Signal processing method and system for touch integrated display module

Abstract

The invention relates to the technical field of signal processing, and discloses a touch integrated display module signal processing method and system. The method comprises the following steps: performing signal acquisition on a display driving circuit of a touch display module during a display frame interval to obtain mixed signal data; performing common-mode noise separation on the mixed signal data according to a common-mode noise reference to obtain touch signal data; identifying a differential mode noise component generated by a pixel driving voltage according to the touch signal data to obtain a differential mode noise separation result and a multi-point touch signal; extracting mixed coordinate data containing real touch points and phantom points according to the multi-point touch signal; based on the differential mode noise separation result, space geometric constraint verification and signal intensity inspection are conducted on the phantom points in the mixed coordinate data, and a touch positioning result is obtained.The phantom point recognition accuracy is ensured, the misjudgment rate is effectively reduced, and the touch reliability of the display module is improved.

Description

Technical Field

[0001] This invention relates to the field of signal processing technology, and in particular to a signal processing method and system for a touch-integrated display module. Background Technology

[0002] Integrated design combines the touch sensing circuit and display driving circuit into a single module, achieving device thinner and lighter designs and cost optimization. However, it also introduces complex signal interference problems. During touch detection, electromagnetic interference generated by the display driving circuit can severely affect the accuracy of the touch signal, especially in high-resolution, high-refresh-rate display scenarios where this interference is more pronounced.

[0003] Existing signal processing technologies for touch display modules primarily rely on single hardware filters or simple software algorithms to handle signal interference. However, these methods cannot effectively distinguish and process common-mode noise and differential-mode noise generated by the display driver circuit. Common-mode noise manifests as in-phase interference across all touch channels, while differential-mode noise is differential interference caused by variations in pixel drive voltage. Their spectral characteristics and interference mechanisms are completely different, requiring different separation strategies. Traditional methods often employ uniform filtering, resulting in the loss of useful touch signals while eliminating noise, thus affecting the sensitivity and accuracy of touch detection. Summary of the Invention

[0004] This invention provides a signal processing method and system for a touch-integrated display module. This invention ensures the accuracy of phantom dot recognition, effectively reduces the false judgment rate, and improves the touch reliability of the display module.

[0005] In a first aspect, the present invention provides a signal processing method for a touch-integrated display module, the signal processing method comprising: During the display frame interval, the display driving circuit of the touch display module is sampled to obtain mixed signal data; The mixed signal data is separated into common-mode noise according to a common-mode noise reference to obtain touch signal data; Based on the touch signal data, the differential mode noise component generated by the pixel driving voltage is identified to obtain the differential mode noise separation result and the multi-touch signal; Based on the multi-touch signal, extract mixed coordinate data containing real touch points and phantom points; Based on the differential noise separation results, the phantom points in the mixed coordinate data are subjected to spatial geometric constraint verification and signal strength test to obtain the touch positioning results.

[0006] In conjunction with the first aspect, in a first implementation of the first aspect of the present invention, the step of acquiring signals from the display driving circuit of the touch display module during the display frame interval to obtain mixed signal data includes: Establish a timing synchronization mechanism between the display refresh cycle and the touch scan cycle, and collect timing control signals for the display frame interval; According to the timing control signal, during the horizontal blanking period, the signal acquisition window is opened to synchronously detect the differential signal generated by the display driving circuit; The differential signal is digitized to obtain an original digital signal containing common-mode noise, differential-mode noise, and touch information; The original digital signal is preprocessed by DC bias elimination and signal amplitude normalization to obtain mixed signal data containing common-mode noise, differential-mode noise and touch information.

[0007] In conjunction with the first aspect, in a second implementation of the first aspect of the present invention, the step of separating the mixed signal data according to a common-mode noise reference to obtain touch signal data includes: Statistical analysis of the DC component of the mixed signal data over a continuous acquisition period is performed to obtain the initial reference value of the common-mode noise in the display driving circuit. An adaptive tracking algorithm is established based on the initial reference value, and the weighted average filter coefficient is dynamically adjusted according to the changes in the displayed content to obtain the common mode noise reference value. The mixed signal data is subtracted point by point from the common-mode noise reference value to eliminate the in-phase interference component, thereby obtaining touch signal data containing differential-mode noise and touch information.

[0008] In conjunction with the first aspect, in a third implementation of the first aspect of the present invention, the step of subtracting the mixed signal data from the common-mode noise reference value point by point to eliminate the in-phase interference component, thereby obtaining touch signal data containing differential-mode noise and touch information, includes: A timing alignment mechanism is established between the mixed signal data and the common-mode noise reference value to ensure that each signal acquisition point matches the common-mode noise reference value at the corresponding time, thereby obtaining a timing-synchronized data pair; Based on the time-synchronized data pairs, the signal amplitude value of each signal acquisition point is subtracted from the corresponding common-mode noise reference value to obtain the calculation result; Based on the calculation results, identify the in-phase interference components generated by the display driving circuit and mark abnormal data points that exceed the normal touch signal range to obtain in-phase interference component identification data; Based on the in-phase interference component identification data, the abnormal data points are filtered to eliminate residual in-phase interference components, resulting in touch signal data containing differential noise and touch information.

[0009] In conjunction with the first aspect, in a fourth implementation of the first aspect of the present invention, the step of identifying the differential-mode noise component generated by the pixel driving voltage based on the touch signal data to obtain the differential-mode noise separation result and the multi-touch signal includes: Fast Fourier Transform analysis and signal power spectral density calculation are performed on the touch signal data to obtain frequency domain analysis data containing the switching frequency characteristics of the pixel TFT. Based on the frequency domain analysis data and pixel driving voltage, the frequency components of differential mode noise are identified to obtain the frequency characteristic parameters of differential mode noise. Based on the frequency characteristic parameters, the notch frequency and notch depth parameters of the adaptive notch filter are set, and the touch signal data is filtered and separated by the adaptive notch filter to obtain the differential noise separation result and the multi-touch signal.

[0010] In conjunction with the first aspect, in a fifth implementation of the first aspect of the present invention, the step of setting the notch frequency and notch depth parameters of the adaptive notch filter based on the frequency characteristic parameters, and filtering and separating the touch signal data through the adaptive notch filter to obtain differential mode noise separation results and multi-touch signals, includes: The center frequency position of the adaptive notch filter is calculated based on the frequency components in the frequency characteristic parameters, and the notch frequency of the adaptive notch filter is set based on the center frequency position. The notch depth parameter of the adaptive notch filter is set according to the differential mode noise signal strength in the frequency characteristic parameters. The touch signal data is input into the adaptive notch filter for filtering to obtain filtered output data. The differential noise component and the touch signal component are then separated and identified in the filtered output data to obtain the differential noise separation result and the multi-touch signal.

[0011] In conjunction with the first aspect, in a sixth implementation of the first aspect of the present invention, the step of extracting mixed coordinate data including real touch points and phantom points based on the multi-touch signal includes: The multi-touch signal is divided into multiple spatial partitions according to the touch detection area, and the touch signal data of each spatial partition is obtained; The touch signal data is subjected to touch signal threshold detection to determine the potential touch point data of each spatial partition; Based on the potential touch point data, analyze the signal correlation between adjacent partitions and calculate the correlation between each potential touch point and the signals of adjacent partitions to obtain the touch point correlation relationship; Based on the aforementioned touch point association relationship, a centroid algorithm is used to calculate the centroid position of the touch signal in spatial distribution, resulting in mixed coordinate data containing real touch points and phantom points.

