Touch-control area detection method and apparatus, and electronic device, storage medium and program product

Abstract

Provided in the embodiments of the present application are a touch-control area detection method and apparatus, and an electronic device, a storage medium and a program product. The touch-control area detection method comprises: acquiring first touch-control data obtained by a user performing a touch-control operation on a touch-control screen, wherein the first touch-control data at least comprises touch-control data with regard to an intended touch-control area of the touch-control screen; and on the basis of the first touch-control data, performing touch-control area detection by means of a machine learning model, so as to obtain coordinate information of the intended touch-control area corresponding to the first touch-control data. The solution achieves a better effect in terms of detecting a touch-control area of a touch-control screen.

Description

Touch area detection methods, devices, electronic devices, storage media and software products Technical Field

[0001] This application relates to the field of touch technology, and in particular to a touch area detection method, apparatus, electronic device, storage medium, and program product. Background Technology

[0002] With the rapid popularization and continuous development of smartphones and other mobile devices, touchscreens have become an indispensable part of them. Full-screen design is gradually becoming a major trend in mobile devices, providing users with a better visual and user experience. However, this design also brings new challenges. For example, due to the increased screen-to-body ratio, users are more prone to accidental touches when holding the phone, such as unintentionally touching the screen edge with their fingers or palms. These accidental touches can lead to a series of negative consequences that degrade the user experience. Therefore, how to accurately detect touch areas to suppress the negative consequences of accidental touches and improve the user experience of using touchscreens has become an urgent problem to be solved. Summary of the Invention

[0003] In view of this, embodiments of this application provide a touch area detection method, apparatus, electronic device, storage medium, and program product.

[0004] According to a first aspect of the embodiments of this application, a touch area detection method is provided, comprising: acquiring first touch data obtained by a user performing a touch operation on a touch screen, wherein the first touch data includes at least touch data for an intended touch area of ​​the touch screen; and based on the first touch data, performing touch area detection through a machine learning model to obtain coordinate information of the intended touch area corresponding to the first touch data.

[0005] In some alternative embodiments, the first touch data includes touch data for the intended touch area of ​​the touchscreen and touch data for the edge touch area.

[0006] In some optional embodiments, the step of detecting the touch area using a machine learning model based on the first touch data to obtain the coordinate information of the intended touch area corresponding to the first touch data includes: transforming the first touch data according to reference touch data to obtain second touch data, wherein the reference touch data is data that has been determined to be able to be detected by the machine learning model to obtain the coordinate information of the intended touch area corresponding to the reference touch data; and inputting the second touch data into the machine learning model to perform touch area detection to obtain the coordinate information of the intended touch area corresponding to the first touch data.

[0007] In some optional embodiments, the step of transforming the first touch data according to the reference touch data to obtain the second touch data includes: transforming the first touch data according to the reference touch data and preset transformation parameters using an affine transformation algorithm or a second machine learning model to obtain the second touch data.

[0008] In some optional embodiments, the preset transformation parameters are predetermined by: acquiring registered touch data obtained by the user performing touch operations on the touch screen, wherein the registered touch data includes at least touch data for the intended touch area of ​​the touch screen; performing at least one round of transformation processing on the registered touch data, and obtaining the preset transformation parameters from the processing result of the at least one round of transformation processing according to the relationship between the at least one round of transformation processing and a preset transformation termination condition.

[0009] In some optional embodiments, obtaining preset transformation parameters from the processing results of the at least one round of transformation processing based on the relationship between the at least one round of transformation processing and a preset transformation termination condition includes: for each round of transformation processing, determining whether the transformation result obtained by performing the transformation processing on the registered touch data in that round meets a similarity standard with the reference touch data; if the similarity standard is met, then determining the transformation parameters of that round of transformation processing as the preset transformation parameters; if the similarity standard is not met, then iteratively updating the transformation parameters of the transformation processing based on the transformation result of that round of transformation processing, and using the updated transformation parameters to perform transformation processing on the registered touch data in the next round of transformation processing.

[0010] In some optional embodiments, the method further includes: if the number of processing steps of the at least one transformation process reaches a preset number of rounds, and no transformation result is obtained that satisfies the similarity standard with the reference touch data, then the transformation parameter corresponding to the transformation result with the highest similarity to the reference touch data is determined as the preset transformation parameter.

[0011] In some optional embodiments, obtaining the first touch data obtained by the user performing a touch operation on the touch screen includes: obtaining capacitance enhancement data, wherein the capacitance enhancement data is used to characterize: the capacitance data obtained by the user performing a touch operation on the touch screen after enhancing the capacitance data acquisition capability of the edge area of ​​the touch screen relative to the capacitance data acquisition capability of the middle area of ​​the touch screen; and determining the capacitance enhancement data as the first touch data.

[0012] In some optional embodiments, the touch screen includes a first edge region, a middle region, and a second edge region arranged along a first direction. The touch screen also includes a plurality of driving signal lines arranged sequentially along the first direction for collecting capacitance data. At least one first driving signal line is located in the first edge region, at least one second driving signal line is located in the second edge region, and at least one third driving signal line is located in the middle region. The capacitance data acquisition capability of the edge region of the touch screen is enhanced relative to the capacitance data acquisition capability of the middle region by setting the coding time of at least one first driving signal line and / or at least one second driving signal line to be greater than the coding time of each of the third driving signal lines.

[0013] In some optional embodiments, the method further includes: displaying a given touch area on the touch screen corresponding to the touch method selected by the user, to prompt the user to perform a touch operation in the given touch area, wherein the given touch area includes a given intention touch area and a given edge accidental touch area; acquiring registered touch data obtained by the user performing a touch operation in the given touch area on the touch screen; wherein the registered touch data is used to determine preset transformation parameters for the user, the preset transformation parameters are used to transform the user's first touch data to obtain second touch data for input into the machine learning model for touch area detection.

[0014] In some alternative embodiments, the given edge accidental touch area and the given intention touch area are configured to have different visual effects.

[0015] According to a second aspect of the embodiments of this application, a model training method is provided, comprising: acquiring training samples for training a machine learning model, wherein the training samples include target touch data samples obtained by a user performing touch operations on a touch screen, and ground truth values ​​of intent touch areas corresponding to the target touch data samples, wherein the target touch data samples at least include touch data samples for intent touch areas of the touch screen; and inputting the training samples into a machine learning model to train the machine learning model.

[0016] In some optional embodiments, the target touch data sample includes touch data samples of the intended touch area and touch data samples of the edge mis-touch area; the training sample further includes: the ground truth value of the edge mis-touch area corresponding to the target touch data sample.

[0017] In some optional embodiments, obtaining training samples for training the machine learning model includes: obtaining multiple training samples for training the machine learning model, wherein: the multiple target touch data samples of the multiple training samples include multiple first touch data samples and multiple second touch data samples; the multiple first touch data samples are obtained by touching the touch screen using different preset standard touch techniques; the multiple second touch data samples are obtained by touching the touch screen using personalized touch techniques corresponding to different device manufacturers; the step of inputting the training samples into the machine learning model to train the machine learning model includes: inputting the multiple training samples into the machine learning model and performing supervised fine-tuning training on the machine learning model to obtain machine learning models for different device manufacturers.

[0018] In some optional embodiments, the target touch data sample obtained by the user's touch operation on the touch screen in the training sample is obtained by: acquiring the original capacitance data sample, which is used to characterize: the capacitance data sample obtained by the user's touch operation on the touch screen after enhancing the capacitance data acquisition capability of the edge area of ​​the touch screen relative to the capacitance data acquisition capability of the middle area of ​​the touch screen; and determining the original capacitance data sample as the target touch data sample.

[0019] In some optional embodiments, the touch screen includes a first edge region, a middle region, and a second edge region arranged along a first direction. The touch screen also includes a plurality of driving signal lines arranged sequentially along the first direction for collecting capacitance data. At least one first driving signal line is located in the first edge region, at least one second driving signal line is located in the second edge region, and at least one third driving signal line is located in the middle region. The capacitance data acquisition capability of the edge region of the touch screen is enhanced relative to the capacitance data acquisition capability of the middle region by setting the coding time of at least one first driving signal line and / or at least one second driving signal line to be greater than the coding time of each of the third driving signal lines.

[0020] According to a third aspect of the embodiments of this application, a touch area detection device is provided, comprising: a first acquisition module, configured to acquire first touch data obtained by a user performing a touch operation on a touch screen, wherein the first touch data includes at least touch data for an intended touch area of ​​the touch screen; and a detection module, configured to perform touch area detection based on the first touch data using a machine learning model to obtain coordinate information of the intended touch area corresponding to the first touch data.

[0021] According to a fourth aspect of the embodiments of this application, a model training apparatus is provided, comprising: a second acquisition module, configured to acquire training samples for training a machine learning model, wherein the training samples include target touch data samples obtained by a user performing touch operations on a touch screen, and ground truth values ​​of intent touch areas corresponding to the target touch data samples, wherein the target touch data samples at least include touch data samples for intent touch areas of the touch screen; and a training module, configured to input the training samples into a machine learning model to train the machine learning model.

[0022] According to a fifth aspect of the present application, an electronic device is provided, comprising: a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other via the communication bus; the memory is used to store a computer program; and the processor is used to execute the method described in the first or second aspect by running the computer program stored in the memory.

[0023] According to a sixth aspect of the embodiments of this application, a computer storage medium is provided that stores a computer program thereon, which, when executed by a processor, implements the method described in the first or second aspect.

[0024] According to a seventh aspect of the embodiments of this application, a computer program product is provided, including a computer program that, when executed by a processor, implements the method as described in the first or second aspect.

[0025] According to the technical solution provided in the embodiments of this application, since the first touch data obtained by the user's touch operation on the touch screen can be obtained, and the touch area can be detected by a machine learning model based on the first touch data to obtain the coordinate information of the intended touch area corresponding to the first touch data, the actual intended touch area of ​​the user's touch operation on the touch screen can be accurately and effectively detected by the machine learning model. This is beneficial to further suppress the negative situation caused by accidental touch through accurate detection results, thereby improving the user's experience of using the touch screen. Attached Figure Description

[0026] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings.

[0027] Figure 1A is a schematic diagram of the edge suppression region defined for touch screens in related technologies.

[0028] Figure 1B is a flowchart of a technical solution for touch area detection and edge accidental touch suppression in related technologies.

[0029] Figure 2 is a flowchart of a touch area detection method according to an embodiment of this application.

[0030] Figure 3A is a schematic diagram of an example user grip method and its corresponding touch data.

[0031] Figure 3B is a schematic diagram of another example of user grip and its corresponding touch data.

[0032] Figure 4 is a schematic diagram of different areas of an example touchscreen and the capacitance detection circuit network.

[0033] Figure 5 is a flowchart of some optional steps in step S104 of the present application embodiment.

[0034] Figure 6 is a flowchart of some optional steps in the method of pre-determining the preset transformation parameters in the embodiments of this application.

[0035] Figure 7 is a flowchart illustrating the workflow of some application scenario examples of embodiments of this application.

