Mistaken touch prevention control method for touch screen of industrial tablet computer

Abstract

The invention relates to the technical field of computers, and discloses a mistaken touch prevention control method for a touch screen of an industrial tablet computer. The method comprises the following steps: monitoring foreground application switching, extracting control and layout features of a current graphical user interface, generating quantitative feature vectors, and inputting the quantitative feature vectors into a preset scene prototype classification model to judge that a scene is a precise interaction scene, a data entry scene, a discrete control scene or a universal default scene currently; and dynamically loading a corresponding mistaken touch prevention strategy parameter set according to a classification result, and carrying out real-time processing such as edge suppression, contact area filtering and static hovering judgment on an original touch event flow. According to the method and the device, the self-adaptive adjustment of the mistaken touch prevention strategy is realized, and the response efficiency and smoothness of high-intensity operation are improved while the high-precision operation accuracy is guaranteed.

Description

Technical Field

[0001] This invention belongs to the field of computer technology, and specifically relates to a method for preventing accidental touches on a touchscreen used in industrial tablet PCs. Background Technology

[0002] With the deep integration of industrial automation and human-machine interaction technologies, industrial panel PCs, as key operating terminals, have been widely used in complex operational scenarios such as intelligent manufacturing, on-site control, logistics scheduling, and equipment maintenance. Their core interactive component—the touchscreen—must maintain stable response under harsh conditions such as high noise, strong vibration, multiple gestures in parallel, and operation while wearing gloves. To ensure operational reliability, existing anti-mistouch mechanisms generally employ filtering algorithms based on fixed thresholds or static contact area determination strategies, such as filtering unintentional touches by setting minimum touch area, duration, or pressure thresholds.

[0003] However, such rigid rules are difficult to adapt to the ever-changing application needs in industrial settings: when drawing precision or fine-tuning parameters, users need extremely high touch sensitivity and positioning accuracy, but existing strategies, due to excessive suppression of small or rapid touches, are prone to losing effective input; conversely, in high-speed interactive scenarios such as batch data entry or emergency equipment operation, overly conservative anti-accidental touch logic will introduce operation delays or action interruptions, significantly reducing work efficiency.

[0004] Touchscreen anti-mistouch control methods for industrial panel PCs focus on dynamically recognizing user operation intentions and adjusting touch response strategies accordingly. The core of this approach lies in how to ensure system robustness while achieving adaptive adaptation to different interaction modes, avoiding a decline in human-machine collaboration efficiency due to a "one-size-fits-all" approach to protection.

[0005] Existing anti-accidental touch solutions typically rely on single sensor data or preset scene labels for judgment, lacking comprehensive perception capabilities of the currently running application type, user operating habits, and environmental interference factors.

[0006] Most systems cannot identify in real time whether a user is in drawing, input, or control mode; they can only execute a unified policy based on global configuration. Even if some solutions introduce application identification mechanisms, they often adopt discrete switching methods, which can lead to policy lag or state oscillation during rapid application switching. Existing methods do not fully integrate multi-dimensional touch features with contextual information, resulting in a high false positive rate.

[0007] The aforementioned defects are particularly prominent in industrial operation scenarios where high frequency switching, high precision, and high efficiency coexist, severely restricting the smoothness and safety of human-computer interaction. There is an urgent need for an intelligent anti-mistouch control method that can achieve scene perception, dynamic adjustment, and seamless transition. Summary of the Invention

[0008] The technical problem to be solved by this invention is that the existing industrial tablet PC touch screen anti-mistouch control method adopts a fixed and single anti-mistouch strategy, which cannot adaptively adjust to the changes in touch interaction requirements when users switch between different application scenarios such as precision drawing, data entry, and equipment operation. This results in excessive sensitivity and easy mis-touch when high-precision operation is required, or excessive restrictions due to overly strict filtering conditions when fast response operation is required, thereby reducing the accuracy of operation and overall work efficiency.

[0009] To address the aforementioned technical problems, this invention provides a method for preventing accidental touches on a touchscreen for an industrial tablet PC. This method monitors the switching of the current foreground application on the industrial tablet PC in real time, automatically acquires and analyzes the layout features of the current application's graphical user interface, classifies the graphical user interface into scene prototypes, dynamically loads and activates a set of anti-accidental touch strategy parameters that match the current scene prototype based on the classification results, and finally uses the activated anti-accidental touch strategy parameter set to filter the original touch event stream generated by the touchscreen to generate legitimate touch commands.

