is 1.

[0132] Based on the above method of setting the weight matrix, in another preset weight matrix of the second convolution kernel in the embodiment of the application, the weight of the (n+1) column in the (n+1) row is in the first value range, the weight of the other rows except the (n+1) row in the (n+1) column is zero, and the weight of the other columns except the (n+1) column in the (2n+1) column is in the second value range.

[0133] For example, the weight of the (2) column in the (2) row is in the first value range, the weight of the other rows except the (2) row in the (2) column is zero, and the weight of the other columns except the (2) column in the (3) column is in the second value range.

[0134] In the embodiment of the application, the processing unit can substitute the preset weight matrix and the sampling data in each second convolution matrix into the convolution formula, and the convolution data of each sampling data in the data matrix can be obtained.

[0135] For example, if the size of the data matrix is 3*3, the convolution formula is:

[0136] cov=K 11 *N1+K 12 *N2+K 13 *N3+K 21 *N4+K 22 *M+K 23 *N5+K 31 *N6+K 32 *N7+K 33 *N8。

[0137] Wherein, cov represents the convolution result, K 11 represents the weight of the first column in the first row of the weight matrix, K 12 represents the weight of the second column in the first row of the weight matrix.13 K represents the weight in the first row and third column of the weight matrix. 21 K represents the weight in the second row and first column of the weight matrix. 22 K represents the weight in the second row and second column of the weight matrix. 23 K represents the weight in the second row and third column of the weight matrix. 31 K represents the weight in the third row and first column of the weight matrix. 32 K represents the weight in the third row and second column of the weight matrix. 33 This represents the weight in the third row and third column of the weight matrix.

[0138] For example, please refer to Figure 5c , Figure 5c This is an example diagram of convolutional data of each sampled data in a data matrix provided in an embodiment of this application. Figure 5c In the diagram, F represents the convolutional data of each sampled data in the data matrix. Specifically, F1 represents the convolutional data calculated by convolving E1, F2 represents the convolutional data calculated by convolving E2, F3 represents the convolutional data calculated by convolving E3, F4 represents the convolutional data calculated by convolving E4, F5 represents the convolutional data calculated by convolving E5, F6 represents the convolutional data calculated by convolving E6, F7 represents the convolutional data calculated by convolving E7, F8 represents the convolutional data calculated by convolving E8, and F9 represents the convolutional data calculated by convolving E9.

[0139] In this embodiment, the processing unit performs convolution processing on the first noise reduction matrix to obtain convolution data of each sampled data in the first noise reduction matrix. This is to reduce the noise of each sampled data in the first noise reduction matrix while retaining the basic characteristics of each sampled data in the first noise reduction matrix.

[0140] Please refer to Figure 5d , Figure 5d This is a schematic flowchart illustrating another method for convolution processing of a noise reduction matrix provided in an embodiment of this application. Figure 5d The execution entity of the method in the middle can be Figure 1 The processing unit within. For example... Figure 5d As shown, the method includes: S503 to S505.

[0141] S503, The processing unit uses the second convolution kernel to traverse the sampled data in the first noise reduction matrix to obtain each second convolution matrix.

[0142] Specifically, the method for obtaining each second convolution matrix in S503 is the same as in S501, and will not be repeated here.

[0143] S504, the processing unit determines the weight of the second convolution kernel assigned to the second convolution matrix.

[0144] The method for determining the weight of the second convolution kernel assigned to the second convolution matrix in the embodiments of the present application is as follows:

[0145] Firstly, in each second convolution matrix, the third processing data and the fourth processing data are determined.

[0146] The third processing data is the sampling data corresponding to each second convolution matrix. For example, S in the second row and the second column of E1, N2 in E2, S in the second row and the second column of E3, N4 in E4, M in E5, N5 in E6, S in the second row and the second column of E7, N7 in E8, and N8 in E9.

[0147] The fourth processing data is the data in each second convolution matrix except the third processing data. For example, the data in E1 except S in the second row and the second column, S, S, S, S, S, N4, M, and N5 in E2, the data in E3 except S in the second row and the second column, and so on.