[0036] Figure 8 is a flowchart of a model training method according to an embodiment of this application.

[0037] Figure 9 shows an example of a target touch data sample before and after annotation.

[0038] Figure 10 is a schematic diagram of a touch area detection device according to an embodiment of this application.

[0039] Figure 11 is a schematic diagram of a model training device according to an embodiment of this application.

[0040] Figure 12 is a schematic diagram of the structure of an electronic device according to an embodiment of this application. Detailed Implementation

[0041] To enable those skilled in the art to better understand the technical solutions in the embodiments of this application, the technical solutions in the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art should fall within the protection scope of the embodiments of this application.

[0042] With the rapid popularization and continuous development of smartphones and other mobile devices, touchscreens have become an indispensable part of them. Full-screen design is gradually becoming a major trend in mobile devices, providing users with a better visual and user experience. However, this design also brings new challenges. For example, due to the increased screen-to-body ratio, users are more prone to accidental touches when holding the phone, such as, but not limited to, accidentally touching the edge of the screen with their fingers or palms. These accidental touches can lead to a series of negative consequences that reduce the user experience, such as, but not limited to, accidental application operations, frequent typing errors, accidental message sending, and frequent screen wake-ups.

[0043] For example, taking the reduction of accidental touches at the edges as an example, most existing solutions in related technologies employ edge anti-accidental touch algorithms. These algorithms extract features such as the shape, position, area, and operation state of the touch to distinguish between normal touches and accidental touches. For instance, referring to Figure 1A, some manufacturers define an "edge suppression area" for the touchscreen according to their product requirements (it should be understood that the landscape and portrait screens in Figure 1A are only examples). Within this area, it is necessary to distinguish between real and accidental touches as much as possible. For example, when a touch event occurs in the middle area (i.e., the white area) of Figure 1A, it is always considered a real touch; however, when a touch event occurs in the "edge suppression area" (i.e., the gray area) of Figure 1A, an edge anti-accidental touch algorithm is needed to determine whether the touch event is a real or accidental touch based on a series of complex conditions. However, solutions including but not limited to those shown in Figure 1A still have shortcomings in the accuracy of suppressing accidental touches and cannot fully meet user needs. Furthermore, most solutions are not very efficient in debugging, which is not conducive to maintenance and optimization, requiring developers to invest a significant amount of time.

[0044] For example, referring to Figure 1B, a flowchart of a related technology solution for touch area detection and edge mis-touch suppression is shown. As shown in Figure 1B, in this related technology solution, taking a capacitive touch screen as an example, the capacitance data of the current frame is first obtained, then the touch frame corresponding to the touch area in the capacitance data is searched, and the corresponding coordinates are calculated. Then, touch frame state analysis and coordinate point state analysis are performed. Finally, combined with manufacturer requirements (such as information on edge suppression areas) and the aforementioned analysis results, edge mis-touch suppression and reporting of touch area coordinates are performed. As can be seen from the above description of the related technology solution, the edge mis-touch suppression of this solution depends on multiple conditions (such as touch frame state, coordinate point state, and manufacturer requirements). These conditions need to be combined and coordinated to ensure the effectiveness of the solution. Moreover, the involvement of the above multiple conditions makes the judgment logic of the solution itself extremely cumbersome. In addition, this solution involves a large number of preset thresholds, each of which needs to be carefully selected by the R&D personnel to improve the accuracy of the existing solution. However, the complexity of this solution cannot guarantee the accuracy of touch area detection and edge mis-touch suppression, nor is it conducive to later maintenance and further optimization. Furthermore, the requirements of the same manufacturer can change frequently. For example, manufacturers typically propose an initial edge mis-touch suppression requirement, which is then adjusted during development. Moreover, different manufacturers have different requirements for edge mis-touch suppression. This results in the lack of universality of the above technical solution. Adding, deleting, or modifying requirements from the same manufacturer, or different requirements from different manufacturers, will lead to modifications or even reconstruction of the above technical solution, consuming a significant amount of time and energy for R&D personnel, resulting in low development efficiency and difficulty in guaranteeing accuracy. It should be noted that the technical solution shown in Figure 1B above may not be an existing technology.

[0045] It should be understood that the embodiments of this application provide a new touch area detection scheme, which can realize the detection of the user's intended touch area and the determination of coordinate information through a machine learning model. This can at least improve the accuracy of touch area detection. The accurate results of this touch area detection can help to more effectively suppress negative situations caused by accidental touches and improve versatility, thereby improving the user's experience of using the touch screen.

[0046] The specific implementation of the embodiments of this application will be further described below with reference to the accompanying drawings.

[0047] Figure 2 is a flowchart of a touch area detection method according to an embodiment of the present application. According to a first aspect of the present application, a touch area detection method is provided. Referring to Figure 2, the method includes steps S102 and S104, specifically:

[0048] S102: Obtain first touch data obtained by the user performing a touch operation on the touch screen, wherein the first touch data includes at least touch data for the intended touch area of ​​the touch screen.

[0049] It should be noted that the touch area detection method in this application embodiment can be applied to touch area detection of any touch screen. For example, the touch screen can be a capacitive touch screen, a resistive touch screen, etc. For ease of explanation, a capacitive touch screen will be used as an example below, and the touch data collected is capacitance data.

[0050] Optionally, the touch area detection method in this application embodiment can be used in the use stage of the touch screen, that is, in the stage after the touch screen has been manufactured and installed on the terminal device.

[0051] In this embodiment, the type of the first touch data can be determined according to the type of touchscreen. For example, taking a capacitive touchscreen as an example, the first touch data can be capacitance data. Here, capacitance data can be mutual capacitance data or self-capacitance data, depending on whether the touchscreen is based on the mutual capacitance principle or the self-capacitance principle. Taking mutual capacitance data as an example, mutual capacitance data can be collected at various locations on the touchscreen. When the user does not touch the touchscreen, the fluctuation of the mutual capacitance data collected at various locations on the touchscreen is very small. However, when the user presses on a part of the touchscreen to achieve touch control or accidentally touches it, the mutual capacitance data collected through that part of the area changes, causing a significant increase in the difference between the mutual capacitance data and the value when there is no touch or accidental touch. Therefore, by detecting the change in the difference in mutual capacitance data (e.g., the difference from a reference value (e.g., 0)), it can be determined whether there is a touch or accidental touch, thereby facilitating the determination of the user's intended touch area, accurately realizing touch area detection, and suppressing accidental touches.

[0052] In this embodiment, the intended touch area can be the area the user actually wants to touch on the touchscreen, or it can be considered the actual touch area. Generally, if a user accidentally touches any location during a touch operation, the accidentally touched area is not the area the user actually wanted to touch. It should be understood that in this embodiment, by acquiring first touch data that includes at least the intended touch area for the touch area, the user can subsequently use a machine learning model to accurately detect the touch area according to their intention. This enables accurate and effective detection of the actual intended touch area for the user's touch operation on the touchscreen, and further helps to suppress negative situations caused by accidental touches through accurate detection results, thereby improving the user's touchscreen experience.

[0053] In some alternative embodiments, the first touch data includes touch data for the intended touch area of ​​the touchscreen and touch data for the edge touch area.

[0054] Based on this, in this embodiment of the application, first touch data is obtained, which includes touch data of the intended touch area of ​​the touch screen and touch data of the edge mis-touch area. This facilitates the subsequent use of a machine learning model to accurately and effectively detect the actual intended touch area of ​​the user's touch operation on the touch screen. It also helps to further suppress the negative situation caused by edge mis-touch through accurate detection results, thereby suppressing the user's edge mis-touch and improving the user's experience of using the touch screen.

[0055] Optionally, if the touch screen is a capacitive touch screen, then the first touch data obtained in the embodiments of this application can be capacitance data, such as the original capacitance data collected from the user's touch, or the capacitance data after noise reduction processing of the original capacitance data. Using the noise-reduced data is beneficial to reducing detection errors.

[0056] In some optional embodiments, the touch screen is a capacitive touch screen, and the above step S102 includes: acquiring capacitance enhancement data, wherein the capacitance enhancement data is used to characterize: the capacitance data obtained by the user performing a touch operation on the touch screen after enhancing the capacitance data acquisition capability of the edge area of ​​the touch screen relative to the capacitance data acquisition capability of the middle area of ​​the touch screen; and determining the capacitance enhancement data as the first touch data.

[0057] In some cases, users may use different grip techniques on the edge area of ​​the touchscreen, yet the shape of the touch data (capacitive data) at the edge area shows little difference. For example, in the left column of Figure 3A, the user's thumb "continuously grips" the edge area of ​​the touchscreen (as shown in Figure 3A, both contacts are accidental edge touches), and the right column of Figure 3A shows examples of the corresponding capacitive data (touch data). Similarly, in the left column of Figure 3B, the user's thumb "continuously clicks" on the edge area (as shown in Figure 3B, one contact is an intended touch, and the other is an accidental edge touch), and the right column of Figure 3B shows examples of the corresponding capacitive data (touch data). However, comparing the right side of Figure 3A and the right column of Figure 3B, the shape of the capacitive data (such as the magnitude of the capacitance difference, the affected area, and the touch shape) is almost identical when the user's thumb is continuously gripping (Figure 3A) and continuously clicking (Figure 3B). In other words, some touch data cannot reflect the differences between certain grip techniques in the edge area, and therefore cannot distinguish the user's true intention (i.e., it cannot determine whether it is an intentional touch or an accidental edge touch).

[0058] It should be understood that in the above-mentioned optional embodiments of this application, by acquiring capacitance enhancement data and then determining the capacitance enhancement data as the first touch data, since the capacitance enhancement data is the capacitance data obtained by the user's touch operation on the touch screen after enhancing the capacitance data acquisition capability of the edge area of ​​the touch screen relative to the capacitance data acquisition capability of the middle area of ​​the touch screen, the data quality corresponding to the edge area in the first touch data is enhanced, so that the first touch data itself can better reflect the user's operation intention, which is conducive to subsequent better touch area detection through machine learning models to determine the coordinate information of the intended touch area, and facilitates the suppression of negative situations such as edge mis-touch, thereby improving the user's experience of using the touch screen.

[0059] In this embodiment, the capacitance data acquisition capability of the edge region of the touch screen can be enhanced relative to the capacitance data acquisition capability of the middle region of the touch screen in any way. For example, Figure 4 is a schematic diagram of different regions of an example touch screen and the capacitance detection circuit network. In some optional embodiments, as shown in sub-figures a and b of Figure 4, the touch screen 100 includes a first edge region 101, a middle region 103 and a second edge region 102 arranged along a first direction F. The touch screen 100 includes a plurality of driving signal lines arranged sequentially in the first direction F for acquiring capacitance data. At least one first driving signal line 1041 among the plurality of driving signal lines is located in the first edge region 101, at least one second driving signal line 1042 among the plurality of driving signal lines is located in the second edge region 103, and at least one third driving signal line 1043 among the plurality of driving signal lines is located in the middle region 103. Optionally, the capacitive data acquisition capability of the edge region of the touch screen can be enhanced relative to the capacitive data acquisition capability of the middle region of the touch screen by setting the coding time of at least one first driving signal line 1041 and / or at least one second driving signal line 1042 to be greater than the coding time of each third driving signal line 1043, so that the capacitive data acquisition capability of the first edge region 101 and / or the second edge region 102 of the touch screen 100 is enhanced relative to the capacitive data acquisition capability of the middle region 103 of the touch screen 100.