[0010] This invention provides a method for preventing accidental touches on a touchscreen in an industrial tablet PC, comprising the following steps: Monitor the current foreground application status of the industrial tablet PC's operating system, and trigger the interface feature extraction process when a switch in the foreground application is detected. Obtain the graphical user interface feature data of the currently active foreground application. The graphical user interface feature data includes the number, position, and size information of all interactive controls in the interface, as well as the layout information of the non-control areas. Based on the graphical user interface feature data, a set of quantized feature vectors for characterizing interface interaction attributes are calculated and generated. The quantized feature vector is input into a preset scene prototype classification model for calculation and matching, and the scene prototype category to which the current graphical user interface belongs is output. Based on the scenario prototype category, retrieve and load the corresponding anti-accidental touch strategy parameter set from the preset strategy parameter repository; The set of parameters for the anti-mistouch strategy is set as the currently effective filtering strategy and applied to filter the raw touch event stream reported by the touch screen controller in real time. Touch events that do not conform to the current strategy are discarded, and the filtered touch events are forwarded to the operating system.

[0011] As one embodiment of the present invention, obtaining the graphical user interface feature data of the currently active foreground application specifically includes: By calling the operating system's accessibility application programming interface or user interface automation framework interface, the control tree structure of the current foreground application window is traversed to obtain the class name identifier, boundary rectangle size and position in the absolute screen coordinate system for each interactive control. Meanwhile, by capturing the frame buffer image data of the current screen, an image segmentation algorithm is performed on the image data to identify a large, continuous background area or drawing canvas area that does not contain any identified interactive controls.

[0012] As one embodiment of the present invention, the calculation and generation of a set of quantized feature vectors for characterizing interface interaction attributes specifically includes: Calculate the density of interactive controls, which is defined as the ratio of the total number of interactive controls to the total screen area; Calculate the average control size, which is defined as the arithmetic mean of the areas of all interactive controls; The layout dispersion of controls is calculated by measuring the degree of concentration or dispersion of the distribution by calculating the standard deviation of the geometric centers of all interactive controls. Calculate the canvas area ratio, which is defined as the ratio of the area of ​​the largest identified continuous background area or drawing canvas area to the total screen area; The existence identifier of the virtual keyboard is calculated by searching for the existence of a window class name specific to the input method in the control tree structure.

[0013] As one embodiment of the present invention, the step of inputting the quantized feature vector into a preset scene prototype classification model for calculation and matching, and outputting the scene prototype category to which the current graphical user interface belongs specifically includes: The scenario prototype classification model is a rule-based decision tree engine; The decision tree engine makes its decisions based on the following rules: If the proportion of the canvas area is greater than the first preset threshold and the average control size is less than the second preset threshold, then the scene prototype category is determined to be a precision interaction scene. If the virtual keyboard existence flag is true, then the scene prototype category is determined to be a data entry scene; If the average control size is greater than the third preset threshold and the interactive control density is less than the fourth preset threshold, then the scene prototype category is determined to be a discrete control scene. Otherwise, it will be classified as the general default scenario.

[0014] The first preset threshold, the second preset threshold, the third preset threshold, and the fourth preset threshold are calibration parameters permanently stored in non-volatile memory.

[0015] As one embodiment of the present invention, the anti-accidental touch strategy parameter set is a data structure containing multiple parameter fields, specifically including: Minimum touch contact area threshold, used to filter out noisy touch points with too small an area; Maximum touch contact area threshold, used to filter out large-area contacts caused by the palm or arm; Edge suppression distance defines the width of the boundary area around the screen where no touch response occurs; The static touch detection time is determined when the position of the touch point does not change or the change is less than the movement tolerance within this time. It is then identified as static hovering and may be identified as palm pressure or ignored. Multi-touch association rules define the processing logic when multiple concurrent touch points are detected, distinguishing between gesture operations or combinations of main touch points and interfering touch points.