[0148] Secondly, the deviation value of the fourth processing data and the reference data is determined, and the weight of the second convolution kernel assigned to the second convolution matrix is determined according to the deviation value.

[0149] Specifically, in the embodiments of the present application, when the deviation value of the fourth processing data and the reference data is determined, it is first determined whether the fourth processing data is greater than the reference data or less than or equal to the reference value, that is, the fourth processing data greater than the reference data is determined and the fourth processing data less than or equal to the reference data is determined. Then, each fourth processing data is subtracted from the reference data to determine the deviation value. For example, please refer to Figure 5e , Figure 5e is an example diagram for determining the deviation value provided by the embodiments of the present application.

[0150] Figure 5e The second convolution matrix shown in the above table is E5. Please refer to E 5a, the fourth processing data corresponding to N1 is 129, N1 is the fourth processing data greater than the reference data, and the deviation value of N1 from the reference data is |129-128|=1. The fourth processing data corresponding to N2 is 130, N2 is the fourth processing data greater than the reference data, and the deviation value of N2 from the reference data is |130-128|=2. The fourth processing data corresponding to S is 128, S is the fourth processing data equal to the reference data, and the deviation value of S from the reference data is |128-128|=0. The fourth processing data corresponding to N4 is 121, N4 is the fourth processing data less than the reference data, and the deviation value of N4 from the reference data is |121-128|=7. The fourth processing data corresponding to N5 is 119, N5 is the fourth processing data less than the reference data, and the deviation value of N5 from the reference data is |119-128|=9. The fourth processing data corresponding to S is 128, S is the fourth processing data equal to the reference data, and the deviation value of S from the reference data is |128-128|=0. The fourth processing data corresponding to N7 is 132, N7 is the fourth processing data greater than the reference data, and the deviation value of N7 from the reference data is |132-128|=4. The fourth processing data corresponding to N8 is 135, N8 is the fourth processing data greater than the reference data, and the deviation value of N8 from the reference data is |135-128|=7.

[0151] Please refer to E 5b , the fourth processing data corresponding to N1 is 129, N1 is the fourth processing data greater than the reference data, and the deviation value of N1 from the reference data is |129-128|=1. The fourth processing data corresponding to N2 is 119, N2 is the fourth processing data less than the reference data, and the deviation value of N2 from the reference data is |119-128|=9. The fourth processing data corresponding to S is 128, S is the fourth processing data equal to the reference data, and the deviation value of S from the reference data is |128-128|=0. The fourth processing data corresponding to N4 is 121, N4 is the fourth processing data less than the reference data, and the deviation value of N4 from the reference data is |121-128|=7. The fourth processing data corresponding to N5 is 130, N5 is the fourth processing data greater than the reference data, and the deviation value of N5 from the reference data is |130-128|=2. The fourth processing data corresponding to S is 128, S is the fourth processing data equal to the reference data, and the deviation value of S from the reference data is |128-128|=0. The fourth processing data corresponding to N7 is 132, N7 is the fourth processing data greater than the reference data, and the deviation value of N7 from the reference data is |132-128|=4. The fourth processing data corresponding to N8 is 135, N8 is the fourth processing data greater than the reference data, and the deviation value of N8 from the reference data is |135-128|=7.

[0152] In the embodiment of the present application, the maximum deviation value is determined from the deviation values of the fourth processing data greater than the reference data and the reference data. For example, in the embodiment of the present application, the maximum deviation value is 7, and the corresponding fourth processing data is N8. 5a In the embodiment of the present application, the maximum deviation value is determined from the deviation values of the fourth processing data greater than the reference data and the reference data. For example, in the embodiment of the present application, the maximum deviation value is 7, and the corresponding fourth processing data is N8.

[0153] For another example, in the embodiment of the present application, the maximum deviation value is 7, and the corresponding fourth processing data is N8. 5b In the embodiment of the present application, the maximum deviation value is determined from the deviation values of the fourth processing data greater than the reference data and the reference data. For example, in the embodiment of the present application, the maximum deviation value is 7, and the corresponding fourth processing data is N8.