[0012] In conjunction with the first aspect, in the seventh implementation of the first aspect of the present invention, the step of performing spatial geometric constraint verification and signal strength test on the phantom points in the mixed coordinate data based on the differential noise separation result to obtain the touch positioning result includes: Based on the differential noise separation results, the signal purity of each coordinate point is analyzed and a phantom point signal strength discrimination threshold is set to obtain the signal strength discrimination benchmark. The spatial geometric relationship calculation is performed on the distance between touch points in the mixed coordinate data to identify abnormal coordinate combinations where the distance between adjacent touch points is less than a preset value; The signal strength of each coordinate point in the mixed coordinate data is compared with the signal strength discrimination benchmark to filter out coordinates with abnormal signal strength. Based on the intersection operation of the abnormal coordinate combination and the abnormal signal strength coordinate, the coordinate point that simultaneously violates the spatial constraints and signal strength requirements is determined as the phantom point, and the phantom point identification result is obtained. Remove the phantom point coordinates determined by the phantom point recognition result from the mixed coordinate data to obtain the touch positioning result.

[0013] In conjunction with the first aspect, in the eighth implementation of the first aspect of the present invention, the step of determining the coordinate point that simultaneously violates both spatial constraints and signal strength requirements as a phantom point based on the intersection operation of the abnormal coordinate combination and the abnormal signal strength coordinates, and obtaining the phantom point identification result, includes: Each coordinate point in the abnormal coordinate combination is matched one by one with each point in the abnormal signal strength coordinate to obtain coordinate matching relationship data; Based on the coordinate matching relationship data, overlapping coordinate points that exist simultaneously in abnormal coordinate combinations and abnormal signal strength coordinates are selected to obtain a set of dual violation coordinates. For each coordinate point in the set of dual violation coordinates, calculate the degree of violation of spatial constraints and the degree of signal strength deviation, and establish phantom point confidence data; Based on the phantom point confidence data, coordinate points that simultaneously violate spatial constraints and signal strength requirements are identified as phantom points, and phantom point recognition results are generated.

[0014] Secondly, the present invention provides a touch-integrated display module signal processing system, the touch-integrated display module signal processing system comprising: The signal acquisition module is used to acquire signals from the display driving circuit of the touch display module during the display frame interval to obtain mixed signal data; A common-mode noise separation module is used to separate the mixed signal data according to a common-mode noise reference to obtain touch signal data; The identification module is used to identify the differential noise components generated by the pixel driving voltage based on the touch signal data, and obtain the differential noise separation result and the multi-point touch signal; The extraction module is used to extract mixed coordinate data containing real touch points and phantom points based on the multi-touch signal; The verification module is used to perform spatial geometric constraint verification and signal strength verification on the phantom points in the mixed coordinate data based on the differential noise separation results, so as to obtain the touch positioning results.

[0015] The technical solution provided by this invention overcomes the limitation of the two processes being independent in existing technologies by establishing a collaborative processing mechanism for differential-mode noise separation and phantom point elimination, achieving mutual promotion and synergistic optimization. The use of time-synchronous acquisition technology during display frame intervals ensures precise matching between touch signal acquisition and display drive signals, providing a high-quality data foundation for noise separation. The established adaptive common-mode noise benchmark tracking algorithm and differential-mode noise frequency domain feature recognition algorithm can dynamically adjust processing parameters according to changes in display content, maintaining long-term stable processing performance. Precise coordinate positioning is achieved through spatial partitioning and a centroid algorithm, avoiding capacitive coupling interference between adjacent touch points. The established multi-verification mechanism, including spatial geometric constraint verification, signal strength testing, and intersection operation confirmation, effectively reduces the misjudgment rate of phantom point recognition. Attached Figure Description

[0016] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1 A schematic flowchart illustrating the signal processing method for a touch-integrated display module provided in an embodiment of this application; Figure 2 A schematic block diagram of the signal processing system for the touch-integrated display module provided in the embodiments of this application. Detailed Implementation

[0018] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0019] The flowchart shown in the attached diagram is for illustrative purposes only and does not necessarily include all content and operations / steps, nor does it necessarily have to be performed in the described order. For example, some operations / steps can be broken down, combined, or partially merged, so the actual execution order may change based on the actual situation.

[0020] It should also be understood that the terminology used in this specification is for the purpose of describing particular embodiments only and is not intended to limit the scope of the application. As used in this specification and the appended claims, the singular forms “a,” “an,” and “the” are intended to include the plural forms unless the context clearly indicates otherwise.

[0021] It should also be further understood that the term “and / or” as used in this application specification and the appended claims means any combination of one or more of the associated listed items and all possible combinations, and includes such combinations.

[0022] The following detailed description of some embodiments of this application is provided in conjunction with the accompanying drawings. Unless otherwise specified, the following embodiments and features described herein can be combined with each other.

[0023] Please see Figure 1 , Figure 1 This is a flowchart illustrating the signal processing method for a touch-integrated display module provided in an embodiment of this application, as shown below. Figure 1 As shown, the touch-integrated display module signal processing method provided in this application embodiment includes steps S100 to S600.

[0024] Step S100: During the display frame interval, the display driving circuit of the touch display module is sampled to obtain mixed signal data; It is understood that the executing entity of this invention can be a touch-integrated display module signal processing system, or it can be a terminal or a server; no specific limitation is made here. This embodiment of the invention will be described using a server as an example.

[0025] Specifically, a timing synchronization mechanism is constructed to ensure a corresponding control relationship between the display refresh cycle and the touch scan cycle, thereby enabling stable extraction of the differential drive signal without interfering with normal image output. This synchronization mechanism analyzes the timing control signal output by the display driver circuit, paying particular attention to the horizontal blanking interval between each frame of the displayed image. During this period, no actual content is refreshed in the pixels, thus providing a sampling window with minimal interference for noise analysis and touch data capture. Based on this, the control signal acquisition unit automatically opens the acquisition channel during each horizontal blanking period and synchronously detects changes in the differential signal in the display driver circuit in real time, thereby ensuring that the signal acquisition window is aligned with the display drive cycle and avoiding frequency aliasing or time misalignment between the touch signal and the display signal. The acquired analog differential signal is digitized by a high-precision analog-to-digital conversion module (such as a 16-bit ADC) to generate raw digital signal data. This data contains common-mode interference signals generated by the display driver, differential-mode noise components generated by TFT pixel switching, and real capacitance changes caused by user touch actions. These three factors coexist in the sampled signal in a time-overlapping and spectral-overlapping manner. The raw digital signal data undergoes uniform preprocessing to eliminate DC bias, thus removing the influence of static levels in the display circuit and preventing them from misleading common-mode reference calculations. The amplitude range of each sampling point is normalized, remapping all signal values ​​to a unified numerical range. After these steps, the final output signal is a highly consistent mixed signal containing common-mode noise, differential-mode noise, and touch information.