[0016] As one embodiment of the present invention, the step of retrieving and loading the corresponding anti-accidental touch strategy parameter set from the preset strategy parameter repository according to the scenario prototype category specifically includes: For precise interaction scenarios, a high-inhibition strategy parameter set is loaded, with a minimum touch contact area threshold of 10 square millimeters, a maximum touch contact area threshold of 150 square millimeters, an edge suppression distance of 15 millimeters, and a static touch judgment duration of 300 milliseconds, in order to filter out interference from non-precise pen tips or fingertips to the greatest extent. For data entry scenarios, a balanced strategy parameter set is loaded, with a maximum touch contact area threshold of 300 square millimeters and a more tolerant multi-touch association rule, to allow fast keyboard input while suppressing palm pressure outside the input area; For discrete control scenarios, a high-tolerance strategy parameter set is loaded, with a maximum touch contact area threshold of 600 square millimeters and zero edge suppression distance to ensure fast and reliable response to large-sized function buttons.

[0017] As one embodiment of the present invention, the real-time filtering of the raw touch event stream reported by the touch screen controller specifically includes: Each received raw touch data packet contains a touch identifier, coordinates, contact area, pressure value, and timestamp. Perform the following check sequence in sequence: First, check if the touch point coordinates are within the effective area of ​​the screen center defined by the currently effective edge suppression distance; otherwise, discard the data packet directly. Next, check whether the contact area of ​​the touch point is within the range formed by the minimum and maximum touch contact area thresholds that are currently in effect; otherwise, discard the data packet directly. Then, a historical position and timestamp queue is maintained for each touch point identifier. It is determined whether the touch point constitutes a stationary touch. If the touch point is determined to be stationary and the duration exceeds the stationary touch determination time, the touch point identifier is added to a temporary ignore list. All subsequent data packets belonging to the identifier are discarded until the touch point is lifted.

[0018] Only data packets that pass the entire inspection sequence are deemed legitimate touch events and formatted before being forwarded to the operating system's input subsystem.

[0019] The present invention also provides a touchscreen anti-mistouch control system for industrial tablet PCs, comprising: The application status monitoring module is used to monitor the current foreground application status of the industrial tablet PC's operating system and to issue a trigger signal when a switch of the foreground application is detected. The interface feature extraction module, coupled to the application status monitoring module, is used to obtain the graphical user interface feature data of the currently activated foreground application after receiving the trigger signal. The quantization feature generation module is used to calculate and generate a set of quantization feature vectors to characterize the interface interaction attributes based on the graphical user interface feature data. The scene prototype classification module is used to receive the quantized feature vector and output the scene prototype category to which the current graphical user interface belongs; The strategy loading module is used to retrieve and load the corresponding anti-accidental touch strategy parameter set from the strategy parameter repository according to the scenario prototype category; The touch event filtering engine is used to set the anti-accidental touch policy parameter set as the currently effective filtering policy and to perform real-time filtering processing on the raw touch event stream from the touch screen controller.

[0020] Furthermore, the interface feature extraction module includes a user interface structure analysis unit and a user interface visual analysis unit. The user interface structure analysis unit is configured to traverse the control tree and extract control attributes by calling the operating system's user interface automation framework interface. The user interface visual analysis unit is configured to identify non-control background or canvas areas by acquiring the screen frame buffer and performing image processing on it.

[0021] Furthermore, the scene prototype classification module integrates a rule-based decision engine. This engine embeds a set of conditional judgment logic that compares each component of the input quantized feature vector with a set of preset thresholds to determine a unique scene prototype category output.

[0022] Furthermore, the strategy loading module internally maintains a mapping table data structure. The key of the mapping table is an identifier of the scene prototype category, and the value is a complete anti-mistouch strategy parameter set structure. Based on the input scene prototype category identifier, the module searches for and reads the corresponding parameter set data in the mapping table and writes it into a shared memory area accessible to the touch event filtering engine.

[0023] Furthermore, the touch event filtering engine is implemented as a low-level input event interception service. This service runs in the operating system kernel mode or a high-priority user-mode process, and it is directly connected to the event output of the touchscreen hardware driver to ensure that the filtering operation is completed before any touch event is received by the upper-layer application of the operating system. The filtering operation includes edge region checks, contact area validity checks, and static hover state checks on each raw touch data packet; only touch data packets that sequentially pass all checks are allowed to pass.

[0024] Compared with the prior art, the beneficial effects of the present invention are as follows: 1. Achieved dynamic adaptability of the anti-accidental touch strategy. This invention can automatically identify the currently running application scenario, without requiring manual user intervention or pre-adaptation by developers, and can match the optimal anti-accidental touch logic for different interactive tasks, fundamentally solving the contradiction that fixed strategies cannot meet the needs of all scenarios.