[0154] In the embodiment of the present application, in order to avoid the situation that the deviation value of the fourth processing data and the reference data is too large, and the convolution result is too close to the reference data when the third processing data corresponding to the sampling data is convoluted, and the characteristics of the sampling data are submerged. The second convolution kernel is the weight assigned to the fourth processing data corresponding to the maximum deviation value in the deviation values of the fourth processing data greater than the reference data and the reference data (for example, N8 corresponding to 7 in the embodiment of the present application) and the fourth processing data corresponding to the maximum deviation value in the deviation values of the fourth processing data less than or equal to the reference data and the reference data (for example, N5 corresponding to 9 in the embodiment of the present application, N2 corresponding to 9 in the embodiment of the present application). 5a and E 5b In the embodiment of the present application, in order to avoid the situation that the deviation value of the fourth processing data and the reference data is too large, and the convolution result is too close to the reference data when the third processing data corresponding to the sampling data is convoluted, and the characteristics of the sampling data are submerged. The second convolution kernel is the weight assigned to the fourth processing data corresponding to the maximum deviation value in the deviation values of the fourth processing data greater than the reference data and the reference data (for example, N8 corresponding to 7 in the embodiment of the present application) and the fourth processing data corresponding to the maximum deviation value in the deviation values of the fourth processing data less than or equal to the reference data and the reference data (for example, N5 corresponding to 9 in the embodiment of the present application, N2 corresponding to 9 in the embodiment of the present application). 5a 5b In the embodiment of the present application, in order to avoid the situation that the deviation value of the fourth processing data and the reference data is too large, and the convolution result is too close to the reference data when the third processing data corresponding to the sampling data is convoluted, and the characteristics of the sampling data are submerged. The second convolution kernel is the weight assigned to the fourth processing data corresponding to the maximum deviation value in the deviation values of the fourth processing data greater than the reference data and the reference data (for example, N8 corresponding to 7 in the embodiment of the present application) and the fourth processing data corresponding to the maximum deviation value in the deviation values of the fourth processing data less than or equal to the reference data and the reference data (for example, N5 corresponding to 9 in the embodiment of the present application, N2 corresponding to 9 in the embodiment of the present application).

[0155] In the embodiment of the present application, in order to retain the characteristics of the sampling data corresponding to the third processing data, the second convolution kernel assigns the weight of the third processing data (for example, M) to the first numerical range.

[0156] In the embodiment of the present application, in order to prevent the weight of the fourth processing data from being too high and reducing the noise resistance, the second convolution kernel assigns the weight of other fourth processing data (for example, the sampling data other than M and N5 in the embodiment of the present application, the sampling data other than M and N2 in the embodiment of the present application) to the second numerical range. 5a 5b In the embodiment of the present application, in order to prevent the weight of the fourth processing data from being too high and reducing the noise resistance, the second convolution kernel assigns the weight of other fourth processing data (for example, the sampling data other than M and N5 in the embodiment of the present application, the sampling data other than M and N2 in the embodiment of the present application) to the second numerical range.

[0157] ​​Based on the above weight distribution method, in the embodiment of the application, the weight of the fourth processing data greater than the reference data and corresponding to the maximum deviation value of the deviation value of the fourth processing data from the reference data and the weight of the fourth processing data less than or equal to the reference data and corresponding to the maximum deviation value of the deviation value of the fourth processing data from the reference data are zero, and the weight of the second convolution kernel for other fourth processing data is in the second value range.

[0158] Specifically, the first value range is greater than the second value range, and the other fourth processing data is data other than the fourth processing data greater than the reference data and corresponding to the maximum deviation value of the deviation value of the fourth processing data from the reference data and the fourth processing data less than or equal to the reference data and corresponding to the maximum deviation value of the deviation value of the fourth processing data from the reference data.

[0159] In the embodiment of the application, the first value range is [0.4, 0.6], and the second value range is [0.05, 0.2].

[0160] S505, the processing unit obtains the convolution data of each sampling data in the data matrix according to the weight and each second convolution matrix.