[0026] Step S200: Perform common-mode noise separation on the mixed signal data according to the common-mode noise reference to obtain touch signal data; Specifically, statistical analysis of the DC component of the mixed signal data is performed over multiple consecutive acquisition cycles. By dynamically calculating the mean of each sampling point within a sliding window, a stable and representative initial reference value for common-mode interference is extracted. This reference value primarily reflects the inherent in-phase electromagnetic interference characteristics of the display driver circuit when there is no effective touch input. Based on this, an adaptive tracking algorithm model is constructed. Starting from the initial reference value, this model continuously absorbs background interference characteristics from each new round of signal sampling by setting a first-order low-pass filter structure. The weighted average coefficient α of the filter is adjusted in real time according to the static or dynamic characteristics of the displayed content. When the displayed content changes little or is static, a lower filter coefficient (e.g., 0.1) is selected to enhance the stability of the reference value. When the displayed content changes frequently or the refresh rate is high, the filter coefficient is appropriately increased (e.g., 0.3) to improve response speed and tracking accuracy. This ensures that the common-mode noise reference value maintains dynamic adaptability and suppression effectiveness under different display environments. By performing a point-by-point subtraction operation between the mixed signal data and the common-mode noise reference value, interference components with in-phase signal characteristics are effectively eliminated, leaving mainly differential-mode noise and the true touch signal information caused by capacitance changes in the remaining signal components. Since common-mode noise is superimposed on multiple touch nodes as global interference, its signal characteristics exhibit high consistency across channels. Therefore, benchmark comparison and subtraction can suppress this type of interference to the maximum extent without destroying the non-homogeneity and local variation characteristics of the touch signal. To prevent long-term drift of the reference value from affecting the suppression effect, a dynamic threshold is set. When the reference value changes by more than a set limit (e.g., ±10%) within several consecutive sampling periods, a re-initialization process is automatically triggered to ensure that the filtering model maintains both accuracy and adaptability. The final output signal data is the intermediate signal that retains differential-mode noise and touch characteristic components after common-mode interference components are eliminated.

[0027] Step S300: Identify the differential noise component generated by the pixel driving voltage based on the touch signal data, and obtain the differential noise separation result and multi-point touch signal; Specifically, the touch signal data is used as the input signal, and frequency domain features are extracted and identified for the differential-mode noise components still contained within it. By performing a Fast Fourier Transform (FFT) on the touch signal data, the time-domain signal is transformed into a frequency-domain representation, and the power spectral density distribution of the signal is calculated based on this, characterizing the contribution ratio of each frequency component to the total signal energy. In the frequency domain analysis results, a set of obvious periodic interference peaks can be observed. These peak frequencies are related to the voltage drive cycle caused by the switching action of the pixel TFTs and have a specific concentrated distribution trend, thus forming a frequency feature region for noise identification. Combining the known switching frequency and operating timing of the TFT pixel scanning in the display driving voltage, the frequency components of the differential-mode noise are identified by comparing the frequency domain analysis data with a preset frequency template, thereby extracting the most representative differential-mode noise frequency feature parameters under the current environment. This feature parameter set includes the location of the dominant frequency point, as well as its bandwidth range, energy intensity, and harmonic distribution characteristics. Based on the aforementioned frequency characteristic parameters, the core parameters of the adaptive notch filter are set, including the notch center frequency and notch depth. The notch frequency must be perfectly aligned with the identified dominant noise frequency, while the notch depth is determined based on the noise signal power and set to approximately -40dB to ensure that interference components are sufficiently suppressed without affecting the effective frequency band of the signal. The adaptive notch filter filters and separates the touch signal data, specifically weakening and eliminating frequency components containing differential-mode interference, ultimately outputting the differential-mode noise separation result and the multi-touch signal.

[0028] Step S400: Extract mixed coordinate data containing real touch points and phantom points based on the multi-touch signal; Specifically, spatial partitioning is performed on multi-touch signals to construct a localized processing framework, thereby improving the sensitivity and accuracy of touch point recognition. The entire touch detection area is divided according to a preset grid structure, for example, into 64 spatial sub-regions of 8×8. Each region is defined as an independent touch processing unit, and touch signal data within this local region is collected and its signal strength and temporal characteristics are recorded, thus obtaining a set of local signals with spatial positioning attributes. Based on this, a dynamically updated touch signal amplitude threshold is set for the signal data within each spatial partition. This threshold is set to be more than three times the current noise baseline level to enhance the sensitivity to real touch events. When the touch signal amplitude in a certain region exceeds this threshold, it is initially identified as a potential touch point. Because touch screens may exhibit capacitive coupling or parasitic induction effects under high-density touch or boundary interference conditions, resulting in false or derived signals in adjacent regions, recognition based solely on thresholds is insufficient to confirm their authenticity. To this end, the signal relationships between all potential touch points and their eight adjacent zones were analyzed. A signal correlation matrix between touch points was constructed by calculating factors such as temporal consistency, amplitude synchronization, and change trends. This correlation value reflects the spatial continuity and logical consistency of each touch point with its surrounding signals and is a key indicator for determining whether it has a real physical trigger source. Based on the touch point correlation, the centroid algorithm was used to calculate the spatial distribution of touch signals in the area where each potential point is located. The coordinates and signal strength of each signal point were used as weights to form a weighted summation model, thereby estimating the centroid position of the overall touch signal in the area. Finally, a mixed coordinate dataset including all potential real touch points and accompanying phantom points was obtained.

[0029] Step S500: Based on the differential noise separation results, perform spatial geometric constraint verification and signal strength test on the phantom points in the mixed coordinate data to obtain the touch positioning results.

[0030] Specifically, based on the differential-mode noise separation results, the purity of the signal corresponding to each touch coordinate point in the time and frequency domains is analyzed, and the signal energy ratio or signal-to-noise ratio of each coordinate point is calculated to assess its physical reliability. This analysis process establishes a signal strength discrimination benchmark adapted to the current electromagnetic interference environment and touch density. This threshold is set below 70% of the current average touch signal energy to identify false signals with abnormally low amplitudes, parasitic responses, or interference coupling. A systematic analysis of the geometric spatial relationships between all touch points in the mixed coordinate data is performed, calculating the Euclidean distance between any two touch points and identifying point pairs with distances less than a set lower limit (e.g., 15mm). Under normal usage conditions, the minimum distance between a user's finger or pen tip in physical space has certain physiological or structural limitations; therefore, excessively close point pairs suggest the presence of phantom points caused by capacitive crosstalk, mirror signals, or hardware parasitic paths. Simultaneously, the signal strength of each coordinate point is compared item by item with the previously defined strength discrimination benchmark. Points with signal amplitudes significantly lower than the standard threshold are filtered out and recorded as signal strength anomaly coordinates. Subsequently, through set intersection operations, these spatially violating geometric spacing constraints anomaly points are mapped and cross-validated with the set of points with substandard signal strength. This extracts coordinate points that simultaneously meet two independent anomaly conditions; these points are then identified as high-confidence phantom points. To avoid mistakenly deleting real touch information, the spatial distribution of phantom points is assessed for continuity. When a phantom point is relatively independent and does not form a stable continuous touch chain with the surrounding area, its phantom attribute is further enhanced, resulting in higher confirmation. The identified phantom points are systematically removed from the mixed coordinate data, retaining the remaining coordinate points that meet the spatial logic and signal strength requirements as the final touch positioning result output.

[0031] In one specific embodiment, the process of performing step S100 may specifically include the following steps: Establish a timing synchronization mechanism between the display refresh cycle and the touch scan cycle, and collect timing control signals for the display frame interval; During the horizontal blanking period, the signal acquisition window is opened according to the timing control signal to synchronously detect the differential signal generated by the display driver circuit; The differential signal is digitized to obtain the original digital signal containing common-mode noise, differential-mode noise, and touch information; The original digital signal is preprocessed by DC bias elimination and signal amplitude normalization to obtain mixed signal data containing common-mode noise, differential-mode noise and touch information.