[0025] 2. Improved accuracy of high-precision operations. In precision interaction scenarios such as industrial design and drawing review, this invention employs a more stringent suppression strategy to effectively filter unintentional contact caused by the palm, wrist, or accidental finger touch, ensuring that delicate operations are not interfered with and improving the accuracy and quality of work.

[0026] 3. Improved smoothness and efficiency in high-intensity operations. In scenarios such as device control panels and rapid data entry, this invention adopts a more lenient strategy, allowing for larger areas and faster touch input, preventing legitimate operations from being filtered out incorrectly, ensuring instant response and smoothness, and improving overall work efficiency.

[0027] 4. High versatility and scalability. This invention is based on the universality analysis of graphical user interface layouts, is not dependent on specific applications, and is equally effective for new applications that emerge in the future. Furthermore, the classification rules and corresponding strategy parameter sets of the scenario prototypes are configurable and extensible, enabling flexible adaptation to the needs of more complex industrial environments. Attached Figure Description

[0028] Figure 1 This is a schematic diagram of the overall technical solution architecture of the touch screen anti-accidental touch control method for industrial tablet PCs proposed in this invention; Figure 2 This is a schematic diagram of the core principle framework of scene prototype classification and dynamic matching of anti-accidental touch strategy based on graphical user interface features in this invention. Figure 3 This is a flowchart illustrating the logical process of interface feature extraction and quantization feature vector generation triggered by foreground application switching in this invention. Figure 4 This is a logical framework diagram of the rule determination and category output of the scene prototype classification model in this invention; Figure 5 This is a schematic diagram illustrating the structural definition and loading mechanism of the parameter set for multiple types of accidental touch prevention strategies in this invention. Detailed Implementation

[0029] Please refer to Figures 1 to 5 This invention provides a touchscreen anti-mistouch control method for industrial tablet PCs. The method dynamically extracts the structure and visual features of the current graphical user interface by sensing the switching events of the foreground application in real time, generates a quantized feature vector, and automatically identifies the current interaction scenario type based on a preset scenario prototype classification model. Then, it loads the matching anti-mistouch strategy parameter set from the strategy parameter repository, and finally performs multi-level real-time filtering on the original touch event stream to improve the response efficiency of high-intensity operations while ensuring the accuracy of high-precision operations.

[0030] The specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0031] The method begins by monitoring the current foreground application status of the industrial tablet PC's operating system. When a switch in the foreground application is detected, the interface feature extraction process is triggered.

[0032] This monitoring process is completed by an application status monitoring service that resides in the background of the operating system.

[0033] This service continuously receives notification messages about foreground application activation or switching by registering the application lifecycle callback interface provided by the operating system.

[0034] Whenever such a notification is received, it is determined to be a foreground application switching event, and a trigger signal is sent to the subsequent processing module to start the complete interface feature extraction and strategy reconfiguration process.

[0035] This mechanism ensures that the system can respond instantly to any application switching behavior without relying on application-specific adaptation or manual user intervention.

[0036] Upon receiving the trigger signal, the step of acquiring the graphical user interface feature data of the currently active foreground application is executed. This step consists of two parallel sub-processes: One approach is to call the operating system's accessibility application programming interface or user interface automation framework interface to traverse the control tree structure of the current foreground application window and obtain the class name identifier, boundary rectangle size and position in the absolute screen coordinate system for each interactive control. Secondly, by capturing the frame buffer image data of the current screen, an image segmentation algorithm is performed on the image data to identify a large, continuous background area or drawing canvas area that does not contain any identified interactive controls.

[0037] The control tree traversal process starts from the root window node, recursively visits all child nodes, and filters out control types with interactive capabilities, including but not limited to buttons, text boxes, sliders, check boxes, and custom drawing area containers.

[0038] For each valid control, record the coordinates of its top-left and bottom-right corners in the screen coordinate system, and then calculate its width, height, and area.

[0039] Meanwhile, the class name string of the control is used for subsequent semantic analysis, such as determining whether a virtual keyboard window exists.

[0040] The image segmentation process employs a binarization segmentation algorithm based on connected component analysis.

[0041] First, the framebuffer image is converted to grayscale, and then a global threshold is set for binarization to obtain a black and white image.

[0042] Next, morphological closing operations are performed on the binary image to eliminate small holes and noise, and then all continuous white regions are identified through the four-neighbor connected component labeling algorithm.