[0060] For example, as shown in sub-figures a and b of Figure 4, the touchscreen 100 in this example includes multiple drivers TX0, TX1, ..., TXend, and multiple receivers RX0, RX1, ..., RXend. Each driver is connected to a drive signal line (also called a drive electrode), and each receiver is connected to a receive signal line (also called a receive electrode). The multiple drive signal lines and multiple receive signal lines form a capacitance detection circuit network (mutual capacitance type). Taking a total of 5 drive signal lines and 5 receive signal lines in the figure as an example, the capacitance data C11 to C55 (only an example) changes in various areas of the touchscreen 100 caused by touch can be detected and collected. It should be understood that the number and proportion of structures in Figure 4 are only for ease of illustration and understanding, and they are not necessarily consistent with reality, nor are they intended to limit the embodiments of this application.

[0061] Optionally, the first direction F can be the width or length direction of the touchscreen 100; Figure 4 uses the width direction as an example. It should be understood that the number of first drive signal lines 1041 located in the first edge region 101 can be set as needed, the number of second drive signal lines 1042 located in the second edge region 102 can be set as needed, and the number of third drive signal lines 1043 located in the middle region 103 can be set as needed. As an example, as shown in sub-figure b of Figure 4, there is one first drive signal line 1041 in the first edge region 101, namely the drive signal line connected to the leftmost driver TX0 (corresponding to the left edge of the touchscreen); there is one second drive signal line 1042 in the second edge region 102, namely the drive signal line connected to the rightmost driver TXend (corresponding to the right edge of the touchscreen); and there are three first drive signal lines 1043 in the middle region 103, namely the drive signal lines connected to the driver preceding TXend from the middle driver TX1.

[0062] The coding time can be considered as the time it takes for the drive signal line to collect capacitance data within one cycle. A longer coding time for the drive signal line allows for a longer time for its driver to send drive signals, thus enabling it to collect more and higher-quality capacitance data. Therefore, by adjusting the coding time, the capacitance data acquisition capability of different areas of the touchscreen can be changed. Therefore, in this embodiment, the coding time of at least one first drive signal line 1041 and / or at least one second drive signal line 1042 can be set to be greater than the coding time of each third drive signal line 1043, thereby enhancing the capacitance data acquisition capability of the first edge region 101 and / or the second edge region 102 relative to the capacitance data acquisition capability of the middle region 103 of the touchscreen 100, and improving the quality of the capacitance data acquired by the first edge region 101 and / or the second edge region 102. Optionally, the coding time of each first drive signal line 1041 and each second drive signal line 1042 can be set to be greater than the coding time of each third drive signal line 1043. For example, the coding time of the third drive signal line 1043 can be set to 100µs, and the coding time of the first drive signal line 1041 and the second drive signal line 1042 can be set to 200µs. For instance, assuming that the original coding time of each drive signal line was 100µs, this is equivalent to increasing the coding time of the first drive signal line 1041 and the second drive signal line 1042 from 100µs to 200µs, while the coding time of the third drive signal line 1043 remains unchanged. Alternatively, other values ​​can be set as needed; no specific limitation is made here.

[0063] Based on this, in the embodiments of this application, the capacitance data acquisition capability of the edge area (i.e., the first edge area and / or the second edge area) of the touch screen is enhanced relative to the capacitance data acquisition capability of the middle area of ​​the touch screen through the above-mentioned optional methods. This can effectively obtain capacitance enhancement data, thereby effectively enhancing the data quality corresponding to the edge area in the first touch data. This allows the first touch data itself to better reflect the user's operation intention, which is beneficial for subsequent touch area detection through machine learning models to determine the coordinate information of the intended touch area and to suppress the negative situation of edge mis-touch, thereby improving the user's experience of using the touch screen.

[0064] For example, let's continue with Figures 3A and 3B. After adopting the aforementioned optional enhancement method, as shown on the left side of Figure 3A, when a user holds the edge of the phone's touchscreen using both the thenar eminence and thumb, the suspended area between the thenar eminence and the thumb pad (highlighted by a double-dotted line in Figure 3A) will also be reflected in the collected touch data (capacitive data) (this differs from the example of touch data on the right side of Figure 3A). This is because, under this grip, the suspended area is closer to the edge of the touchscreen, and with the enhanced capacitive data acquisition capability, it can be collected and reflected. As shown on the left side of Figure 3B, when a user holds the edge of the phone's touchscreen using the thenar eminence and taps the edge with their thumb, the suspended area between the thenar eminence and the thumb pad (highlighted by a double-dotted line in Figure 3B) will not be reflected in the touch data (capacitive data) because, under this grip, the suspended area is farther from the edge of the touchscreen. As can be seen, through the above optional enhancement methods, the original indistinct differences in the capacitance data (touch data) collected by the two different grip methods in Figure 3A and Figure 3B have been transformed into significant differences, which can better reflect the user's operation intention. This allows the first touch data obtained based on data enhancement to be used by the machine learning model for more accurate touch area detection, better suppression of edge touches, and improved user experience when using the touch screen.

[0065] S104: Based on the first touch data, a machine learning model is used to detect the touch area and obtain the coordinate information of the intended touch area corresponding to the first touch data.

[0066] In this embodiment of the application, after obtaining the first touch data, the touch area can be detected by a machine learning model based on the first touch data, thereby obtaining the coordinate information of the intended touch area corresponding to the first touch data. In this way, even if there are accidental touches on the touch screen (including but not limited to edge accidental touches), they can be accurately suppressed by reporting the accurate coordinate information of the user's actual intended touch area.

[0067] Optionally, the coordinate information of the intent touch area can be the coordinates of the center point of the intent touch area (for example, if the intent touch area is an ellipse, it can be the coordinates of the center point of the ellipse), or the coordinates of each point on the outer contour of the intent touch area, or the coordinates of each point within the intent touch area, or any combination thereof. It can be set as needed, and there is no single limitation here.

[0068] Based on this, the touch area detection scheme in this embodiment of the application, through the above steps S102 to S104, can obtain the first touch data obtained by the user's touch operation on the touch screen, and based on the first touch data, perform touch area detection through a machine learning model to obtain the coordinate information of the intended touch area corresponding to the first touch data. Therefore, the machine learning model can accurately and effectively detect the actual intended touch area of ​​the user's touch operation on the touch screen, which is conducive to further suppressing the negative situation caused by accidental touch through accurate detection results, thereby improving the user's experience of using the touch screen.

[0069] Furthermore, the use of machine learning models not only makes it more convenient to use, but also improves development and debugging efficiency, facilitates maintenance and optimization, enhances the versatility of the solution, and helps reduce the extra time and effort that R&D personnel need to invest before and after development is completed.

[0070] Because this solution can effectively suppress accidental touches, it can also effectively avoid false detection and reduce latency when pulling up or swiping down menus during touchscreen use.

[0071] It should be understood that the machine learning model in the embodiments of this application can be a pre-trained machine learning model. The machine learning model is implemented based on artificial intelligence technology. The machine learning model can be of any type. For example, it can be a convolutional neural network model (e.g., including but not limited to AlexNet, Inception, VGG, ResNet, MobileNet, FCN, U-Net, DeepLab, BiSeNet, R-CNN, YOLO, SSD, RetinaNet, etc.), a deep language model (e.g., including but not limited to LSTM, GRU, BiLSTM, Transformer, MarianMT, BERT, RoBERTa, XLNet, GPT, T5, S4, Mamba, etc.), or a hybrid variant model based on the above two (e.g., including but not limited to ViT, DeiT, ConViT, Swin Transformer, TransUNet, Swin Unet, DETR, Deformable DETR, ConvS2S, ConvBERT, etc.). The training process of the machine learning model is not limited in any way in the embodiments of this application. For example, it can be trained using the model training method described in the second aspect below. This part will be explained in detail below, but will not be elaborated on here.

[0072] The implementation of step S104 described above is not limited in this embodiment. For example, in some optional embodiments, the first touch data can be directly input into a machine learning model for touch area detection, thereby obtaining the coordinate information of the intent touch data corresponding to the first touch data.

[0073] In some alternative embodiments, referring to the flowchart shown in FIG5, step S104 includes steps S1042 and S1044, specifically:

[0074] S1042: Based on the reference touch data, the first touch data is transformed to obtain the second touch data. The reference touch data is data that has been determined to be able to perform touch area detection through a machine learning model to obtain the coordinate information of the intended touch area corresponding to the reference touch data.

[0075] For example, touch data of typical hand shapes can be selected as reference touch data and stored in a storage unit (which can be located in the terminal device with the touchscreen installed, and may include, but is not limited to, a database, or any type of memory with storage function). This touch data of typical hand shapes can be obtained by touching the touchscreen using different techniques. When storing the reference touch data in the storage unit, it is necessary to ensure that the contents of the database are not lost after the device is powered off or disconnected from power. This touch data of typical hand shapes can come from the same person or from different people. It is necessary to ensure that this touch data of typical hand shapes can be correctly recognized by a machine learning model, that is, that the machine learning model can be used to detect the touch area to obtain the coordinate information of the corresponding intended touch area.

[0076] Since the machine learning model trained in this embodiment has a certain degree of generalization, it can accurately detect touch areas and suppress edge mis-touches on a large amount of touch data. However, the machine learning model is still insufficient to cover the hand shapes of all users. To adapt to this situation, when using the machine learning model, the first touch data can be transformed based on the reference touch data to obtain the second touch data, so that the second touch data can better approximate the reference touch data. Since the reference touch data can achieve accurate touch area detection through the machine learning model, the second touch data can also achieve accurate touch area detection through the machine learning model, thereby adapting to the needs of touch area detection and edge mis-touch suppression under the user's hand shape, and effectively improving the user's experience of using the touch screen.

[0077] Step S1042 described above can be implemented in any suitable manner. In some optional embodiments, step S1042 can be implemented using an affine transformation algorithm or a second machine learning model. Optionally, step S1042 may include: transforming the first touch data using an affine transformation algorithm or a second machine learning model based on the reference touch data and preset transformation parameters to obtain the second touch data.

[0078] Based on this, in the embodiments of this application, the first touch data can be effectively transformed by an affine transformation algorithm or a second machine learning model according to the reference touch data and preset transformation parameters to obtain the second touch data that can better approximate the reference touch data. This enables the subsequent machine learning model to accurately detect the touch area of ​​the second touch data, thereby adapting to the user's hand shape requirements for touch area detection and edge accidental touch suppression, and effectively improving the user's experience of using the touch screen.

[0079] For example, when using an affine transformation algorithm, preset transformation parameters can be used to perform an affine transformation on the first touch data to obtain the second touch data. The resulting second touch data closely approximates the reference touch data, making it easier to input into a machine learning model for touch area detection.