[0161] Specifically, the second convolution kernel assigns weights to each second convolution matrix respectively by using the method in S504, and the processing unit can substitute the weight and the sampling data in each second convolution matrix into the convolution formula to obtain the convolution data of each sampling data in the data matrix.

[0162] For example, if the size of the data matrix is 3*3, the convolution formula is:

[0163] cov=K 11 *N1+K 12 *N2+K 13 *N3+K 21 *N4+K 22 *M+K 23 *N5+K 31 *N6+K 32 *N7+K 33 *N8。

[0164] Wherein, cov represents the convolution result, K 11 represents the weight of the first row and the first column, K 12 represents the weight of the first row and the second column, K 13 represents the weight of the first row and the third column, K 21 represents the weight of the second row and the first column, K 22 represents the weight of the second row and the second column, K 23 represents the weight of the second row and the third column, K31 K represents the weight of the third row and first column. 32 K represents the weight in the third row and second column. 33 This represents the weight of the third row and third column.

[0165] In this embodiment, the processing unit performs convolution processing on the first noise reduction matrix to obtain convolution data of each sampled data in the first noise reduction matrix. This is to reduce the noise of each sampled data in the first noise reduction matrix while retaining the basic characteristics of each sampled data in the first noise reduction matrix.

[0166] S205. The processing unit obtains multiple second target data containing abnormal noise in the first noise reduction matrix based on each convolutional data and the preset second noise range, and replaces the multiple second target data with the convolutional data corresponding to each second target data to obtain the second noise reduction matrix.

[0167] Please refer to Figure 6a , Figure 6a This is a schematic flowchart illustrating a method for obtaining second target data provided in an embodiment of this application. Figure 6a The execution entity of the method in the middle can be Figure 1 The processing unit in the process. As shown in Figure 6a, the method includes: S601 to S602.

[0168] S601, The processing unit determines whether each convolutional data is within the second noise range.

[0169] Specifically, the convolutional data (e.g., F1, F2, etc.) after convolution processing is closer to the baseline data than the data in the first noise reduction matrix D before convolution processing. Therefore, there is a possibility of misjudgment if the sampling data corresponding to the convolutional data is judged by the first noise range. In order to reduce the probability of misjudgment, this application embodiment determines whether each convolutional data is located within the second noise range when judging whether the sampling data corresponding to each convolutional data is within the second noise range, which is smaller than the first noise range.

[0170] S602. If the convolutional data is within the second noise range, the processing unit obtains multiple second target data with abnormal noise in the first noise reduction matrix.

[0171] For example, if F2, F4 and F5 are within the second noise range, then N2 in the first noise reduction matrix corresponding to F2, N4 in the first noise reduction matrix corresponding to F4 and M in the first noise reduction matrix corresponding to F5 are determined as multiple second target data with abnormal noise.

[0172] In this embodiment, multiple second target data are replaced with convolutional data corresponding to each second target data (N2 corresponds to F2, N4 corresponds to F4, and M corresponds to F5). This yields the second denoising matrix. An example diagram of the second denoising matrix can be found in [reference needed]. Figure 6b , Figure 6b This is an example diagram of a second noise reduction matrix provided in an embodiment of this application.

[0173] This embodiment converts the sampled data in the second noise reduction matrix into grayscale values, allowing for a more intuitive observation of the noise reduction effect in the form of an image. Please refer to... Figure 6c , Figure 6c This application provides a grayscale image before and after noise reduction for a five-finger touch operation. Wherein, A represents the grayscale image before noise reduction, and a represents the grayscale image after noise reduction.

[0174] Depend on Figure 6c It can be seen that after noise reduction, image a effectively filters out a large amount of noise, and the noise reduction effect is obvious. At the same time, it also preserves the characteristics of the sampled data (such as the sampled data during touch).

[0175] Please refer to Table 1, which is a distribution table of sampled data before and after denoising using the denoising method of this application, according to an embodiment of this application. In Table 1, a total of 1254 frames of sampled data under non-touch conditions are sampled, with 120*216 sampled data per frame. Therefore, there are a total of 1254*120*216 sampled data in Table 1.