[0043] After excluding the areas that overlap with the known control positions, the remaining largest connected region is identified as the canvas area or the main background area, and its actual area is obtained by converting the number of pixels to the screen resolution.

[0044] Based on the acquired graphical user interface feature data, the steps are performed to calculate and generate a set of quantized feature vectors to characterize the interface interaction attributes.

[0045] The quantized feature vector consists of scalar values ​​in five dimensions: interactive control density, average control size, control layout dispersion, canvas area occupancy, and virtual keyboard presence identifier.

[0046] Interactive control density is defined as the ratio of the total number of interactive controls to the total screen area, and its unit is the number of controls per square millimeter. It is used to measure the density of interactive elements on the interface.

[0047] The average control size is defined as the arithmetic mean of the areas of all interactive controls, reflecting the physical size of a typical control.

[0048] Control layout dispersion measures the degree of concentration or dispersion of the distribution by calculating the standard deviation of the geometric centers of all interactive controls. In the specific calculation, the mean of the horizontal and vertical coordinates of the center points of all controls are first obtained, then the variance of the horizontal and vertical coordinates is calculated separately, and finally the square root of the sum of the two squares is taken as the dispersion index.

[0049] The canvas area ratio is defined as the ratio of the area of ​​the identified large, continuous background area or drawing canvas area that does not contain any identified interactive controls to the total screen area. It is used to determine whether the interface is mainly for free drawing or content display.

[0050] The virtual keyboard existence flag is determined by searching for the existence of a window class name specific to the input method in the control tree structure. If it exists, the flag is set to true; otherwise, it is set to false.

[0051] The five feature values ​​mentioned above together form a five-dimensional real vector, which serves as the sole input to the subsequent classification model.

[0052] Subsequently, the step of inputting the quantized feature vector into a preset scene prototype classification model for calculation and matching, and outputting the scene prototype category to which the current graphical user interface belongs is performed.

[0053] The scenario prototype classification model is a rule-based decision tree engine, which internally contains a set of condition judgment logic.

[0054] The decision tree engine executes the following decision rules in sequence: First, determine if the virtual keyboard existence flag is true. If it is true, directly output the scene prototype category as data entry scene. If not, continue to determine whether the canvas area ratio is greater than the first preset threshold and whether the average control size is less than the second preset threshold. If both are satisfied, the output scene prototype category is a precision interaction scene. If not, further determine whether the average control size is greater than the third preset threshold and whether the density of interactive controls is less than the fourth preset threshold. If satisfied, output the scene prototype category as discrete control scene. If none of the above conditions are met, the output scene prototype category will be the general default scene.

[0055] The first preset threshold, the second preset threshold, the third preset threshold, and the fourth preset threshold are calibration parameters fixed in non-volatile memory, with typical values ​​of 60%, 50 square millimeters, 200 square millimeters, and 0.02 controls per square millimeter, respectively.

[0056] This rule system ensures the certainty and interpretability of classification results, avoiding the uncontrollable risks brought about by black box models.

[0057] Based on the output scenario prototype category, the steps are performed to retrieve and load the corresponding anti-accidental touch strategy parameter set from the preset strategy parameter repository.

[0058] The anti-mistouch strategy parameter set is a data structure containing multiple parameter fields, specifically including minimum touch contact area threshold, maximum touch contact area threshold, edge suppression distance, static touch judgment duration, and multi-touch association rules.

[0059] For precise interaction scenarios, a high-inhibition strategy parameter set is loaded, which has a smaller minimum touch contact area threshold, a smaller maximum touch contact area threshold, a larger edge suppression distance, and a shorter static touch detection time.

[0060] Typical parameter values ​​are: minimum touch contact area threshold is set to 10 square millimeters, maximum touch contact area threshold is set to 150 square millimeters, edge suppression distance is set to 15 millimeters, and static touch detection time is set to 300 milliseconds.

[0061] This configuration is designed to filter out interference from imprecise pen tips or fingertips to the greatest extent possible, preventing unintentional contact with the edge of the palm or wrist from being misinterpreted as valid input.

[0062] For data entry scenarios, a balanced strategy parameter set is loaded, which has a moderate maximum touch contact area threshold and a more lenient multi-touch association rule.

[0063] Typical parameter values ​​are: the maximum touch contact area threshold is set to 300 square millimeters, the edge suppression distance is set to 5 millimeters, the static touch judgment time is set to 800 milliseconds, and the multi-touch association rule allows two touch points to exist simultaneously in the virtual keyboard area without triggering the accidental touch filtering.