[0080] For example, an affine transformation can be achieved using the following formula:

[0081] Where θ represents the rotation angle; sx and sy represent the scaling factors in the x and y directions, respectively; and tx and ty represent the translation amounts in the x and y directions, respectively. The x and y directions can be the horizontal and vertical axes of the coordinate system. Taking the first touch data as the independent variable, an affine transformation is performed using the above formula. By performing an affine transformation on the first touch data according to preset transformation parameters (which can include at least one of θ, sx, sy, tx, and ty), f(first touch data) = second touch data ≈ reference touch data.

[0082] The second machine learning model is a different machine learning model from the one described above, and it can be used to transform the first touch data. Alternatively, the second machine learning model can also be part of another model (such as the aforementioned machine learning model). For the second machine learning model, the preset transformation parameters can be at least a portion of the model parameters. For example, the first touch data can be input into the second machine learning model, which then transforms the first touch data according to the preset transformation parameters to obtain the second touch data. The resulting second touch data closely approximates the reference touch data, facilitating subsequent input into the machine learning model for touch area detection.

[0083] The aforementioned preset transformation parameters can be determined in any suitable manner. For example, in one embodiment, specific values ​​can be set according to needs, experience, etc. In some other optional embodiments, referring to the flowchart shown in FIG6, the preset transformation parameters are predetermined through the following steps T102 to T104:

[0084] T102: Obtain registered touch data obtained from the user's touch operation on the touch screen, wherein the registered touch data includes at least the touch data for the intended touch area of ​​the touch screen.

[0085] For example, operation instructions for user registration can be displayed on the touch screen's display interface (e.g., the system can automatically enter when the terminal device with the touch screen installed is powered on, or the user can actively control the terminal device to enter). The user can follow the prompts in the operation instructions and use the same touch techniques as the reference touch data to enter the registration touch data (which can be capacitive data), so that the user's operation intentions for the reference touch data and the registration touch data are consistent.

[0086] Optionally, the registered touch data may include touch data for the intent control area of ​​the touch screen and touch data for the edge accidental touch area of ​​the touch screen, so as to facilitate edge accidental touch suppression.

[0087] Optionally, in step T102, a given touch area can be displayed on the touchscreen to prompt the user to perform touch operations on the given touch area according to the touch method corresponding to the given touch area. The given touch area includes a given intention touch area and a given edge accidental touch area. Registered touch data is obtained by acquiring the user's touch operations on the given touch area on the touchscreen. This effectively obtains the registered touch data. (This part can be understood with reference to the relevant content below.)

[0088] Optionally, the registered touch data, including touch data for the intended touch area of ​​the touchscreen, can be: touch data for a given intended touch area of ​​the touchscreen. Optionally, the registered touch data, including touch data for the accidental touch area at the edge of the touchscreen, can be: touch data for a given accidental touch area at the edge of the touchscreen.

[0089] Since different parts of the hand can cause accidental edge touches, optionally, multiple accidental edge touch areas may exist within a given touch area displayed on the touchscreen in this application. Of course, in some cases, there may only be one accidental edge touch area.

[0090] Optionally, the accidental touch area at a given edge and the intended touch area are configured with different visual effects. This improves the visualization of different areas of the given touch area for the user, making it easier for the user to distinguish between the accidental touch area at the given edge and the intended touch area. This allows the user to accurately perform touch operations within these areas, thus ensuring accurate registration of touch data. (This section can be understood with reference to the related content below.)

[0091] T104: Perform at least one round of transformation processing on the registered touch data, and obtain preset transformation parameters from the processing results of at least one round of transformation processing based on the relationship between at least one round of transformation processing and preset transformation termination conditions.

[0092] Based on this, through the optional methods of steps T102 to T104 above, registered touch data obtained by the user's touch operation on the touch screen can be acquired, and the registered touch data can be transformed at least once. Based on the relationship between the transformation at least once and the preset transformation termination condition, preset transformation parameters with better effect can be effectively obtained from the processing result of the transformation at least once. When using the preset transformation parameters, the first touch data can be transformed effectively, thereby effectively obtaining second touch data that can better approximate the reference touch data. This allows the subsequent machine learning model to accurately detect the touch area of ​​the second touch data and accurately obtain the coordinate information of the intended touch area corresponding to the first touch data. This enables accurate and effective detection of the actual intended touch area of ​​the user's touch operation on the touch screen, thereby adapting to the touch area detection and edge mis-touch suppression requirements under the user's hand shape and effectively improving the user's experience of using the touch screen.

[0093] It should be understood that the transformation termination conditions in the embodiments of this application can be set as needed. For example, in some optional embodiments, the transformation termination condition may be that the transformation rounds have reached a preset number of rounds, the result of the transformation process and the reference touch data meet a similarity standard, etc. Or it may be other termination conditions.

[0094] In some optional embodiments, step T104, "obtaining preset transformation parameters from the processing results of at least one round of transformation processing based on the relationship between at least one round of transformation processing and preset transformation termination conditions", includes: for each round of transformation processing, determining whether the transformation result obtained by performing the transformation processing on the registered touch data in that round meets the similarity standard with the reference touch data.

[0095] If the similarity criteria are met, the transformation parameters for that round of transformation processing are determined as preset transformation parameters.

[0096] If the similarity criteria are not met, the transformation parameters of the transformation process are iteratively updated based on the transformation results of this round of transformation processing, and the updated transformation parameters are used to transform the registered touch data in the next round of transformation processing.

[0097] Optionally, the similarity criterion can be a similarity score greater than a preset similarity threshold. The preset similarity threshold can be set as needed, such as 90%, 95%, 98%, etc. Alternatively, other similarity criteria can be used. The similarity calculation can employ any suitable similarity index. For example, similarity can be calculated using at least one similarity index, including but not limited to Normalized Cross Correlation (NCC), Structural Similarity (SSIM), Mean-Square Error (MSE), and Cosine Similarity.

[0098] For example, taking a similarity standard of greater than a preset similarity threshold as an example, if the similarity between the transformation result of a certain round of transformation processing and the reference touch data is greater than the preset similarity threshold, then the transformation result can be considered to be basically consistent with the reference touch data. At this time, the transformation effect can be considered to meet the requirements, so the transformation parameters of this round of transformation processing can be determined as preset transformation parameters. The preset transformation parameters can be stored in the storage unit so that they can be directly obtained and used to transform the first touch data in subsequent applications.

[0099] For example, taking a similarity standard of greater than a preset similarity threshold as an example, if the similarity between the transformation result of a certain round of transformation processing and the reference touch data is less than the preset similarity threshold, it is considered that the transformation result is not similar enough to the reference touch data and the transformation effect cannot meet the requirements. Therefore, the transformation parameters of the transformation processing can be iteratively updated, and the updated transformation parameters can be used to continue the next round of transformation processing, so that the transformation results in subsequent rounds can gradually meet the similarity standard (such as greater than the preset similarity threshold) with the reference touch data.

[0100] Optionally, during iterative updates, registered touch data entered by the user using the same method can be used, and / or registered touch data entered by the user using different methods can be used, so that the subsequently obtained preset transformation parameters can better adapt to the transformation processing of the first touch data collected by the user performing touch operations on the touch screen using different methods.

[0101] Based on this, the optional method described in the embodiments of this application sets a similarity standard and, in each round of transformation processing, determines whether the transformation result obtained from the transformation processing of the registered touch data in that round meets the similarity standard with the reference touch data. When the similarity meets the similarity standard, the transformation parameters of that round of transformation processing are determined as preset transformation parameters. When the similarity does not meet the similarity standard, the transformation parameters of the transformation processing are iteratively updated based on the transformation result of that round of transformation processing, and the updated transformation parameters are used to process the registered touch data in the next round of transformation processing. This can effectively achieve better results. The preset transformation parameters are used to effectively transform the first touch data, thereby obtaining second touch data that closely approximates the reference touch data. This allows for accurate touch area detection using a machine learning model on the second touch data, accurately obtaining the coordinate information of the intended touch area corresponding to the first touch data. This enables accurate and effective detection of the actual intended touch area of ​​the user's touch operation on the touchscreen, adapting to the user's hand shape requirements for touch area detection and edge mis-touch suppression, and effectively improving the user experience of the touchscreen.

[0102] In some optional embodiments, the touch area detection method in this application further includes: if at least one transformation process reaches a preset number of rounds and no transformation result is obtained that meets the similarity standard with the reference touch data, then the transformation parameter corresponding to the transformation result with the highest similarity to the reference touch data is determined as the preset transformation parameter.

[0103] Based on this, multiple rounds of parameter iteration and transformation processing can be avoided when transformation results that meet the similarity conditions are not obtained, thereby improving processing efficiency. Furthermore, the transformation parameters corresponding to the transformation results with the highest similarity to the reference touch data are selected as preset transformation parameters. This ensures the effectiveness of the preset transformation parameters based on existing parameter data. When using the preset transformation parameters, it is guaranteed that the first touch data can be effectively transformed using the preset transformation parameters, thereby effectively obtaining second touch data that closely approximates the reference touch data. This allows for accurate touch area detection of the second touch data using a machine learning model, accurately obtaining the coordinate information of the intended touch area corresponding to the first touch data. This enables accurate and effective detection of the actual intended touch area of ​​the user's touch operation on the touch screen, adapting to the user's hand shape touch area detection and edge mis-touch suppression requirements, effectively improving the user experience of the touch screen.

[0104] For example, if the similarity criterion is that the similarity between the transformed result and the reference touch data is greater than 95%, and the preset number of rounds is 100, then the transformation parameters are iteratively updated in each round of transformation processing. Assuming the similarity in the 80th round is 96%, which is greater than 95% and meets the similarity criterion, then the transformation parameters of the 80th round are determined as the preset transformation parameters. If, after 100 rounds of transformation processing, the similarity is less than 95%, and the highest similarity among these 100 rounds is 92% in the 99th round, then the transformation parameters of the 99th round are determined as the preset transformation parameters. It should be understood that this is for ease of understanding and is not intended to limit the embodiments of this application.

[0105] It should be understood that, in addition to using the affine transformation algorithm or the second machine learning model to implement the transformation process as described above, other feasible methods can also be used for transformation processing, and no specific limitations are made in the embodiments of this application.

[0106] S1044: Input the second touch data into the machine learning model to detect the touch area, so as to obtain the coordinate information of the intended touch area corresponding to the first touch data.

[0107] Based on this, in the optional embodiments of steps S1042 to S1044 described above, the first touch data is first transformed according to the reference touch data to obtain the second touch data. After obtaining the second touch data, it can be input into a machine learning model for touch area detection. Since the second touch data can closely approximate the reference touch data, and the reference touch data can achieve accurate touch area detection through the machine learning model, the second touch data can also achieve accurate touch area detection through the machine learning model, accurately obtaining the coordinate information of the intended touch area corresponding to the first touch data. This enables accurate and effective detection of the actual intended touch area of ​​the user's touch operation on the touch screen, thereby better adapting to the user's hand shape requirements for touch area detection and edge mis-touch suppression, and effectively improving the user's experience of using the touch screen.