[0032] Specifically, by analyzing the synchronization signals output by the display controller, particularly the display frame signal, line synchronization signal, and field synchronization signal, a synchronization reference system for timing locking is constructed. In this system, the display refresh cycle serves as the global master clock signal, while the touch scan cycle is scheduled and configured as a passive slave cycle, ensuring that each touch sampling cycle corresponds to the corresponding display frame cycle. In the specific composition of the display refresh cycle, each frame display signal includes a visible effective image cycle and an invisible frame interval. The horizontal blanking interval is the period during pixel switching and signal pause in the display driver; therefore, electromagnetic interference in this area is relatively low, making it an ideal signal acquisition window. To achieve this, the system automatically triggers the opening of the data acquisition window within the horizontal blanking interval by real-time capture of the timing control signal, aligning the signal sampling start time with the stable edge of the blanking period to minimize dynamic interference during pixel switching, thereby improving the accuracy and consistency of differential signal detection. After the acquisition window is activated, the differential signal is transmitted to the analog-to-digital conversion unit via the front-end signal conditioning module, where a 16-bit high-resolution ADC chip performs high-precision digital conversion of the analog signal. In this process, the system continuously samples the differential channel signals at fixed time intervals to obtain a time-series signal sequence. Each sample point contains amplitude information from multiple channels. These differential signals physically superimpose three types of components: common-mode noise caused by the display circuit drive, manifested as in-phase signal disturbances generated synchronously by all channels; differential-mode noise caused by the periodic activation of TFT pixel switches, possessing certain frequency and power characteristics; and effective signal components caused by the minute disturbances to the capacitor array due to the user's actual touch behavior. This part of the signal exhibits obvious characteristics such as non-periodicity, concentrated regional distribution, and local amplitude abrupt changes. A unified standardized preprocessing procedure is performed on the original digital signals. DC bias elimination is performed on each group of signals by calculating the average signal value over a period of time and subtracting it from the original signal, eliminating global offsets introduced by circuit static bias or inconsistent reference levels, thereby eliminating systematic errors. Amplitude normalization is performed on the signals, that is, mapping all sampled values ​​to a unified amplitude range (such as 0-1 or –1-1) based on a set reference amplitude or historical maximum value. Throughout the acquisition and preprocessing process, the system establishes a sliding buffer mechanism to store the complete signal sequence of each round of acquisition into a circular buffer, ensuring the continuity and traceability of data in the processing flow. The buffer cycle length is dynamically adjusted according to the display refresh rate: when the refresh rate is 60Hz, the typical buffer cycle is set to 8 frames, and when the refresh rate is 120Hz, it is set to 16 frames to adapt to the processing requirements of high-frequency data streams.

[0033] In one specific embodiment, the process of performing step S200 may specifically include the following steps: By performing statistical analysis of the DC component of the mixed signal data over a continuous acquisition period, the initial reference value of the common-mode noise in the display driver circuit is obtained. An adaptive tracking algorithm is established based on the initial reference value, and the weighted average filter coefficient is dynamically adjusted according to the changes in the displayed content to obtain the common mode noise reference value. The mixed signal data is subtracted point by point from the common-mode noise reference value to eliminate the in-phase interference component, resulting in touch signal data containing differential-mode noise and touch information.

[0034] Specifically, with the goal of accurately identifying common-mode interference, common features in the signal are extracted from multi-cycle sampling data. A statistical mechanism combining long-term averages and short-term fluctuations is established to separate the DC component in the signal. Since mixed signal data contains common-mode noise, differential-mode noise, and touch response signals, and common-mode noise, as a periodically superimposed in-phase interference component in the display driver circuit structure, exhibits stable behavior and a highly consistent DC offset trend across multiple sampling cycles, the DC component is statistically analyzed by continuously acquiring signal data from multiple cycles to extract the initial common-mode reference value that best represents the systematic background interference. The system sets a time window to buffer continuous signal sampling data and performs a weighted average of the signal values ​​of all channels within this window to obtain the DC reference distribution of each channel under the current environment, forming a set of initial reference values ​​reflecting the background interference state of the display circuit. Because the displayed content alternates between static and dynamic images, the fluctuation of the driving voltage and its impact on common-mode noise change significantly under different display modes. Therefore, it is impossible to accurately reconstruct the common-mode state within the current frame cycle solely based on a fixed-weighted average value. To address this, the system introduces a first-order low-pass filter structure based on the initial reference values ​​to construct an adaptive tracking algorithm. This algorithm dynamically tracks common-mode noise through continuous updates. The filter structure uses a weighted average method for data updates, combining the latest signal value with the reference value from the previous period in each sampling cycle. The weighted average filter coefficient (α) used in the update formula is adaptively adjusted based on the changing characteristics of the currently displayed content: when the screen displays a static image, the common-mode interference is stable due to the slow changes in the display drive signal; therefore, the α value is set to a low value (e.g., 0.1) to improve output stability and reduce the impact of short-term fluctuations. However, when the display is in a rapid refresh or dynamic image switching phase, the in-phase noise caused by the display circuit will change frequently. By increasing the α value (e.g., 0.3), the filter response speed is accelerated, enabling it to quickly follow changes in the displayed content, dynamically correct the common-mode reference value, and thus ensure accurate fitting of the common-mode noise. After obtaining the time-updated common-mode noise reference value, the system enters the common-mode separation processing stage. This stage involves point-by-point subtraction of the mixed signal value at each sampling point with the common-mode reference value corresponding to the current period, thereby achieving real-time interval suppression of in-phase interference components. Because the interference characteristics of common-mode noise are highly synchronized across all channels, its distribution characteristics at all sampling points are approximately the same. This differential operation effectively eliminates it without affecting the retention of differential-mode components and touch signals. The processed signal data is the intermediate result after common-mode noise suppression, containing differential-mode interference generated by the periodic excitation of TFT pixel switches, as well as charge disturbances caused by user touch behaviors such as fingers and capacitive styluses.

[0035] In one specific embodiment, the process of performing point-by-point subtraction operations between the mixed signal data and the common-mode noise reference value to eliminate in-phase interference components and obtain touch signal data containing differential-mode noise and touch information can specifically include the following steps: A timing alignment mechanism between mixed signal data and common-mode noise reference value is established to ensure that each signal acquisition point matches the common-mode noise reference value at the corresponding time, thereby obtaining a time-synchronized data pair; Based on the time-synchronized data pairs, the corresponding common-mode noise reference value is subtracted from the signal amplitude value of each signal acquisition point to obtain the calculation result; Based on the calculation results, identify the in-phase interference components generated by the display driving circuit and mark abnormal data points that exceed the normal touch signal range to obtain in-phase interference component identification data; Based on the in-phase interference component identification data, the abnormal data points are filtered to eliminate the residual in-phase interference components, resulting in touch signal data containing differential noise and touch information.