[0176] Table 1

[0177] Sample data Before denoising After denoising Sample data Before denoising After denoising 107 3 0 149 1 1 108 0 0 148 0 0 109 6 0 147 2 0 110 3 0 146 3 0 111 6 0 145 2 0 112 10 0 144 7 0 113 22 0 143 14 0 114 52 0 142 25 0 115 143 0 141 139 0 116 327 0 140 591 0 117 701 0 139 1142 0 118 1166 0 138 1627 0 119 2549 1 137 2849 0 120 6639 1 136 4514 0 121 18665 14 135 8005 0 122 73806 136 134 22209 5 123 306093 2922 133 88106 234 124 1043369 53908 132 343564 1449 125 2643149 750795 131 1162146 6952 126 4986395 6844111 130 2922197 67113 127 6738541 15368525 129 5166625 1054264 128 6958267 8353246 128 6958267 8353246

[0178] As shown in Table 1, the noise reduction method in the embodiments of this application can effectively filter out the noise in the range of (149, 138) and (107, 118) of the sampled data. That is, the noise with a high deviation value (deviation value greater than 10) is transformed into noise within the allowable noise range through noise reduction. At this time, the noise will not cause difficulties for the subsequent processing of the sampled data.

[0179] In order to test the noise reduction performance of the noise reduction algorithm provided in this application, 120*216 data points, each with a value of 128, were collected on a screen with a resolution of 120*216. Noise with a deviation value within [10, 20] was randomly added. The distribution of the 120*216 data points before and after noise reduction is shown in Table 2.

[0180] Table 2

[0181]

[0182]

[0183] As shown in Table 2, the sampling data added with noise is completely filtered out by the noise reduction method of the embodiment of the present application, and the noise resistance of the screen is effectively improved.

[0184] In summary, in the technical solution of the present application, the sampling data is obtained, and the sampling data is converted into a data matrix; the data matrix is subjected to a pseudo-convolution process to obtain pseudo-convolution data of each sampling data in the data matrix, and a plurality of first target data existing abnormal noise in the data matrix is obtained according to each pseudo-convolution data and a preset first noise range, and the plurality of first target data is not adjacent in the data matrix; the plurality of first target data is replaced by reference data to obtain a first noise reduction matrix, and the reference data is data of the screen not being touched and having no noise; the first noise reduction matrix is subjected to a convolution process to obtain convolution data of each sampling data in the first noise reduction matrix; a plurality of second target data existing abnormal noise in the first noise reduction matrix is obtained according to each convolution data and a preset second noise range, and the plurality of second target data is replaced by the convolution data corresponding to each second target data to obtain a second noise reduction matrix. That is, the present application can perform a pseudo-convolution process on the data matrix, and obtain the first noise reduction matrix according to the first noise range to complete the first noise reduction of the sampling data, and perform a convolution process on the first noise reduction matrix, and obtain the second noise reduction matrix according to the second noise range to complete the second noise reduction of the sampling data, effectively filter the electromagnetic interference in the sampling data, and significantly improve the noise reduction effect.

[0185] It should be understood that the size of the serial number of each step in the above embodiment does not mean the order of execution, and the execution order of each process should be determined according to its function and internal logic, and should not constitute any limitation on the implementation process of the embodiment of the present application.

[0186] Please refer to Figure 7 , Figure 7 is a structural schematic diagram of a noise reduction device of a screen provided by the embodiment of the present application, and the device comprises:

[0187] The acquisition module 71 is configured to acquire sampling data and convert the sampling data into a data matrix.

[0188] The pseudo-convolution module 72 is configured to perform a pseudo-convolution process on the data matrix to obtain pseudo-convolution data of each sampling data in the data matrix, and obtain a plurality of first target data existing abnormal noise in the data matrix according to each pseudo-convolution data and a preset first noise range, and the plurality of first target data is not adjacent in the data matrix.

[0189] The first replacement module 73 is configured to replace the plurality of first target data with reference data to obtain a first noise reduction matrix, and the reference data is data of the screen not being touched and having no noise.