[0064] This configuration balances the larger contact area created by fingers pressed together during fast typing with the need to suppress pressure from the palm outside the input area.

[0065] For discrete control scenarios, a high-tolerance strategy parameter set is loaded, which has a large maximum touch contact area threshold and zero edge suppression distance.

[0066] Typical parameter values ​​are: maximum touch contact area threshold is set to 600 square millimeters, edge suppression distance is set to 0, and static touch detection time is set to 150 milliseconds.

[0067] This configuration ensures reliable response even when the operator is wearing gloves or using large fingertips to operate large function buttons.

[0068] The general default scenario uses compromise parameters, which are suitable for ordinary application interfaces that are not clearly categorized.

[0069] After successfully loading the target policy parameter set, set it as the currently effective filtering policy and apply it to filter the raw touch event stream reported by the touchscreen controller in real time.

[0070] The original touch event stream is continuously reported by the touch screen hardware controller at a fixed sampling frequency. Each data packet contains a touch point identifier, horizontal and vertical coordinates, contact area, pressure value, and timestamp.

[0071] The filtering process is performed by a low-level input event interception service, which runs in the operating system kernel mode and is directly connected to the event output port of the touch screen driver, ensuring that all filtering logic is completed before any touch events are received by the upper-level application of the operating system.

[0072] The filtering operation performs the following check sequence in sequence: First, perform an edge region check. Based on the currently effective edge suppression distance parameters, define a ring-shaped invalid region around the screen.

[0073] Specifically, the left boundary of the effective touch area is the edge suppression distance value, the right boundary is the screen width minus the edge suppression distance value, the top boundary is the edge suppression distance value, and the bottom boundary is the screen height minus the edge suppression distance value.

[0074] If the horizontal coordinate of the current contact is less than the left boundary or greater than the right boundary, or the vertical coordinate is less than the upper boundary or greater than the lower boundary, then the contact is determined to be in an invalid area, the data packet is immediately discarded, and no further processing is performed.

[0075] Next, a contact area validity check is performed. The contact area value of the current touch point is read, and it is determined whether it is strictly greater than the minimum touch contact area threshold and strictly less than the maximum touch contact area threshold.

[0076] If the contact area is less than or equal to the minimum threshold, it is considered sensor noise or slight drift and is discarded; if the contact area is greater than or equal to the maximum threshold, it is considered a palm, arm or large area of ​​obstruction and is also discarded.

[0077] The next inspection stage is only initiated when the contact area is within the open interval (minimum threshold, maximum threshold).

[0078] Then, a stationary hovering state check is performed. The system maintains an independent historical trajectory buffer queue for each contact identifier, which stores the position and timestamp of the most recent 10 sampling points.

[0079] Whenever a new data packet is received, its coordinates are first compared with the oldest coordinates in the queue to calculate the Euclidean distance. If the distance is less than a preset movement tolerance (typically 3 pixels), the contact is considered to be stationary.

[0080] Simultaneously, the time difference between the current timestamp and the oldest timestamp in the queue is calculated. If this time difference is greater than or equal to the currently effective static touch detection duration, the touch point is determined to be a static hover touch point. Once a touch point is determined to be a static hover touch point, its identifier is added to a temporary ignore list.

[0081] After that, all subsequent data packets carrying this identifier are unconditionally discarded until the system detects the lifting event of the contact point (i.e., the contact area returns to 0). At this point, the identifier is removed from the ignore list and its historical trajectory queue is cleared.

[0082] Only data packets that sequentially pass all three checks mentioned above are considered legitimate touch events. These legitimate events are then repackaged into standard input event format and forwarded to the current foreground application via the standard interface of the operating system's input subsystem. This multi-level filtering mechanism ensures that accidental touches are effectively suppressed in different scenarios without sacrificing the integrity and real-time performance of legitimate operations.

[0083] At the system level, the above method is implemented by an integrated touchscreen anti-mistouch control system. This system includes an application status monitoring module, an interface feature extraction module, a quantized feature generation module, a scene prototype classification module, a strategy loading module, and a touch event filtering engine.

[0084] The application status monitoring module is responsible for listening to foreground application switching events in the operating system and sending synchronous trigger signals.

[0085] Upon receiving a trigger signal, the interface feature extraction module activates the user interface structure analysis unit and the user interface visual analysis unit.