[0036] Specifically, a timing pairing mechanism is constructed to ensure that the acquired mixed signal data corresponds to the common-mode noise reference value in the time dimension. Because common-mode noise is not a constant DC quantity, but exhibits slow or periodic fluctuations due to pixel loading order and scanning voltage changes within the display frame cycle and its sub-cycles, it is necessary to ensure that each sampling point in the mixed signal accurately corresponds to the common-mode reference value at the current sampling moment. Otherwise, even if the reference value itself is valid, the timing offset of the subtraction operation will cause residual enhancement or damage to the true signal. Therefore, before performing differential operations, the system uniformly schedules and registers the sampling timestamp, signal timing label, and common-mode reference value update cycle to ensure that each signal amplitude value in the acquisition sequence corresponds to the common-mode reference value at its position in the display cycle, thus forming a matched timing synchronization data pair. This matching mechanism relies on a multi-channel synchronization control strategy using a buffer queue or ring buffer structure, ensuring that each acquired sample is labeled with the corresponding reference value before entering the common-mode separation channel, eliminating calculation offsets caused by cross-cycle data mixing. The system performs a difference operation on each signal sampling point, subtracting the current common-mode noise reference value from the original amplitude value to obtain the difference output reflecting the net effective signal at that sampling point. Considering that some common-mode interference in the display circuit has abrupt characteristics, such as horizontal switching edges, voltage bounce, and capacitor sudden discharge, which can cause reference model mismatch in a short time, residual common-mode components are retained in the calculation results, or periodic erroneous judgments are introduced. Therefore, the system performs behavioral feature recognition on the above difference operation results. By comparing the current amplitude change trend with the system's preset touch signal standard response model, it analyzes the differences to identify abnormal data points that do not conform to the normal touch signal amplitude threshold, response time length, or change rate. Normal touch signals exhibit a continuous increase or decrease in amplitude, lasting for at least several sampling periods, and the peak amplitude has a stable physical upper limit, while common-mode interference manifests as instantaneous fluctuations, periodic pulses, or random spikes. Based on this, the system sets multi-dimensional discrimination conditions, filtering out sampling points from the difference results in each round that exceed the upper limit of signal amplitude, have excessively short duration, or exhibit obvious repetition periodic characteristics. These points are then marked as in-phase interference, generating in-phase interference component identification data containing information on the location of the anomaly, the degree of amplitude anomaly, and the frequency of occurrence. After identification, the system enters a targeted filtering phase. Instead of relying on a unified filter across all channels, it performs local processing on the anomaly points based on the identification data, prioritizing methods such as windowed median filtering, band-limited differential filtering, or conditional notch filters to suppress or replace signals at specific intensities. In median filtering, a dynamic window containing multiple preceding and following normal points is set to locally reconstruct the anomaly point; in band-limited differential filtering, the abnormal frequency band is attenuated based on spectral characteristics; if the identification data indicates that the anomaly has fixed frequency characteristics, a local notch filter is activated for narrowband cancellation.After local filtering correction of all outliers, a purer signal sequence is obtained, which no longer contains common-mode interference components, but retains differential-mode noise and real touch signals.

[0037] In one specific embodiment, the process of performing step S300 may specifically include the following steps: Fast Fourier transform analysis and signal power spectral density calculation are performed on the touch signal data to obtain frequency domain analysis data containing the switching frequency characteristics of the pixel TFT; Differential mode noise frequency components are identified based on frequency domain analysis data and pixel driving voltage to obtain frequency characteristic parameters of differential mode noise. The notch frequency and notch depth parameters of the adaptive notch filter are set based on the frequency characteristic parameters, and the touch signal data is filtered and separated by the adaptive notch filter to obtain the differential noise separation result and the multi-touch signal.

[0038] Specifically, frequency domain analysis is performed on the touch signal data to reveal the hidden differential-mode noise characteristics. This type of noise is caused by the switching behavior of the TFT (Thin Film Transistor) array. Its signal exhibits irregular abrupt changes or periodic perturbations in the time domain, while displaying a specific concentrated frequency distribution in the frequency domain. Therefore, a Fast Fourier Transform (FFT) is used to transform and analyze the touch signal sequence to extract the differential-mode noise component. To obtain good frequency domain resolution and spectral energy density distribution, the system divides the time-domain touch signal into fixed-length time windows and performs a FFT operation on the data within each window. A typical window length is set to 1024 points, with an overlap rate of 50%, thereby improving the continuity and stability of the spectral data while balancing frequency resolution and temporal locality. After performing the Fourier transform, a complex-form frequency domain output signal is obtained. By calculating the square of its magnitude and combining it with the window normalization factor, the signal power spectral density (PSD) curve is obtained. This curve describes the distribution of signal energy in different frequency bands. In PSD analysis, the system focuses on energy-dense, frequency-stable spectral peaks that repeat across multiple time periods. These peaks correspond to the differential-mode noise components generated during the periodic switching of the pixel array. Since each frame of the image activates numerous TFT elements for pixel loading during display, and these loading operations are affected by factors such as row scan timing, column data switching, and charge transport paths, they manifest as clusters of interference peaks in the frequency domain. These interference peaks are located in typical frequency ranges, such as 60Hz, 120Hz, or their harmonics, and within some non-integer harmonic bands. The system compares this frequency domain spectrum with the frequency model of the pixel driving voltage waveform, identifies frequency bands with high overlap as differential-mode interference frequency intervals, and extracts elements such as the dominant frequency, bandwidth, harmonic distribution, and signal amplitude to form a set of differential-mode noise frequency characteristic parameters. An adaptive notch filter structure is constructed based on frequency characteristic parameters. This filter centers on the dominant frequency of differential-mode noise, setting a notch bandwidth of a certain width around this frequency. The notch depth is set according to the noise intensity parameter, with a notch depth of at least -40dB to ensure that interference signals are sufficiently suppressed without affecting the integrity of other frequency components. The center frequency of the notch filter is automatically adjusted based on PSD analysis results, supporting multi-frequency notch filtering and dynamic bandwidth control. That is, when multiple interference peaks are detected simultaneously, the filter constructs multiple notch channels or uses a composite notch structure for parallel filtering. To avoid attenuating the actual touch signal, the notch filter adopts a minimum phase design, maintaining minimal disturbance to the time-domain waveform and ensuring that the timing structure and energy boundary of the touch response signal are not destroyed. Simultaneously, to achieve the algorithm's adaptability under different display refresh rates, different panel structures, and different touch densities, the system introduces a feedback adjustment mechanism. The power spectrum before and after filtering is compared and analyzed, and the filter parameters are adjusted accordingly to improve the differential-mode noise suppression efficiency and the filter's tracking stability.After configuring the adaptive notch filter and applying it to the original touch signal, the system outputs two types of data: one is the high-purity touch signal data after being suppressed by the filter, and the other is the residual data of differential-mode interference weakened by the filter in the frequency domain. This differential-mode noise separation process has good system scalability and algorithm adjustability. In practical applications, it can be customized according to different panel materials, electrode layouts, driver chip models, and user behavior models. For example, in high-resolution OLED panels, due to higher pixel density and TFT activation frequency, the system expands the frequency domain window to 2048 points to enhance spectral resolution; while in low refresh rate scenarios, the center frequency tracking rate of the filter is appropriately reduced to reduce the computational burden.

[0039] In one specific embodiment, the process of setting the notch frequency and notch depth parameters of the adaptive notch filter based on frequency characteristic parameters, and filtering and separating the touch signal data through the adaptive notch filter to obtain the differential mode noise separation result and the multi-touch signal can specifically include the following steps: The center frequency position of the adaptive notch filter is calculated based on the frequency components in the frequency characteristic parameters, and the notch frequency of the adaptive notch filter is set according to the center frequency position. The notch depth parameter of the adaptive notch filter is set according to the differential mode noise signal strength in the frequency characteristic parameters. The touch signal data is input into an adaptive notch filter for filtering to obtain filtered output data. The differential noise component and the touch signal component are then separated and identified from the filtered output data to obtain the differential noise separation result and the multi-touch signal.