[0190] The convolution module 74 is configured to perform convolution processing on the first denoising matrix to obtain convolution data of each sampling data in the first denoising matrix.

[0191] The second replacing module 75 is configured to obtain a plurality of second target data with abnormal noise in the first denoising matrix according to each convolution data and a preset second noise range, replace the plurality of second target data with corresponding convolution data of each second target data, and obtain a second denoising matrix, the second noise range being smaller than the first noise range.

[0192] The convolution module 72 is further configured to perform edge padding on the data matrix based on the reference data to obtain a target matrix.

[0193] The first convolution kernel is used to traverse the sampling data in the target matrix to obtain each first convolution matrix, the first convolution matrix being a matrix corresponding to each sampling data in the target matrix.

[0194] In each first convolution matrix, a plurality of first target data with abnormal noise in the data matrix are obtained according to a denoising condition corresponding to the first noise range.

[0195] The convolution module 72 is further configured to determine first processing data and second processing data in each first convolution matrix, the first processing data being sampling data corresponding to each first convolution matrix, and the second processing data being data other than the first processing data in each first convolution matrix.

[0196] If the first processing data is not within the first noise range and the second processing data is within the first noise range, the plurality of first target data with abnormal noise in the data matrix are obtained.

[0197] The convolution module 74 is further configured to traverse the sampling data in the first denoising matrix using a second convolution kernel to obtain each second convolution matrix, the second convolution matrix being a matrix corresponding to each sampling data in the first denoising matrix.

[0198] The convolution data of each sampling data in the data matrix is obtained according to a preset weight matrix of the second convolution kernel and each second convolution matrix.

[0199] In the convolution module 74, the size of the second convolution kernel is (2n+1)*(2n+1), and the weight matrix includes (2n+1)*(2n+1) weights, wherein the weight in the (n+1)th row and the (n+1)th column is located in a first numerical range, the weights in the (m+1)th row and other columns except the (m+1)th column are zero, the weights in other rows except the (n+1)th row in the (2n+1) rows are located in a second numerical range, n is an integer greater than or equal to 1, the first numerical range is greater than the second numerical range, and the sum of the (2n+1)*(2n+1) weights is 1.

[0200] The size of the second convolution kernel in the convolution module 74 is (2n+1)*(2n+1), and the weight matrix includes (2n+1)*(2n+1) weights, wherein the weight of the (n+1)th row and the (n+1)th column is located in a first numerical range, the weights of the (n+1)th column in other rows except the (n+1)th row are zero, the weights of (2n+1) columns except the (n+1)th column are located in a second numerical range, n is an integer greater than or equal to 1, the first numerical range is greater than the second numerical range, and the sum of the (2n+1)*(2n+1) weights is 1.

[0201] The second replacement module 75 is further configured to determine whether each convolution data is located in the second noise range.

[0202] If the convolution data is located in the second noise range, a plurality of second target data with abnormal noise in the first noise reduction matrix is obtained.

[0203] Those skilled in the art can clearly understand that, for the convenience and brevity of description, only the division of the above functional units and modules is exemplified, and in actual application, the above functions can be completed by different functional units and modules according to needs, that is, the internal structure of the device is divided into different functional units or modules to complete all or part of the functions described above. Each functional unit and module in the embodiment can be integrated in one processing unit, or each unit can be physically present separately, or two or more units can be integrated in one unit. The integrated unit can be realized in the form of hardware or software. In addition, the specific names of each functional unit and module are only for easy distinction, and do not limit the protection scope of the present application. The specific working process of the units and modules in the system can refer to the corresponding process in the foregoing method embodiments, which will not be repeated here.

[0204] As shown in Figure 8 The present application also provides a terminal device 200, which includes a memory 21, a processor 22, and a computer program 23 stored in the memory 21 and executable on the processor 22. When the processor 22 executes the computer program 23, the screen noise reduction method of each embodiment described above is realized.

[0205] The processor 22 can be a central processing unit (CPU), and can also be other general-purpose processors, a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field-programmable gate array (FPGA) or other programmable logic device, discrete gate or transistor logic, discrete hardware components, etc. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor.