[0108] In some optional embodiments, the touch area detection method of this application further includes: displaying a given touch area on a touch screen to prompt the user to perform a touch operation on the given touch area according to the touch method corresponding to the given touch area, wherein the given touch area includes a given intention touch area and a given edge mis-touch area; acquiring registered touch data obtained by the user performing a touch operation on the given touch area on the touch screen; wherein the registered touch data is used to determine preset transformation parameters for the user, and the preset transformation parameters are used to transform the user's first touch data to obtain second touch data for input into a machine learning model for touch area detection.

[0109] Based on this, this application displays a given touch area on the touchscreen, including a given intention touch area and a given edge mis-touch area, to prompt the user to perform touch operations on the given touch area according to the touch method corresponding to the given touch area. Then, it obtains the registered touch data obtained by the user's touch operation in the given touch area, so as to determine the preset transformation parameters for the user. This facilitates the subsequent transformation processing of the user's first touch data using the preset transformation parameters to obtain second touch data. The second touch data can then be used to input into a machine learning model, which can facilitate accurate touch area detection for the user through the machine learning model. This adapts to the user's hand shape and the needs for touch area detection and edge mis-touch suppression, effectively improving the user's experience of using the touchscreen.

[0110] Optionally, the registered touch data, including touch data for the intended touch area of ​​the touchscreen, can be: touch data for a given intended touch area of ​​the touchscreen. Optionally, the registered touch data, including touch data for the accidental touch area at the edge of the touchscreen, can be: touch data for a given accidental touch area at the edge of the touchscreen.

[0111] Since different parts of the hand can cause accidental edge touches, optionally, multiple accidental edge touch areas may exist within a given touch area displayed on the touchscreen in this application. Of course, in some cases, there may only be one accidental edge touch area.

[0112] For example, a user registration system can be added to a terminal device with a touchscreen (such as a mobile phone), allowing users of the device to register. For instance, operation instructions for user registration can be displayed on the touchscreen's interface (e.g., automatically initiated by the system when the terminal device is powered on, or actively initiated by the user). These instructions display a given touch area on the touchscreen, prompting the user to use the corresponding touch technique to perform touch operations within that area. The user can then follow the instructions to perform touch operations using the corresponding touch technique (e.g., the touch technique could be a "preset standard touch technique" or a "personalized touch technique," which can be understood by referring to the relevant descriptions in the model training method embodiments below, and will not be repeated here). The touch operation completely covers the intended touch area and the given edge mis-touch area within the given touch area to obtain registered touch data. This registered touch data can then be used to determine preset transformation parameters for the user.

[0113] In this application, the touch technique corresponding to a given touch area can be the same as the touch technique of a certain reference touch data, thus ensuring that the reference touch data aligns with the user's operational intent in registering the touch data. The reference touch data is data that has been determined to be capable of touch area detection using a machine learning model to obtain the coordinate information of the intended touch area corresponding to the reference touch data.

[0114] Optionally, the user's first touch data can be transformed to obtain the second touch data in the following way: based on reference touch data and preset transformation parameters, the first touch data is transformed using an affine transformation algorithm or a second machine learning model to obtain the second touch data.

[0115] Optionally, the preset transformation parameters can be determined by performing at least one round of transformation processing on the registered touch data, and obtaining the preset transformation parameters from the processing results of at least one round of transformation processing based on the relationship between the at least one round of transformation processing and the preset transformation termination condition.

[0116] Optionally, obtaining the preset transformation parameters from the processing results of at least one round of transformation processing based on the relationship between at least one round of transformation processing and a preset transformation termination condition may include: for each round of transformation processing, determining whether the transformation result obtained by performing the transformation processing on the registered touch data in that round meets the similarity standard with the reference touch data; if the similarity standard is met, then the transformation parameters of that round of transformation processing are determined as the preset transformation parameters; if the similarity standard is not met, then the transformation parameters of the transformation processing are iteratively updated based on the transformation result of that round of transformation processing, and the updated transformation parameters are used to perform transformation processing on the registered touch data in the next round of transformation processing.

[0117] Optionally, if at least one transformation process reaches a preset number of rounds and no transformation result is obtained that meets the similarity standard with the reference touch data, then the transformation parameter corresponding to the transformation result with the highest similarity to the reference touch data is determined as the preset transformation parameter.

[0118] Optionally, the second touch data can be input into a machine learning model for touch area detection to obtain the coordinate information of the intended touch area corresponding to the first touch data.

[0119] It should be understood that the above-mentioned optional content such as preset transformation parameters, second touch data, and reference touch data has been described in detail above, and will not be repeated here.

[0120] In some optional embodiments, the given edge accidental touch area and the given intention touch area can be configured with different visual effects. This improves the visualization of different areas of the given touch area for the user, making it easier for the user to distinguish between the given edge accidental touch area and the given intention touch area. This allows the user to accurately perform touch operations within their given edge accidental touch area and given intention touch area, thereby accurately obtaining registered touch data.

[0121] Optionally, the different visual effects described above can be implemented in any form, as long as the user can accurately distinguish between the accidental touch area and the intended touch area. For example, the accidental touch area and the intended touch area can be configured with different colors (e.g., the intended touch area can be configured as a blue area and the intended touch area as a red area; or, for example, the intended touch area can be configured as a blue outline area and the intended touch area as a red outline area; blue and red are just examples here, and different colors can be chosen to achieve the desired effect), thereby achieving different visual effects. Alternatively, the accidental touch area and the intended touch area can be configured with different text and / or graphic labels (e.g., using text labels, the intended touch area can be labeled "Accidental Touch Area," while the intended touch area can be labeled "Actual Touch Area"; the text here is just an example, and more complex or simpler styles can also be used), thereby achieving different visual effects. Other implementation methods can also be used, without limitation.

[0122] The workflow of some application scenario examples of this application embodiment will be described below with reference to Figure 7. This workflow can be applied to touch area detection in edge accidental touch suppression scenarios. As shown in Figure 7:

[0123] First, the current frame and the noise-reduced first touch data are obtained. The first touch data can have an edge enhancement effect. For example, capacitance enhancement data can be obtained and determined as the first touch data. Such first touch data can have an edge enhancement effect (refer to the previous embodiment for understanding).

[0124] Next, search the touch frame using the first touch data;

[0125] Next, it is determined whether there is a touch frame in the edge area of ​​the touch screen. If there is a touch frame in the edge area of ​​the touch screen, the first touch data is transformed to obtain the second touch data. For example, the first touch data can be transformed using an affine transformation algorithm or a second machine learning model based on the reference touch data and preset transformation parameters to obtain the second touch data (refer to the previous embodiment for understanding).

[0126] Next, the second touch data is input into the machine learning model for touch area detection to obtain the coordinate information of the intended touch area corresponding to the first touch data;

[0127] Then, the coordinate information of the intended touch area can be reported, and accidental touch coordinates can be suppressed.

[0128] It should be understood that the workflow shown in Figure 7 above is not intended to limit any aspect of the embodiments of this application.

[0129] It is understood that the foregoing description of the touch area detection scheme is merely an exemplary description of the embodiments of this application and is not intended to limit the embodiments of this application in any way.

[0130] In summary, the touch area detection scheme provided in this application can obtain the first touch data obtained by the user's touch operation on the touch screen, and based on the first touch data, perform touch area detection through a machine learning model to obtain the coordinate information of the intended touch area corresponding to the first touch data. Therefore, the machine learning model can accurately and effectively detect the actual intended touch area of ​​the user's touch operation on the touch screen, which is conducive to further suppressing the negative situation caused by accidental touch through accurate detection results, thereby improving the user's experience of using the touch screen.

[0131] According to a second aspect of the embodiments of this application, a model training method is provided. Referring to the flowchart shown in FIG8, the method includes the following steps S202 and S204, specifically:

[0132] S202: Obtain training samples for training the machine learning model, wherein the training samples include target touch data samples obtained by the user performing touch operations on the touch screen, and ground truth values ​​of the intention touch areas corresponding to the target touch data samples. The target touch data samples include at least touch data samples for the intention touch areas of the touch screen.

[0133] Optionally, if the touchscreen is a capacitive touchscreen, then the target touch data samples in the training samples can be capacitive data samples. These capacitive data samples can be mutual capacitance data or self-capacitance data, depending on whether the touchscreen is based on the mutual capacitance principle or the self-capacitance principle. Optionally, the ground truth value of the intended touch area corresponding to the target touch data sample can include labeled data that annotates the intended touch area in the target touch data sample.

[0134] In this embodiment, by acquiring at least target touch data samples of the intended touch area for the touch area and the ground truth values ​​of the intended touch areas corresponding to the target touch data samples, a machine learning model is trained. This enables the trained machine learning model to accurately detect touch areas according to the user's intent, that is, to accurately and effectively detect the actual intended touch area of ​​the user's touch operation on the touch screen. Furthermore, it is beneficial to further suppress negative situations caused by accidental touches through accurate detection results, thereby achieving accidental touch suppression and improving the user's experience of using the touch screen.

[0135] In some optional embodiments, the target touch data sample includes touch data samples of the intended touch area and touch data samples of the edge mis-touch area; the training sample also includes: the ground truth of the edge mis-touch area corresponding to the target touch data sample.

[0136] Optionally, the true value of the edge mis-touch area corresponding to the target touch data sample may include annotation data that marks the edge mis-touch area in the target touch data sample.

[0137] Based on this, in this embodiment of the application, target touch data samples are obtained, including touch data samples of the intended touch area of ​​the touch screen and touch data samples of the edge mis-touch area. The machine learning model is trained using such target touch data samples, the ground truth of the intended touch area corresponding to the target touch data samples, and the ground truth of the edge mis-touch area corresponding to the target touch data samples. This allows the trained machine learning model to accurately and effectively detect the actual intended touch area of ​​the user's touch operation on the touch screen. Furthermore, accurate detection results can help suppress negative situations caused by edge mis-touches, thereby suppressing edge mis-touches and improving the user's touch screen experience.

[0138] After obtaining the touch data sample, the user can manually annotate it, or a preset annotation model can be used to annotate the target touch data sample to obtain the ground truth value of the intended touch area and the ground truth value of the edge mis-touch area. For example, as shown in FIG9, an example of a target touch data sample before and after annotation is shown. The left side of FIG9 is the target touch data sample, and the right side of FIG9 is the target touch data sample after annotation. In this example, the touch data corresponding to the intended touch area is annotated as 200, and the touch data corresponding to the edge mis-touch area is annotated as 100. Of course, this example is not intended to limit any embodiment of this application.

[0139] The training samples obtained in this application embodiment can be multiple training samples from different groups of people, so that the machine learning model can have good versatility after being trained by the training samples.

[0140] In some optional embodiments, the touch screen is a capacitive touch screen. Optionally, the target touch data sample obtained by the user performing touch operations on the touch screen in the above training samples can be obtained in the following way: obtain the original capacitance data sample, which is used to characterize: the capacitance data sample obtained by the user performing touch operations on the touch screen after enhancing the capacitance data acquisition capability of the edge area of ​​the touch screen relative to the capacitance data acquisition capability of the middle area of ​​the touch screen; and determine the original capacitance data sample as the target touch data sample.