[0040] Specifically, the system extracts differential-mode noise frequency characteristic parameters from the frequency domain data. These parameters include the dominant frequency position, frequency bandwidth, harmonic structure, and relative energy intensity. The dominant frequency of differential-mode noise appears near integer multiples of 60Hz, 120Hz, and 240Hz, accompanied by a certain non-integer harmonic distribution. Therefore, the system locates the frequency position corresponding to the energy peak in the spectral data and performs frequency mapping calibration based on the timing parameters in the pixel TFT driving model to determine the most likely dominant frequency position of the differential-mode noise source. This frequency is used as the center frequency of the adaptive notch filter, and the notch passband position is set accordingly. Considering that the dominant frequency of differential-mode noise may shift slightly due to timing fluctuations or changes in display load, the system configures an adjustable bandwidth mechanism for the notch filter. Based on the signal intensity of the differential-mode noise in the frequency characteristic parameters, i.e., the ratio of the peak power spectral density of the noise band to the average energy of the surrounding frequency bands, the system dynamically sets the notch depth parameter of the notch filter. This parameter directly determines the strength of the filter's suppression of the target frequency energy. If the differential-mode noise intensity is significantly higher than the touch signal amplitude, the system sets the notch depth to a higher value, between -40dB and -60dB, to completely suppress strong interference signals. If the spectral energy is relatively similar, the notch depth is appropriately reduced to around -20dB to minimize interference with nearby real-world signals. To adapt to rapid frequency changes in high-resolution displays or high refresh rate scenarios, the adaptive filter's depth parameter has dynamic recalculation capabilities, periodically refreshing by monitoring noise energy fluctuations to enhance the algorithm's real-time responsiveness. After dynamically configuring the center frequency and notch depth, the system inputs the preprocessed touch signal data into the adaptive notch filter for filtering. The filter uses either an IIR (Infinite Impulse Response) or FIR (Finite Impulse Response) structure, depending on the processing platform's computing resources and real-time requirements. For high-precision filtering tasks, an IIR notch filter is preferred. Its implementation uses a double second-order structure superposition for fine control, ensuring the filter has a sharp attenuation response in the target frequency band while maintaining a flat gain in other frequency bands, preventing the real touch signal from being incorrectly filtered. After the filtering operation is completed, the output signal obtained by the system is the purified signal after frequency attenuation. In order to complete the differential noise separation and touch signal extraction, the filtered output data is subjected to component identification analysis. The system compares the signals before and after filtering, and extracts the filtered signal part through difference analysis. This part is the identified differential noise residual; at the same time, the waveform retained after filtering is regarded as the target multi-touch signal.To ensure the validity of the separation results, the system performs spectral verification on the extracted differential mode noise residual, requiring that its frequency components be highly consistent with the initially set notch frequency, and that the frequency deviation not exceed the set tolerance range. At the same time, the system performs time continuity verification and amplitude stability detection on the retained touch signal data to confirm that it has normal touch response characteristics, such as continuity, gradual rise / fall edge characteristics, and local area amplitude peak characteristics, in order to eliminate high-frequency pseudo responses that are not completely filtered out.

[0041] In one specific embodiment, the process of performing step S400 may specifically include the following steps: The multi-touch signal is divided into multiple spatial partitions according to the touch detection area, and the touch signal data of each spatial partition is acquired; Touch signal threshold detection is performed on the touch signal data to determine the potential touch point data of each spatial partition; Based on the potential touch point data, the signal correlation between adjacent zones is analyzed and the correlation between each potential touch point and the signals of adjacent zones is calculated to obtain the touch point correlation relationship; Based on the correlation of touch points, the centroid algorithm is used to calculate the centroid position of the touch signal in spatial distribution, and obtain mixed coordinate data including real touch points and phantom points.

[0042] Specifically, based on the geometric structure of the physical touch panel, the entire touch detection area is divided into multiple spatial partitions according to fixed logic, making each partition an independent signal analysis unit, thereby constructing a localized, parallel-processing touch recognition framework. The entire touch area is divided into several rectangular or square grids according to a coordinate grid division method, such as dividing it into 64 sub-regions in an 8×8 manner. Each sub-region corresponds to a set of capacitive signal channels, and its sampled data reflects the intensity, waveform, and timing characteristics of the touch response within that local area. After signal acquisition, the system maps the signals of each channel to the corresponding spatial units according to the divided partitions, forming a structured data distribution model. After the division is completed and the signal data of each region is acquired, the system performs threshold detection on the signal amplitude based on the set touch recognition logic to initially screen potential touch points. This threshold is a signal judgment threshold dynamically calculated based on the background noise obtained after the previous noise suppression processing, and is set as the upper limit of the range of the noise mean plus 3 times the standard deviation to ensure sufficient suppression of false triggering and misjudgment. During the threshold determination process, the system detects whether the signal amplitude exceeds the threshold within each partition and records the sampling points that meet the conditions as potential touch points, while marking the center coordinates of the area where the point is located as a preliminary coordinate estimate. However, under conditions of multi-point dense touch or electromagnetic interference, false triggering of neighboring areas due to capacitive coupling or signal crosstalk is very likely to occur. To address this, the system introduces a spatial correlation analysis mechanism to compare and cross-validate the signals of adjacent areas for all potential touch points. In this stage, the system extracts touch signal data from eight spatial neighborhoods around each potential touch point, calculates its correlation index with the surrounding area signals, and constructs the index using normalized cross-correlation functions, signal energy similarity, or consistency of change trends, and expresses it numerically as the degree of consistency between the touch point and its surrounding area. If the correlation reaches a preset threshold, such as above 0.85, it indicates that the point has strong spatial continuity and conforms to the diffusion characteristics of real physical contact behavior; conversely, if the point has a large amplitude but no obvious synchronicity with the surrounding area, further verification is required. Once the spatial relationships of all potential touch points are identified and stored, the system enters the real coordinate calculation stage based on the centroid algorithm. This stage comprehensively considers the signal intensity distribution of each touch point and its surroundings, accurately calculating the centroid coordinates of the local touch area using a spatial weighted method. The basic logic of centroid calculation is to use each touch point as its centroid, perform weighted summation based on its signal intensity, and map it to the average position point on a two-dimensional coordinate plane using a formula. After the centroid coordinate calculation is completed, the system obtains a set of coordinates containing all identified touch events, where each set of coordinates is contributed by multiple locally highly correlated signals, thus simultaneously covering real touch points and phantom points formed by signal coupling and other factors.

[0043] In one specific embodiment, the process of executing step S500 may specifically include the following steps: Based on the differential noise separation results, the signal purity of each coordinate point is analyzed and a phantom point signal strength discrimination threshold is set to obtain the signal strength discrimination benchmark. Calculate the spatial geometric relationship between the distances between touch points in the mixed coordinate data to identify abnormal coordinate combinations where the distance between adjacent touch points is less than a preset value; The signal strength of each coordinate point in the mixed coordinate data is compared with the signal strength discrimination benchmark to filter out coordinates with abnormal signal strength. Based on the intersection operation of abnormal coordinate combinations and abnormal signal strength coordinates, the coordinate points that simultaneously violate spatial constraints and signal strength requirements are determined as phantom points, and the phantom point identification results are obtained. Remove the phantom point coordinates determined by the phantom point recognition results from the mixed coordinate data to obtain the touch positioning results.