[0206] The memory 21 can be an internal storage unit of the terminal device 200. The memory 21 can also be an external storage device of the terminal device 200, such as a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, etc. Further, the memory 21 can include both an internal storage unit and an external storage device of the terminal device 200. The memory 21 is used to store computer programs and other programs and data required by the terminal device 200. The memory 21 can also be used to temporarily store data that has been output or will be output.

[0207] The embodiments of the present application further provide a computer readable storage medium, which stores a computer program. The computer program is executed by a processor to implement the screen noise reduction method of the above-mentioned embodiments.

[0208] The embodiments of the present application provide a computer program product. When the computer program product is run on a mobile terminal, the mobile terminal is caused to implement the screen noise reduction method of the above-mentioned embodiments.

[0209] The integrated unit, if implemented in the form of a software function unit and sold or used as an independent product, can be stored in a computer readable storage medium. Based on such understanding, all or part of the processes in the above-mentioned embodiment methods can be completed by a computer program instructing related hardware, and the computer program can be stored in a computer readable storage medium. The computer program can implement the steps of each method embodiment when executed by a processor. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or some intermediate forms. The computer readable storage medium at least includes any entity or device capable of carrying the computer program code to the photographing device / terminal equipment, recording medium, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signal, telecommunication signal, and software distribution medium. For example, U disk, mobile hard disk, magnetic disk or optical disk, etc. In some jurisdictions, according to legislation and patent practice, the computer readable storage medium can not be an electrical carrier signal and a telecommunication signal.

[0210] In the above embodiments, the description of each embodiment has its own focus, and the parts not described or recorded in detail in a certain embodiment can be referred to the related description of other embodiments.

[0211] Those skilled in the art can realize that the units and algorithm steps of each example described in combination with the embodiments disclosed herein can be realized by electronic hardware or a combination of computer software and electronic hardware. Whether the functions are executed in hardware or software depends on the specific application and design constraints of the technical solution. Professional technicians can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of the present application.

[0212] The units described as separate components can or can not be physically separated, and the components shown as units can or can not be physical units, that is, they can be located in one place, or they can be distributed on multiple network units. Part or all of the units can be selected according to actual needs to achieve the purpose of the embodiments of the present application.

[0213] The above examples are only used to illustrate the technical solutions of the present application, but not limit the same; although the present application has been described in detail with reference to the foregoing examples, those skilled in the art should understand that the technical solutions recorded in the foregoing examples can be modified, or some technical features can be replaced by equivalent ones; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application, and should be included in the protection scope of the present application.

Claims

1. A method of noise reduction of a screen, characterized in that, The method comprises the following steps: acquiring sampling data and converting the sampling data into a data matrix; performing a pseudo-convolution operation on the data matrix to obtain pseudo-convolution data of each sampling data in the data matrix, and obtaining a plurality of first target data with abnormal noise in the data matrix according to each pseudo-convolution data and a preset first noise range, wherein the plurality of first target data are not adjacent in the data matrix; replacing the plurality of first target data with reference data to obtain a first noise-reduced matrix, wherein the reference data are data of the screen without touch control and noise; performing a convolution operation on the first noise-reduced matrix to obtain convolution data of each sampling data in the first noise-reduced matrix; obtaining a plurality of second target data with abnormal noise in the first noise-reduced matrix according to each convolution data and a preset second noise range, and replacing the plurality of second target data with convolution data corresponding to each second target data to obtain a second noise-reduced matrix, wherein the second noise range is smaller than the first noise range.

2. The noise reduction method of claim 1, wherein, The method of performing a pseudo-convolution operation on the data matrix to obtain pseudo-convolution data of each sampling data in the data matrix, and obtaining a plurality of first target data with abnormal noise in the data matrix according to each pseudo-convolution data and a preset first noise range comprises the following steps: performing edge filling on the data matrix based on the reference data to obtain a target matrix; traversing the sampling data in the target matrix by using a first convolution kernel to obtain a plurality of first convolution matrices, wherein each first convolution matrix is a matrix corresponding to each sampling data in the target matrix; in each first convolution matrix, obtaining a plurality of first target data with abnormal noise in the data matrix according to a noise reduction condition corresponding to the first noise range.