[0086] The user interface structure analysis unit traverses the control tree and extracts control properties by calling the operating system's user interface automation framework interface; The user interface visual analysis unit obtains screen images by accessing frame buffer memory and executes image segmentation algorithms to identify canvas areas.

[0087] The quantization feature generation module receives two sets of feature data, performs the aforementioned five quantization calculations, and outputs a standardized feature vector.

[0088] The scene prototype classification module integrates a rule-based decision engine, which compares each component of the input vector with a fixed threshold and outputs a unique scene category identifier.

[0089] The strategy loading module maintains a key-value mapping table, where the key is the scene category identifier and the value is a complete set of parameters for the anti-accidental touch strategy.

[0090] This module looks up the corresponding parameter set in a table based on the input category identifier and writes it into the shared memory area.

[0091] The touch event filtering engine continuously polls the shared memory region to obtain the latest policy parameters and performs the aforementioned multi-level filtering logic on the original touch event stream.

[0092] The entire system ensures stability and low latency under high concurrency through strict timing synchronization and memory isolation mechanisms.

[0093] In summary, this invention achieves adaptive optimization of anti-mistouch control for industrial tablet PC touchscreens by constructing a complete closed loop from front-end application perception, interface feature quantification, intelligent scene classification to dynamic strategy loading and real-time event filtering. This solution does not rely on code modifications for specific applications, possesses universality, real-time performance, and high reliability, and effectively solves the problem of insufficient adaptability of traditional fixed strategies in diverse industrial scenarios.

[0094] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0095] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A touch screen anti-mis-touch control method for an industrial tablet computer, characterized in that, The method comprises the following steps: Monitoring the current foreground application state of the operating system of the industrial tablet computer, and triggering an interface feature extraction process when a foreground application switching is detected; Obtaining the graphical user interface feature data of the currently activated foreground application, which includes the number, position, size information of all interactive controls in the interface, and the layout information of the non-control area; Based on the graphical user interface feature data, a set of quantitative feature vectors representing the interface interaction attributes are calculated and generated; The quantitative feature vectors are input into a preset scene prototype classification model for calculation and matching, and the scene prototype category to which the current graphical user interface belongs is output; According to the scene prototype category, the corresponding anti-mis-touch strategy parameter set is retrieved and loaded from the preset strategy parameter storage library; The anti-mis-touch strategy parameter set is set as the currently effective filtering strategy, and is applied to filter the raw touch events reported by the touch screen controller in real time, discard the touch events that do not meet the current strategy, and forward the filtered touch events to the operating system. 2.The touch screen anti-mis-touch control method for an industrial tablet according to claim 1, wherein, The method for obtaining the graphical user interface feature data of the currently activated foreground application comprises the following steps: By calling the auxiliary function application programming interface or user interface automation framework interface of the operating system, the class name identifier, boundary rectangle size and position in the screen absolute coordinate system of each interactive control are obtained by traversing the control tree structure of the current foreground application window; At the same time, by intercepting the frame buffer image data of the current screen, an image segmentation algorithm is performed on the image data to identify a large area, continuous background area or drawing canvas area that does not contain any identified interactive control. 3.The touch screen anti-mis-touch control method for an industrial tablet according to claim 2, wherein, The method for calculating and generating a set of quantitative feature vectors representing the interface interaction attributes comprises the following steps: The interactive control density is calculated, which is defined as the ratio of the total number of interactive controls to the total screen area; The average control size is calculated, which is defined as the arithmetic mean of the areas of all interactive controls; The control layout dispersion is calculated by calculating the standard deviation of the geometric centers of all interactive controls to measure the concentration or dispersion of their distribution; The canvas area ratio is calculated, which is defined as the ratio of the area of the identified large area, continuous background area or drawing canvas area that does not contain any identified interactive control to the total screen area; The virtual keyboard existence identifier is calculated by checking whether there is a window class name specific to the input method in the control tree structure. 4.The touch screen anti-mis-touch control method for an industrial tablet according to claim 3, wherein, The method for inputting the quantitative feature vectors into a preset scene prototype classification model, performing calculation and matching, and outputting the scene prototype category to which the current graphical user interface belongs comprises the following steps: The scene prototype classification model is a rule-based decision tree engine; The decision tree engine makes a decision according to the following rules: If the canvas area ratio is greater than a first preset threshold, and the average control size is less than a second preset threshold, the scene prototype category is determined to be a precise interaction scene; If the virtual keyboard existence identifier is true, the scene prototype category is determined to be a data entry scene; If the average control size is greater than a third preset threshold and the interactive control density is less than a fourth preset threshold, the scene prototype category is determined as a discrete control scene; Otherwise, it is determined as a general default scene; The first preset threshold, the second preset threshold, the third preset threshold and the fourth preset threshold are calibration parameters fixed in a non-volatile memory. 5.The touch screen anti-mis-touch control method for an industrial tablet according to claim 4, wherein, The anti-mis-touch strategy parameter set is a data structure including multiple parameter fields, specifically including: A minimum touch contact area threshold for filtering out noise touch points with too small areas; A maximum touch contact area threshold for filtering out large-area contacts generated by palms or arms; An edge suppression distance defining a boundary area width around the screen that does not respond to any touch; A stationary touch judgment duration, when the position of a touch point does not change or the change is less than a movement tolerance within the duration, the touch point is determined as a stationary hover, which can be identified as a palm pressure or ignored subsequently; A multi-point touch association rule defining a processing logic when multiple concurrent touch points are detected, distinguishing between gesture operations or a combination of a main touch point and an interference touch point. 6.The touch screen anti-mis-touch control method for an industrial tablet according to claim 5, wherein, The anti-mis-touch strategy parameter set corresponding to the scene prototype category is retrieved and loaded from a preset strategy parameter repository, including: For a precise interaction scene, a high-suppression strategy parameter set is loaded, with a minimum touch contact area threshold of 10 square millimeters, a maximum touch contact area threshold of 150 square millimeters, an edge suppression distance of 15 millimeters, and a stationary touch judgment duration of 300 milliseconds; For a data entry scene, a balanced strategy parameter set is loaded, with a maximum touch contact area threshold of 300 square millimeters and a more tolerant multi-point touch association rule; For a discrete control scene, a high-tolerance strategy parameter set is loaded, with a maximum touch contact area threshold of 600 square millimeters and a zero edge suppression distance.