[0044] Specifically, based on the differential-mode noise separation results, the signal amplitude corresponding to each touch coordinate point is extracted from the retained multi-touch signal as an important parameter for determining its validity and interference level. The differential-mode noise separation process has eliminated the main periodic noise components through frequency analysis and notch filtering. Therefore, in the filtered signal, coordinate points with higher amplitudes and more continuous waveforms represent capacitive responses caused by genuine touch behavior, while signals with lower amplitudes and no local adjacency originate from non-real behaviors such as crosstalk, mirror response, or panel noise coupling. The system normalizes the amplitude values ​​of each coordinate point and statistically analyzes the mean and standard deviation of the signal strength across the entire set of coordinate points. Combined with the overall noise level in the sampling frame, a dynamic threshold is set as the signal strength discrimination benchmark for distinguishing between high-confidence touch signals and suspected phantom signals. This threshold is set to be higher than 70% of the overall signal mean or a median weight between the mean and maximum values ​​to avoid false coordinates caused by extremely small amplitude responses. The system analyzes the geometric relationships between coordinate points from a spatial structure perspective. Touch behavior is inherently limited by human operating scale in physical space. Therefore, two independent and effective touch points will not be infinitely close to each other on the screen surface, especially in scenarios where fingers are used for operation. A minimum distance of at least 10 millimeters should be maintained between real touch points. Phantom points, on the other hand, are often induced by coupling phenomena, exhibiting a characteristic of being closely aligned with the main signal and distributed in a regular geometric symmetry. Based on this understanding, the system calculates pairwise Euclidean distances for all mixed coordinate data. Using a distance matrix, it quickly identifies coordinate combinations in all point pairs where the distance between them is less than a preset minimum geometric threshold, marking these combinations as potential spatial anomalies. During this process, the system further detects whether these anomaly point pairs form regular geometric patterns such as equilateral, rectangular, or square shapes to enhance the geometric reliability of phantom point pattern recognition. The system compares each coordinate point in the identified spatial anomaly point pairs with a previously set signal strength discrimination benchmark, filtering out points with signal strength below the threshold; these points are defined as signal strength anomaly coordinates. The system uses an intersection operation strategy to logically intersect and merge combinations of spatial geometric anomalies and signal strength anomalies, extracting coordinate points that simultaneously satisfy the characteristics of proximity and weak signal as high-confidence phantom points. The phantom point coordinates determined by the phantom point identification results are removed from the mixed coordinate data, retaining the remaining verified high-confidence real coordinate points to form the final touch positioning output. To avoid erroneously deleting real points, the system performs multi-frame time window verification on points to be removed, checking their continuity and positional stability across several frames. If a point persists across multiple periods and exhibits time-varying trajectory characteristics, it is determined to have some real component and is not deleted, but instead marked as a weak-confidence point for further processing by upper-layer applications.

[0045] In one specific embodiment, the process of determining phantom points by performing an intersection operation based on anomalous coordinate combinations and signal strength anomaly coordinates to identify phantom points can specifically include the following steps: Each coordinate point in the abnormal coordinate combination is matched one by one with each point in the abnormal signal strength coordinate to obtain coordinate matching relationship data; Based on the coordinate matching relationship data, overlapping coordinate points that exist simultaneously in abnormal coordinate combinations and abnormal signal strength coordinates are selected to obtain a set of dual violation coordinates. For each coordinate point in the set of double-violation coordinates, calculate the degree of spatial constraint violation and signal strength deviation, and establish phantom point confidence data; Based on the phantom point confidence data, coordinate points that simultaneously violate spatial constraints and signal strength requirements are identified as phantom points, and phantom point recognition results are generated.

[0046] Specifically, anomalous coordinate combinations obtained through spatial geometric analysis are represented by pairs or groups of triplets of coordinates whose distances to each other are less than the minimum physical allowable spacing (e.g., 10 mm). Signal strength anomaly coordinate sets, filtered through amplitude statistics and dynamic threshold determination, represent weak signal points in the current frame whose signal amplitude is below 70% of the average effective touch value. Through a coordinate-level one-to-one matching operation, each point in the anomalous combination is compared with each point in the strength anomaly set. For example, a coordinate error within 2 pixels is considered the same point to accommodate sampling errors and numerical quantization offsets. During the traversal, each time a duplicate point appearing in both sets is identified, a coordinate matching relationship is recorded, and a mapping table of "coordinate point → appearance set label" is established. After establishing the matching relationship, intersection points that simultaneously appear in the anomalous coordinate combination and the signal strength anomaly set are selected from the matching results; these are "double violation coordinate points." These coordinate points violate the adjacency constraint in spatial structure and exhibit weak signal characteristics in electrical signal behavior, possessing a very high probability of being phantom points. The system incorporates these points into a key marker set and proceeds to the confidence assessment phase. In this phase, two evaluation dimensions are constructed for each double-violation coordinate point: the first is the "degree of spatial constraint violation," calculated as the ratio of the actual distance between the point and its neighbors to the minimum legal distance. For example, if the distance is only 4mm while the minimum limit is 10mm, the violation rate is 60%. The second is the "degree of signal strength deviation," normalized by the difference between the current point's amplitude and the average amplitude of all valid points in the current frame, and reflected as a deviation ratio. For example, if the average effective amplitude is 1.0, and the point's amplitude is 0.3, the deviation is 70%. Both can be normalized to percentage indicators. The system jointly inputs these two deviation indicators into the phantom point confidence function model. The confidence score is defined as the probability score of a point being a phantom point. The calculation formula is C = α × P_space + β × P_intensity, where P_space is the spatial deviation ratio, P_intensity is the intensity deviation ratio, and α and β are weighting coefficients, set to 0.5:0.5 or 0.4:0.6 in the system design, with the specific ratio determined through experimental calibration to reflect the relative influence of signal features in the recognition decision. When the C value is higher than a preset threshold, such as 0.75 or 0.8, the system classifies the point as a phantom point and records it in the phantom point recognition result dataset. During the confidence score construction process, a third dimension parameter, namely "repetition frequency" or "temporal consistency," is added to improve model stability. For example, whether the point appears continuously in three consecutive sampling periods and whether its confidence score shows an upward trend are statistically analyzed. If cross-frame consistency exists, the phantom judgment intensity is appropriately reduced to avoid misclassifying stable but weakly amplitude boundary touch points as phantom points.Simultaneously, the system should include a confidence score and source label for coordinate points identified as phantom points, allowing upper-layer calling modules to adjust the system in different application scenarios. For example, in drawing, input, drag-and-drop, or anti-mistouch modes, the processing strategy can be dynamically adjusted based on the confidence threshold. The system outputs all coordinate points that meet both violation criteria and have a confidence score higher than the set threshold to the phantom point recognition result set, marking them as objects to be removed. These coordinates are then removed from the officially generated touch positioning data, achieving data purification.

[0047] Please see Figure 2 , Figure 2 This is a schematic block diagram of the signal processing system for a touch-integrated display module provided in an embodiment of this application, such as... Figure 2 As shown, the touch-integrated display module signal processing system includes: Signal acquisition module 211 is used to acquire signals from the display driving circuit of the touch display module during the display frame interval to obtain mixed signal data; The common-mode noise separation module 222 is used to separate the mixed signal data according to the common-mode noise reference to obtain touch signal data; The identification module 233 is used to identify the differential noise component generated by the pixel driving voltage based on the touch signal data, and obtain the differential noise separation result and the multi-point touch signal; Extraction module 244 is used to extract mixed coordinate data containing real touch points and phantom points based on multi-touch signals; The verification module 255 is used to perform spatial geometric constraint verification and signal strength verification on phantom points in mixed coordinate data based on differential noise separation results, so as to obtain touch positioning results.