3. The noise reduction method of claim 2, wherein, The method of obtaining a plurality of first target data with abnormal noise in the data matrix according to a noise reduction condition corresponding to the first noise range in each first convolution matrix comprises the following steps: in each first convolution matrix, determining first processing data and second processing data, wherein the first processing data is the sampling data corresponding to each first convolution matrix, and the second processing data is data in each first convolution matrix except the first processing data; if the first processing data is not within the first noise range and the second processing data is within the first noise range, obtaining a plurality of first target data with abnormal noise in the data matrix.

4. The noise reduction method of claim 2, wherein, The method of performing a convolution operation on the first noise-reduced matrix to obtain convolution data of each sampling data in the first noise-reduced matrix comprises the following steps: traversing the sampling data in the first noise-reduced matrix by using a second convolution kernel to obtain a plurality of second convolution matrices, wherein each second convolution matrix is a matrix corresponding to each sampling data in the first noise-reduced matrix; obtaining convolution data of each sampling data in the data matrix according to a preset weight matrix of the second convolution kernel and each second convolution matrix.

5. The noise reduction method of claim 4, wherein, The size of the second convolution kernel is (2n+1)*(2n+1), and the weight matrix includes (2n+1)*(2n+1) weights, wherein the weight of the (n+1)th column in the (n+1)th row is located in a first numerical range, the weights of the other columns except the (n+1)th column in the (n+1)th row are zero, the weights of the other rows except the (n+1)th row in the (2n+1)th column are located in a second numerical range, the n is an integer greater than or equal to 1, the first numerical range is greater than the second numerical range, and the sum of the (2n+1)*(2n+1) weights is 1.

6. The noise reduction method of claim 4, wherein, The size of the second convolution kernel is (2n+1)*(2n+1), and the weight matrix includes (2n+1)*(2n+1) weights, wherein the weight of the (n+1)th column in the (n+1)th row is located in a first numerical range, the weights of the other columns except the (n+1)th column in the (n+1)th row are zero, the weights of the other rows except the (n+1)th row in the (2n+1)th column are located in a second numerical range, the n is an integer greater than or equal to 1, the first numerical range is greater than the second numerical range, and the sum of the (2n+1)*(2n+1) weights is 1.

7. The noise reduction method according to any one of claims 1 to 6, characterized in that, The second target data in the first denoising matrix with abnormal noise is obtained according to each convolution data and a preset second noise range. It is determined whether each convolution data is located in the second noise range. If the convolution data is located in the second noise range, the second target data in the first denoising matrix with abnormal noise is obtained.

8. A noise reduction device for a screen, characterized by It includes: An acquisition module is configured to acquire sampling data and convert the sampling data into a data matrix. A convolution module is configured to perform convolution processing on the data matrix to obtain convolution data of each sampling data in the data matrix, and obtain first target data with abnormal noise in the data matrix according to each convolution data and a preset first noise range, wherein the first target data are not adjacent in the data matrix. A first replacement module is configured to replace the first target data with reference data to obtain a first denoising matrix, wherein the reference data is data of the screen without touch control and noise. A convolution module is configured to perform convolution processing on the first denoising matrix to obtain convolution data of each sampling data in the first denoising matrix. A second replacement module is configured to obtain second target data with abnormal noise in the first denoising matrix according to each convolution data and a preset second noise range, and replace the second target data with corresponding convolution data of the second target data to obtain a second denoising matrix, wherein the second noise range is smaller than the first noise range.

9. A terminal device, comprising: The screen denoising method includes a memory, a processor, and a computer program stored in the memory and executable on the processor, and the processor executes the computer program to realize the screen denoising method of any one of claims 1 to 7.

10. A computer-readable storage medium storing a computer program, the computer program comprising instructions that, when executed by a computer, cause the computer to perform the method of any one of claims 1 to 9. The computer program, when executed by the processor, implements the screen noise reduction method as claimed in any one of claims 1 to 7.

Images