7. The touch screen anti-mis-touch control method for an industrial tablet according to claim 6, wherein, The original touch event stream reported by the touch screen controller is filtered in real time, including: For each received original touch data packet, it contains a touch point identifier, coordinates, contact area, pressure value and timestamp; The following check sequence is executed in turn: First, check if the touch point coordinates are within the screen center valid area defined by the currently effective edge suppression distance, if not, discard the data packet directly; Second, check if the touch point contact area is within the interval formed by the currently effective minimum and maximum touch contact area thresholds, if not, discard the data packet directly; Then, maintain a history position and timestamp queue for each touch point identifier, determine if the touch point constitutes a stationary touch, if the touch point is determined as a stationary hover and the duration is greater than the stationary touch judgment duration, add the touch point identifier to the temporary ignore list, discard all subsequent data packets belonging to the identifier until the touch point is lifted; Only the data packets that pass through all the check sequences are determined as legal touch events and are formatted and forwarded to the operating system. 8.The touch screen anti-mis-touch control method for an industrial tablet according to claim 7, wherein, The image segmentation algorithm uses a binary segmentation algorithm based on connected domain analysis, including: Convert the frame buffer image to a grayscale image and perform global threshold binaryzation processing; Perform morphological closing operation on the binary image to eliminate small holes and noise; All continuous white regions are identified by a four-neighbor connected component labeling algorithm; After excluding the regions overlapping with the known control positions, the largest remaining connected component is identified as the canvas region or the main background region. 9.The touch screen anti-mis-touch control method for an industrial tablet according to claim 8, wherein, The control layout dispersion is calculated as follows: First, the mean values of the horizontal and vertical coordinates of the geometric centers of all the interactive controls are calculated; Then, the variances of the horizontal and vertical coordinates are calculated respectively; Finally, the square root of the sum of the squares of the two variances is taken as the control layout dispersion index. 10.The touch screen anti-mis-touch control method for an industrial tablet according to claim 9, wherein, The judgment logic of the stationary touch determination duration includes: The Euclidean distance between the current touch point and the oldest sample point in the history trajectory queue is calculated; If the distance is less than a preset movement tolerance, it is considered that the touch point is in a stationary state; The time difference between the current timestamp and the oldest timestamp in the queue is calculated; If the time difference is greater than or equal to the currently effective stationary touch determination duration, it is determined that the touch point is a stationary hover touch point.

Images