[0141] In some cases, users may use different grip techniques on the edge area of ​​the touchscreen, yet the shape of the touch data (capacitive data) at the edge area shows little difference. For example, in the left column of Figure 3A, the user's thumb "continuously grips" the edge area of ​​the touchscreen (as shown in Figure 3A, both contacts are accidental edge touches), and the right column of Figure 3A shows examples of the corresponding capacitive data (touch data). Similarly, in the left column of Figure 3B, the user's thumb "continuously clicks" on the edge area (as shown in Figure 3B, one contact is an intended touch, and the other is an accidental edge touch), and the right column of Figure 3B shows examples of the corresponding capacitive data (touch data). However, comparing the right side of Figure 3A and the right column of Figure 3B, the shape of the capacitive data (such as the magnitude of the capacitance difference, the affected area, and the touch shape) is almost identical when the user's thumb is continuously gripping (Figure 3A) and continuously clicking (Figure 3B). In other words, some touch data cannot reflect the differences between certain grip techniques in the edge area, and therefore cannot distinguish the user's true intention (i.e., it cannot determine whether it is an intentional touch or an accidental edge touch).

[0142] It should be understood that in the above-mentioned optional embodiments of this application, by acquiring the original capacitance data sample and then determining the original capacitance data sample as the target touch data sample, since the original capacitance data sample is the capacitance data sample obtained by the user's touch operation on the touch screen after enhancing the capacitance data acquisition capability of the edge area of ​​the touch screen relative to the capacitance data acquisition capability of the middle area of ​​the touch screen, the data quality corresponding to the edge area in the target touch data sample is enhanced, so that the target touch data sample itself can better reflect the user's operation intention. After training the machine learning model with such target touch data sample, it is beneficial to perform better touch area detection through the trained machine learning model, and it is easier to suppress the negative situation of edge mis-touch, thereby improving the user's experience of using the touch screen.

[0143] In this embodiment, the capacitance data acquisition capability of the edge region of the touch screen can be enhanced relative to the capacitance data acquisition capability of the middle region of the touch screen in any way. In some optional embodiments, as shown in FIG4, the touch screen 100 includes a first edge region 101, a middle region 103, and a second edge region 102 arranged along a first direction F. The touch screen 100 includes a plurality of driving signal lines arranged sequentially along the first direction F for acquiring capacitance data. At least one first driving signal line 1041 among the plurality of driving signal lines is located in the first edge region 101, at least one second driving signal line 1042 among the plurality of driving signal lines is located in the second edge region 103, and at least one third driving signal line 1043 among the plurality of driving signal lines is located in the middle region 103. Optionally, the capacitive data acquisition capability of the edge region of the touch screen can be enhanced relative to the capacitive data acquisition capability of the middle region of the touch screen by setting the coding time of at least one first driving signal line 1041 and / or at least one second driving signal line 1042 to be greater than the coding time of each third driving signal line 1043, so that the capacitive data acquisition capability of the first edge region 101 and / or the second edge region 102 of the touch screen 100 is enhanced relative to the capacitive data acquisition capability of the middle region 103 of the touch screen 100.

[0144] It should be understood that this part of the relevant content has been described in detail in the previous embodiment of the touch area detection method, so it will not be repeated here. For details, please refer to the previous understanding.

[0145] Based on this, in the embodiments of this application, the capacitance data acquisition capability of the edge area (i.e., the first edge area and / or the second edge area) of the touch screen is enhanced relative to the capacitance data acquisition capability of the middle area of ​​the touch screen through the above-mentioned optional methods. This can effectively obtain the original capacitance data sample, thereby effectively enhancing the data quality corresponding to the edge area in the target touch data sample. This allows the target touch data sample itself to better reflect the user's operation intention. After training the machine learning model with such target touch data sample, it is beneficial to perform better touch area detection through the trained machine learning model, and to suppress the negative situation of edge mis-touch, thereby improving the user's experience of using the touch screen.

[0146] S204: Input the training samples into the machine learning model to train the machine learning model.

[0147] After obtaining the training samples, they can be input into the machine learning model to train it until the training termination condition is met, thus completing the training of the machine learning model. The training termination condition can be set as needed. For example, it can be that the training ends after a predetermined number of training epochs (e.g., 1000, 5000, 10000, etc.), or it can be that the loss value of the machine learning model's loss function is less than a certain threshold.

[0148] In this embodiment, the machine learning model is implemented based on artificial intelligence technology. This machine learning model can be of any type. For example, it can be a convolutional neural network model (including but not limited to AlexNet, Inception, VGG, ResNet, MobileNet, FCN, U-Net, DeepLab, BiSeNet, R-CNN, YOLO, SSD, RetinaNet, etc.), a deep language model (including but not limited to LSTM, GRU, BiLSTM, Transformer, MarianMT, BERT, RoBERTa, XLNet, GPT, T5, S4, Mamba, etc.), or a hybrid variant model based on the above two (including but not limited to ViT, DeiT, ConViT, Swing Transformer, TransUNet, Swing Unet, DETR, Deformable DETR, ConvS2S, ConvBERT, etc.).

[0149] Based on this, the model training scheme of steps S202 to S204 in this embodiment of the application can obtain training samples for training the machine learning model. These training samples include target touch data samples obtained by the user's touch operation on the touch screen and the ground truth values ​​of the intention touch area corresponding to the target touch data samples. The target touch data samples at least include touch data samples for the intention touch area of ​​the touch screen. The training samples are then input into the machine learning model for training. Therefore, after the model training is completed, a machine learning model that can accurately detect the user's intention touch area can be obtained. This facilitates the subsequent use of the machine learning model to accurately and effectively detect the actual intention touch area of ​​the user's touch operation on the touch screen. This is beneficial for further suppressing negative situations caused by accidental touches through accurate detection results, thereby improving the user's experience of using the touch screen.

[0150] Furthermore, since the machine learning model trained using the above-mentioned model training method can be used subsequently, this machine learning model is not only more convenient to use, but also more efficient in development and debugging, easier to maintain and optimize, has better versatility, and helps reduce the extra time and effort that R&D personnel need to invest before and after development is completed.

[0151] In some optional embodiments, step S202 above includes: acquiring multiple training samples for training a machine learning model, wherein: the multiple target touch data samples of the multiple training samples include multiple first touch data samples and multiple second touch data samples.

[0152] In this embodiment, multiple first touch data samples are obtained by touching the touch screen using different preset standard touch techniques.

[0153] In this embodiment, the preset standard touch technique can be a touch technique applicable to various device manufacturers, also known as a "common technique". The preset standard touch technique can be pre-set. For example, taking the touch data sample including touch data samples corresponding to the intention touch area and touch data samples corresponding to the edge touch area as an example, some example preset touch techniques may include, but are not limited to: holding the edge of the touch screen with the thenar eminence and middle, ring, and little fingers of one hand at the same time, and drawing a line or clicking in the middle area of ​​the touch screen with the thumb of the holding hand; holding the edge of the touch screen with the thenar eminence and thumb of one hand at the same time, and drawing a line or clicking in the middle area of ​​the touch screen with the fingers of the other hand; pinching the edge of the touch screen with all five fingers of one hand at the same time, and drawing a line or clicking in the middle area of ​​the touch screen with the fingers of the other hand, etc.

[0154] For example, when collecting the first touch data sample, different groups of people (such as different men, women, the elderly, children, etc.) can use various preset standard touch methods to collect data, so as to cover the operating habits of users with various hand shapes for the "common methods" as much as possible, and the collected first touch data sample is labeled to obtain the corresponding true value.

[0155] Optionally, after obtaining multiple training samples including the first touch data sample, the machine learning model can be trained using multiple training samples. The trained machine learning model will have better overall performance in touch area detection and edge accidental touch suppression under the operation of different users' preset standard touch methods.

[0156] In this embodiment, multiple second touch data samples are obtained by touching the touch screen using personalized touch techniques corresponding to different device manufacturers.

[0157] In this embodiment, the personalized touch method can be a touch method specific to a certain device manufacturer, which can also be called a "non-common method". It depends on the unique and personalized needs of different manufacturers. Therefore, this embodiment does not specifically limit the specific situation of the personalized touch method.

[0158] For example, when collecting second touch data samples, different groups of people (such as different men, women, the elderly, children, etc.) can use personalized touch methods from different device manufacturers to collect data, so as to cover the operating habits of users with various hand shapes for "non-common methods" as much as possible, and adapt to the special needs of different device manufacturers for touch area detection and edge accidental touch suppression. The collected second touch data samples are then labeled to obtain the corresponding true values.

[0159] Optionally, step S204 above may include: inputting multiple training samples into a machine learning model, performing supervised fine-tuning training on the machine learning model, to obtain a machine learning model for different device manufacturers.

[0160] It should be understood that by inputting multiple training samples, including the first touch data sample and the second touch data sample, into the machine learning model for supervised fine-tuning training in the above manner, a machine learning model that can accurately detect the user's intended touch area can be obtained after the model fine-tuning training is completed. Furthermore, this machine learning model can effectively adapt to the special needs of touch area detection and edge mis-touch suppression of different device manufacturers. This facilitates the subsequent use of this machine learning model to accurately and effectively detect the actual intended touch area of ​​the user's touch operation on touch screens of different device manufacturers. This helps to further suppress negative situations caused by mis-touches through accurate detection results, thereby improving the user experience when using touch screens of different device manufacturers and further effectively improving the generalization and versatility of the machine learning model.

[0161] It is understood that the foregoing description of the model training scheme is merely an exemplary description of the embodiments of this application and is not intended to limit the embodiments of this application in any way.

[0162] As can be seen from the technical solutions of the embodiments of this application above, this technical solution starts by improving the quality of touch screen edge data to ensure screen edge data quality. It then pre-trains a machine learning model with strong generalization capabilities. This machine learning model can theoretically approximate any function, thus eliminating the need for developers to design complex processes. Finally, it considers the user's special hand shape, further enhancing the versatility of the technical solution. The more comprehensive technical details contribute to improved accuracy. Furthermore, the use of a machine learning model reduces developer input, saving development manpower. In addition, the system solution uses a machine learning model, benefiting from the universal approximation principle of artificial intelligence, possessing a certain degree of versatility. Only additional data needs to be collected according to the manufacturer's requirements, and then the general machine learning model can be retrained or fine-tuned. In other words, this technical solution can ensure the overall performance of the solution without investing too many R&D personnel, and can quickly respond to changes in device manufacturer requirements or different requirements from different device manufacturers. This effectively saves manpower costs, ensures the accuracy of touch area detection and edge mis-touch suppression, and improves efficiency.

[0163] Referring to FIG10, according to a third aspect of the embodiments of this application, a touch area detection device 1000 is provided, comprising:

[0164] The first acquisition module 1002 is used to acquire first touch data obtained by the user performing touch operation on the touch screen, wherein the first touch data includes at least touch data for the intended touch area of ​​the touch screen;

[0165] The detection module 1004 is used to detect the touch area based on the first touch data using a machine learning model, and obtain the coordinate information of the intended touch area corresponding to the first touch data.

[0166] In this embodiment of the application, through the above-described touch area detection scheme, since the first acquisition module 1002 can acquire the first touch data obtained by the user's touch operation on the touch screen, and the detection module 1004 can perform touch area detection based on the first touch data through a machine learning model to obtain the coordinate information of the intended touch area corresponding to the first touch data, the actual intended touch area of ​​the user's touch operation on the touch screen can be accurately and effectively detected through the machine learning model. This is beneficial to further suppress the negative situation caused by accidental touch through accurate detection results, thereby improving the user's experience of using the touch screen.

[0167] In some alternative embodiments, the first touch data includes touch data for the intended touch area of ​​the touchscreen and touch data for the edge touch area.

[0168] In some optional embodiments, the detection module 1004 is specifically used to: transform the first touch data according to the reference touch data to obtain the second touch data, wherein the reference touch data is data that has been determined to be able to perform touch area detection through the machine learning model to obtain the coordinate information of the intended touch area corresponding to the reference touch data; and input the second touch data into the machine learning model to perform touch area detection to obtain the coordinate information of the intended touch area corresponding to the first touch data.

[0169] In some optional embodiments, the detection module 1004 is specifically used to: transform the first touch data using an affine transformation algorithm or a second machine learning model based on reference touch data and preset transformation parameters to obtain second touch data.

[0170] In some optional embodiments, the preset transformation parameters are predetermined by: acquiring registered touch data obtained by the user performing touch operations on the touch screen, wherein the registered touch data includes at least touch data for the intended touch area of ​​the touch screen; performing at least one round of transformation processing on the registered touch data, and obtaining the preset transformation parameters from the processing result of the at least one round of transformation processing according to the relationship between the at least one round of transformation processing and a preset transformation termination condition.

[0171] In some optional embodiments, obtaining preset transformation parameters from the processing results of the at least one round of transformation processing based on the relationship between the at least one round of transformation processing and a preset transformation termination condition includes: for each round of transformation processing, determining whether the transformation result obtained by performing the transformation processing on the registered touch data in that round meets a similarity standard with the reference touch data; if the similarity standard is met, then determining the transformation parameters of that round of transformation processing as the preset transformation parameters; if the similarity standard is not met, then iteratively updating the transformation parameters of the transformation processing based on the transformation result of that round of transformation processing, and using the updated transformation parameters to perform transformation processing on the registered touch data in the next round of transformation processing.

[0172] In some optional embodiments, the method for determining the preset transformation parameter further includes: if the number of processing steps of the at least one transformation process reaches a preset number of rounds, and no transformation result is obtained that meets the similarity standard with the reference touch data, then the transformation parameter corresponding to the transformation result with the highest similarity to the reference touch data is determined as the preset transformation parameter.

[0173] In some optional embodiments, the first acquisition module 1002 is specifically used to: acquire capacitance enhancement data, wherein the capacitance enhancement data is used to characterize: capacitance data obtained by the user performing touch operation on the touch screen after enhancing the capacitance data acquisition capability of the edge area of ​​the touch screen relative to the capacitance data acquisition capability of the middle area of ​​the touch screen; and to determine the capacitance enhancement data as the first touch data.

[0174] In some optional embodiments, the touch screen includes a first edge region, a middle region, and a second edge region arranged along a first direction. The touch screen also includes a plurality of driving signal lines arranged sequentially along the first direction for collecting capacitance data. At least one first driving signal line is located in the first edge region, at least one second driving signal line is located in the second edge region, and at least one third driving signal line is located in the middle region. The capacitance data acquisition capability of the edge region of the touch screen is enhanced relative to the capacitance data acquisition capability of the middle region by setting the coding time of the at least one first driving signal line and / or at least one second driving signal line to be greater than the coding time of each of the third driving signal lines, thereby enhancing the capacitance data acquisition capability of the first edge region and / or the second edge region of the touch screen relative to the capacitance data acquisition capability of the middle region of the touch screen.

[0175] In some optional embodiments, the touch area detection device 1000 is further configured to: display a given touch area on the touch screen to prompt the user to perform a touch operation in the given touch area according to the touch method corresponding to the given touch area, wherein the given touch area includes a given intention touch area and a given edge mis-touch area; acquire registered touch data obtained by the user performing a touch operation in the given touch area on the touch screen; wherein the registered touch data is used to determine preset transformation parameters for the user, and the preset transformation parameters are used to transform the user's first touch data to obtain second touch data for input into the machine learning model for touch area detection.

[0176] In some alternative embodiments, the given edge accidental touch area and the given intention touch area are configured to have different visual effects.

[0177] The touch area detection device 1000 and its optional embodiments in this application have been described in detail in the foregoing touch area detection method embodiments. Therefore, its related content and beneficial effects can be understood by referring to the content of the optional embodiments of the above method, and will not be repeated here.

[0178] Referring to FIG11, according to a fourth aspect of an embodiment of this application, a model training apparatus 1100 is provided, comprising:

[0179] The second acquisition module 1102 is used to acquire training samples for training a machine learning model. The training samples include target touch data samples obtained by the user performing touch operations on the touch screen and ground truth values ​​of the intention touch areas corresponding to the target touch data samples. The target touch data samples include at least touch data samples for the intention touch areas of the touch screen.

[0180] The training module 1104 is used to input the training samples into the machine learning model and train the machine learning model.

[0181] In this embodiment of the application, through the above-described model training scheme, the second acquisition module 1102 can acquire training samples for training the machine learning model. These training samples include target touch data samples obtained by the user's touch operation on the touch screen and the ground truth values ​​of the intended touch areas corresponding to the target touch data samples. The target touch data samples at least include touch data samples for the intended touch areas of the touch screen. Furthermore, the training module 1104 can input the training samples into the machine learning model to train it. Therefore, after the model training is completed, a machine learning model that can accurately detect the user's intended touch areas can be obtained. This facilitates the subsequent use of the machine learning model to accurately and effectively detect the actual intended touch areas of the user's touch operation on the touch screen. This is beneficial for further suppressing negative situations caused by accidental touches through accurate detection results, thereby improving the user's experience of using the touch screen.

[0182] In some optional embodiments, the target touch data sample includes touch data samples of the intended touch area and touch data samples of the edge mis-touch area; the training sample further includes: the ground truth value of the edge mis-touch area corresponding to the target touch data sample.

[0183] In some optional embodiments, the second acquisition module 1102 is specifically used to: acquire multiple training samples for training a machine learning model, wherein: the multiple target touch data samples of the multiple training samples include multiple first touch data samples and multiple second touch data samples; the multiple first touch data samples are obtained by touching the touch screen through different preset standard touch techniques; the multiple second touch data samples are obtained by touching the touch screen through personalized touch techniques corresponding to different device manufacturers; the training module 1104 is specifically used to: input the multiple training samples into the machine learning model, and perform supervised fine-tuning training on the machine learning model to obtain machine learning models for different device manufacturers.

[0184] In some optional embodiments, the second acquisition module 1102 is specifically used to: acquire an original capacitance data sample, the original capacitance data sample being used to characterize: the capacitance data sample obtained by the user performing a touch operation on the touch screen after enhancing the capacitance data acquisition capability of the edge area of ​​the touch screen relative to the capacitance data acquisition capability of the middle area of ​​the touch screen; and determine the original capacitance data sample as the target touch data sample.

[0185] In some optional embodiments, the touch screen includes a first edge region, a middle region, and a second edge region arranged along a first direction. The touch screen also includes a plurality of driving signal lines arranged sequentially along the first direction for collecting capacitance data. At least one first driving signal line is located in the first edge region, at least one second driving signal line is located in the second edge region, and at least one third driving signal line is located in the middle region. The capacitance data acquisition capability of the edge region of the touch screen is enhanced relative to the capacitance data acquisition capability of the middle region by setting the coding time of the at least one first driving signal line and / or at least one second driving signal line to be greater than the coding time of each of the third driving signal lines, thereby enhancing the capacitance data acquisition capability of the first edge region and / or the second edge region of the touch screen relative to the capacitance data acquisition capability of the middle region of the touch screen.

[0186] The model training device 1100 and its optional embodiments in this application have been described in detail in the foregoing model training method embodiments. Therefore, their related content and beneficial effects can be understood by referring to the content of the optional embodiments of the above method, and will not be repeated here.

[0187] According to a fifth aspect of the present application, an electronic device is provided, comprising: a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other via the communication bus; the memory is used to store a computer program; and the processor is used to execute the touch area detection method described in the first aspect or the model training method described in the third aspect by running the computer program stored in the memory.

[0188] Figure 12 shows a structural block diagram of an optional electronic device according to an embodiment of this application. This embodiment does not limit the specific implementation of the electronic device 1200. As an example, referring to Figure 12, the electronic device 1200 provided in this embodiment includes: a processor 1202, a communications interface 1204, a memory 1206, and a communication bus 1208. Wherein:

[0189] The processor 1202, communication interface 1204, and memory 1206 communicate with each other via communication bus 1208.

[0190] Communication interface 1204 is used for communication with other electronic devices or servers.

[0191] The processor 1202 is used to execute the computer program 1210, specifically the relevant steps in any of the aforementioned touch area detection method or model training method embodiments.

[0192] Specifically, computer program 1210 may include program code that includes computer operation instructions.

[0193] The processor 1202 may be a CPU, a GPU (Graphics Processing Unit), an Application Specific Integrated Circuit (ASIC), or one or more integrated circuits configured to implement the embodiments of this application. The one or more processors included in the smart device may be processors of the same type, such as one or more CPUs; or they may be processors of different types, such as one or more CPUs and one or more ASICs.

[0194] Memory 1206 is used to store computer program 1210. Memory 1206 may include high-speed RAM memory, and may also include non-volatile memory, such as at least one disk storage device.

[0195] Specifically, computer program 1210 can be used to cause processor 1202 to execute the touch area detection method or model training method in any of the foregoing embodiments.

[0196] The specific implementation of each step in computer program 1210 can be found in the corresponding steps and units described in any of the aforementioned touch area detection method or model training method embodiments, and will not be repeated here. Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working process of the device and module described above can be referred to the corresponding process description in the aforementioned method embodiments, and will not be repeated here.

[0197] The electronic device 1200 in this application embodiment has been described in detail in the aforementioned touch area detection method or model training method embodiments. Therefore, its related content and beneficial effects can be understood by referring to the above method embodiments, and will not be repeated here.

[0198] According to a sixth aspect of the embodiments of this application, the embodiments of this application also provide a computer storage medium storing a computer program thereon. When the computer program is executed by a processor, it implements the touch area detection method or model training method described in any of the foregoing method embodiments. The computer storage medium includes, but is not limited to, compact disc read-only memory (CD-ROM), random access memory (RAM), floppy disk, hard disk, or magneto-optical disk, etc.

[0199] According to a seventh aspect of the embodiments of this application, the embodiments of this application also provide a computer program product, including a computer program that, when executed by a processor, implements the touch area detection method or model training method as described in any of the embodiments of the plurality of methods described above.

[0200] The touch area detection device 1000 / model training device 1100 / electronic device 1200 / computer storage medium / computer program product embodiments in this application have been described in detail in the foregoing touch area detection method and model training method embodiments. Therefore, their related content and beneficial effects can be understood by referring to the above method embodiments, and will not be repeated here.

[0201] It should be understood that the various figures in the embodiments of this application are for the purpose of illustrating the structure, and the dimensions of each structure may not be drawn according to the actual scale.

[0202] Furthermore, it should be noted that the user-related information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to sample data used for training the model, data used for analysis, stored data, displayed data, etc.) involved in the embodiments of this application are all information and data authorized by the user or fully authorized by all parties. Moreover, the collection, use and processing of related data must comply with the relevant laws, regulations and standards of the relevant countries and regions, and corresponding operation entry points are provided for users to choose to authorize or refuse.

[0203] It should be noted that, depending on the implementation needs, the various components / steps described in the embodiments of this application can be broken down into more components / steps, or two or more components / steps or parts of the operation of components / steps can be combined into new components / steps to achieve the purpose of the embodiments of this application.

[0204] The methods described in the embodiments of this application can be implemented in hardware, firmware, or as software or computer code that can be stored in a recording medium (such as a CD-ROM, RAM, floppy disk, hard disk, or magneto-optical disk), or as computer code downloaded over a network that is originally stored in a remote recording medium or a non-transitory machine-readable medium and will be stored in a local recording medium. Thus, the methods described herein can be stored on a recording medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware (such as an Application Specific Integrated Circuit (ASIC) or a Field Programmable Gate Array (FPGA)). It is understood that the computer, processor, microprocessor controller, or programmable hardware includes storage components (e.g., Random Access Memory (RAM), Read-Only Memory (ROM), Flash Memory, etc.) capable of storing or receiving software or computer code, which, when accessed and executed by the computer, processor, or hardware, implements the methods described herein. Furthermore, when a general-purpose computer accesses code used to implement the methods shown herein, the execution of the code transforms the general-purpose computer into a dedicated computer for executing the methods shown herein.

[0205] Those skilled in the art will recognize that the units and method steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for specific applications, but such implementations should not be considered beyond the scope of the embodiments of this application.

[0206] The above embodiments are only used to illustrate the embodiments of this application, and are not intended to limit the embodiments of this application. Those skilled in the art can make various changes and modifications without departing from the spirit and scope of the embodiments of this application. Therefore, all equivalent technical solutions also fall within the scope of the embodiments of this application, and the patent protection scope of the embodiments of this application should be defined by the claims.

Claims

1. A method for detecting a touch area, comprising: The system acquires first touch data obtained by the user performing touch operations on the touch screen, wherein the first touch data includes at least touch data for the intended touch area of ​​the touch screen; Based on the first touch data, a machine learning model is used to detect the touch area and obtain the coordinate information of the intended touch area corresponding to the first touch data.

2. The method according to claim 1, wherein, The first touch data includes touch data for the intended touch area of ​​the touch screen and touch data for the edge accidental touch area.

3. The method according to claim 1 or 2, wherein, The step of detecting the touch area using a machine learning model based on the first touch data to obtain the coordinate information of the intended touch area corresponding to the first touch data includes: Based on the reference touch data, the first touch data is transformed to obtain the second touch data. The reference touch data is data that has been determined to be able to perform touch area detection through the machine learning model to obtain the coordinate information of the intended touch area corresponding to the reference touch data. The second touch data is input into the machine learning model to detect the touch area, so as to obtain the coordinate information of the intended touch area corresponding to the first touch data.

4. The method according to claim 3, wherein, The step of transforming the first touch data according to the reference touch data to obtain the second touch data includes: Based on the reference touch data and preset transformation parameters, the first touch data is transformed using an affine transformation algorithm or a second machine learning model to obtain the second touch data.

5. The method according to claim 4, wherein, The preset transformation parameters are determined in advance in the following manner: The system acquires registered touch data obtained from a user's touch operation on a touchscreen, wherein the registered touch data includes at least touch data for the intended touch area of ​​the touchscreen; The registered touch data is subjected to at least one round of transformation processing, and based on the relationship between the at least one round of transformation processing and a preset transformation termination condition, preset transformation parameters are obtained from the processing results of the at least one round of transformation processing.

6. The method according to claim 5, wherein, The step of obtaining preset transformation parameters from the processing results of the at least one round of transformation processing based on the relationship between the at least one round of transformation processing and the preset transformation termination condition includes: For each round of transformation processing, it is determined whether the transformation result obtained by performing the transformation processing on the registered touch data in that round meets the similarity standard with the reference touch data; If the similarity criteria are met, then the transformation parameters for this round of transformation processing are determined as the preset transformation parameters; If the similarity criterion is not met, the transformation parameters of the transformation process are iteratively updated based on the transformation result of this round of transformation processing, and the updated transformation parameters are used to transform the registered touch data in the next round of transformation processing.

7. The method according to claim 6, wherein, The method further includes: If the number of processing steps for at least one transformation reaches a preset number of rounds, and no transformation result is obtained that satisfies the similarity standard with the reference touch data, then the transformation parameter corresponding to the transformation result with the highest similarity to the reference touch data is determined as the preset transformation parameter.

8. The method according to claim 1 or 2, wherein, The acquisition of the first touch data obtained by the user's touch operation on the touch screen includes: Acquire capacitance enhancement data, wherein the capacitance enhancement data is used to characterize: capacitance data obtained by the user's touch operation on the touch screen after enhancing the capacitance data acquisition capability of the edge area of ​​the touch screen relative to the capacitance data acquisition capability of the middle area of ​​the touch screen; The capacitance enhancement data is determined as the first touch data.

9. The method according to claim 8, wherein, The touch screen includes a first edge region, a middle region, and a second edge region arranged along a first direction. The touch screen also includes a plurality of driving signal lines arranged sequentially along the first direction for collecting capacitance data. At least one first driving signal line among the plurality of driving signal lines is located in the first edge region, at least one second driving signal line among the plurality of driving signal lines is located in the second edge region, and at least one third driving signal line among the plurality of driving signal lines is located in the middle region. The capacitive data acquisition capability of the edge region of the touchscreen is enhanced relative to the capacitive data acquisition capability of the center region of the touchscreen in the following manner: The coding time of at least one first driving signal line and / or at least one second driving signal line is set to be greater than the coding time of each of the third driving signal lines, so that the capacitive data acquisition capability of the first edge area and / or the second edge area of ​​the touch screen is enhanced relative to the capacitive data acquisition capability of the middle area of ​​the touch screen.

10. The method according to claim 1 or 2, wherein, The method further includes: A given touch area is displayed on the touch screen to prompt the user to perform touch operations in the given touch area according to the touch method corresponding to the given touch area. The given touch area includes a given intention touch area and a given edge accidental touch area. Obtain the registered touch data obtained by the user performing a touch operation on a given touch area on the touch screen; The registered touch data is used to determine preset transformation parameters for the user. The preset transformation parameters are used to transform the user's first touch data to obtain second touch data for input into the machine learning model for touch area detection.

11. The method according to claim 10, wherein, The given edge accidental touch area and the given intention touch area are configured with different visual effects.

12. A model training method, comprising: Obtain training samples for training a machine learning model, wherein the training samples include target touch data samples obtained by the user performing touch operations on the touch screen, and ground truth values ​​of the intention touch areas corresponding to the target touch data samples, wherein the target touch data samples include at least touch data samples for the intention touch areas of the touch screen; The training samples are input into the machine learning model to train the machine learning model.

13. The method according to claim 12, wherein, The target touch data sample includes touch data samples of the intended touch area and touch data samples of the edge accidental touch area; The training samples also include: the ground truth values ​​of the edge accidental touch areas corresponding to the target touch data samples.

14. The method according to claim 12 or 13, wherein, The step of obtaining training samples for training the machine learning model includes: obtaining multiple training samples for training the machine learning model, wherein: The multiple target touch data samples of the multiple training samples include multiple first touch data samples and multiple second touch data samples; The plurality of first touch data samples are obtained by touching the touch screen using different preset standard touch techniques; The plurality of second touch data samples are obtained by touching the touch screen using personalized touch techniques corresponding to different device manufacturers; The step of inputting the training samples into the machine learning model and training the machine learning model includes: inputting the multiple training samples into the machine learning model and performing supervised fine-tuning training on the machine learning model to obtain machine learning models for different device manufacturers.

15. The method according to claim 12 or 13, wherein, The target touch data samples obtained from user touch operations on the touchscreen in the training samples are obtained in the following way: Obtain raw capacitance data samples, which are used to characterize: capacitance data samples obtained by the user's touch operation on the touch screen after enhancing the capacitance data acquisition capability of the edge area of ​​the touch screen relative to the capacitance data acquisition capability of the middle area of ​​the touch screen. The original capacitance data sample is determined as the target touch data sample.

16. The method according to claim 15, wherein, The touch screen includes a first edge region, a middle region, and a second edge region arranged along a first direction. The touch screen also includes a plurality of driving signal lines arranged sequentially along the first direction for collecting capacitance data. At least one first driving signal line among the plurality of driving signal lines is located in the first edge region, at least one second driving signal line among the plurality of driving signal lines is located in the second edge region, and at least one third driving signal line among the plurality of driving signal lines is located in the middle region. The capacitive data acquisition capability of the edge region of the touchscreen is enhanced relative to the capacitive data acquisition capability of the center region of the touchscreen in the following manner: The coding time of at least one first driving signal line and / or at least one second driving signal line is set to be greater than the coding time of each of the third driving signal lines, so that the capacitive data acquisition capability of the first edge area and / or the second edge area of ​​the touch screen is enhanced relative to the capacitive data acquisition capability of the middle area of ​​the touch screen.

17. A touch area detection device, comprising: The first acquisition module is used to acquire first touch data obtained by the user performing touch operation on the touch screen, wherein the first touch data includes at least touch data for the intended touch area of ​​the touch screen; The detection module is used to detect the touch area based on the first touch data using a machine learning model, and obtain the coordinate information of the intended touch area corresponding to the first touch data.

18. A model training device, comprising: The second acquisition module is used to acquire training samples for training the machine learning model. The training samples include target touch data samples obtained by the user performing touch operations on the touch screen and ground truth values ​​of the intention touch areas corresponding to the target touch data samples. The target touch data samples include at least touch data samples for the intention touch areas of the touch screen. The training module is used to input the training samples into the machine learning model and train the machine learning model.

19. An electronic device comprising: The system includes a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other via the communication bus; the memory is used to store a computer program; and the processor is used to execute the method of any one of claims 1-17 by running the computer program stored in the memory.

20. A computer storage medium having a computer program stored thereon, the computer program being executed by a processor to implement the method as described in any one of claims 1-17.

21. A computer program product comprising a computer program that, when executed by a processor, implements the method as described in any one of claims 1-17.

Images