[0048] Through the synergistic cooperation of the aforementioned components, this invention overcomes the limitation of the independent processing of two processes in the prior art by establishing a collaborative processing mechanism for differential-mode noise separation and phantom point elimination. The differential-mode noise separation result provides a signal purity benchmark for phantom point recognition, while the spatial geometric constraint verification in the phantom point elimination process, in turn, optimizes the recognition accuracy of differential-mode noise, achieving mutual promotion and synergistic optimization of the two processing links. By establishing a timing synchronization mechanism during the display frame interval, this invention achieves precise timing matching between touch signal acquisition and display driving signals, effectively avoiding interference from the display refresh process on touch detection. This timing synchronization design ensures the accuracy of signal acquisition timing. This invention establishes an adaptive common-mode noise benchmark tracking algorithm and a differential-mode noise frequency domain feature recognition algorithm, which can dynamically adjust processing parameters according to changes in display content and pixel driving voltage characteristics. This adaptive mechanism ensures that the noise separation effect is unaffected by changes in display content, maintaining long-term stable processing performance. By performing gridded spatial partitioning of the touch detection area, this invention avoids capacitive coupling interference between adjacent touch points, improving the detection accuracy of multi-touch. The spatial partitioning mechanism combined with the centroid algorithm for coordinate positioning enables precise calculation and coordinate extraction of touch point positions. This invention establishes a multi-verification mechanism involving spatial geometric constraint verification, signal strength testing, and intersection operation confirmation. Through comprehensive judgment across multiple dimensions, the accuracy of phantom point recognition is ensured, effectively reducing the false positive rate and improving the reliability of the touch system.

[0049] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0050] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0051] The above-described embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application.

Claims

1. A signal processing method for a touch-integrated display module, characterized in that, include: During the display frame interval, the display driving circuit of the touch display module is sampled to obtain mixed signal data; The mixed signal data is separated into common-mode noise according to a common-mode noise reference to obtain touch signal data; Based on the touch signal data, the differential mode noise component generated by the pixel driving voltage is identified to obtain the differential mode noise separation result and the multi-touch signal; Based on the multi-touch signal, extract mixed coordinate data containing real touch points and phantom points; Based on the differential noise separation results, the phantom points in the mixed coordinate data are subjected to spatial geometric constraint verification and signal strength test to obtain the touch positioning results.

2. The signal processing method for a touch-integrated display module according to claim 1, characterized in that, The step of acquiring signals from the display driving circuit of the touch display module during the display frame interval to obtain mixed signal data includes: Establish a timing synchronization mechanism between the display refresh cycle and the touch scan cycle, and collect timing control signals for the display frame interval; According to the timing control signal, during the horizontal blanking period, the signal acquisition window is opened to synchronously detect the differential signal generated by the display driving circuit; The differential signal is digitized to obtain an original digital signal containing common-mode noise, differential-mode noise, and touch information; The original digital signal is preprocessed by DC bias elimination and signal amplitude normalization to obtain mixed signal data containing common-mode noise, differential-mode noise and touch information.

3. The signal processing method for a touch-integrated display module according to claim 1, characterized in that, The step of separating the mixed signal data according to a common-mode noise benchmark to obtain touch signal data includes: Statistical analysis of the DC component of the mixed signal data over a continuous acquisition period is performed to obtain the initial reference value of the common-mode noise in the display driving circuit. An adaptive tracking algorithm is established based on the initial reference value, and the weighted average filter coefficient is dynamically adjusted according to the changes in the displayed content to obtain the common mode noise reference value. The mixed signal data is subtracted point by point from the common-mode noise reference value to eliminate the in-phase interference component, thereby obtaining touch signal data containing differential-mode noise and touch information.

4. The signal processing method for a touch-integrated display module according to claim 3, characterized in that, The step of subtracting the mixed signal data from the common-mode noise reference value point by point to eliminate in-phase interference components, resulting in touch signal data containing differential-mode noise and touch information, includes: A timing alignment mechanism is established between the mixed signal data and the common-mode noise reference value to ensure that each signal acquisition point matches the common-mode noise reference value at the corresponding time, thereby obtaining a timing-synchronized data pair; Based on the time-synchronized data pairs, the signal amplitude value of each signal acquisition point is subtracted from the corresponding common-mode noise reference value to obtain the calculation result; Based on the calculation results, identify the in-phase interference components generated by the display driving circuit and mark abnormal data points that exceed the normal touch signal range to obtain in-phase interference component identification data; Based on the in-phase interference component identification data, the abnormal data points are filtered to eliminate residual in-phase interference components, resulting in touch signal data containing differential noise and touch information.

5. The signal processing method for a touch-integrated display module according to claim 1, characterized in that, The step of identifying the differential-mode noise component generated by the pixel driving voltage based on the touch signal data to obtain the differential-mode noise separation result and the multi-touch signal includes: Fast Fourier Transform analysis and signal power spectral density calculation are performed on the touch signal data to obtain frequency domain analysis data containing the switching frequency characteristics of the pixel TFT. Based on the frequency domain analysis data and pixel driving voltage, the frequency components of differential mode noise are identified to obtain the frequency characteristic parameters of differential mode noise. Based on the frequency characteristic parameters, the notch frequency and notch depth parameters of the adaptive notch filter are set, and the touch signal data is filtered and separated by the adaptive notch filter to obtain the differential noise separation result and the multi-touch signal.

6. The signal processing method for a touch-integrated display module according to claim 5, characterized in that, The process involves setting the notch frequency and notch depth parameters of an adaptive notch filter based on the frequency characteristic parameters, and then filtering and separating the touch signal data using the adaptive notch filter to obtain differential mode noise separation results and multi-touch signals, including: The center frequency position of the adaptive notch filter is calculated based on the frequency components in the frequency characteristic parameters, and the notch frequency of the adaptive notch filter is set based on the center frequency position. The notch depth parameter of the adaptive notch filter is set according to the differential mode noise signal strength in the frequency characteristic parameters. The touch signal data is input into the adaptive notch filter for filtering to obtain filtered output data. The differential noise component and the touch signal component are then separated and identified in the filtered output data to obtain the differential noise separation result and the multi-touch signal.

7. The signal processing method for a touch-integrated display module according to claim 1, characterized in that, The step of extracting mixed coordinate data containing real touch points and phantom points based on the multi-touch signal includes: The multi-touch signal is divided into multiple spatial partitions according to the touch detection area, and the touch signal data of each spatial partition is obtained; The touch signal data is subjected to touch signal threshold detection to determine the potential touch point data of each spatial partition; Based on the potential touch point data, analyze the signal correlation between adjacent partitions and calculate the correlation between each potential touch point and the signals of adjacent partitions to obtain the touch point correlation relationship; Based on the aforementioned touch point association relationship, a centroid algorithm is used to calculate the centroid position of the touch signal in spatial distribution, resulting in mixed coordinate data containing real touch points and phantom points.

8. The signal processing method for a touch-integrated display module according to claim 1, characterized in that, The spatial geometric constraint verification and signal strength test are performed on the phantom points in the mixed coordinate data based on the differential noise separation results to obtain the touch positioning results, including: Based on the differential noise separation results, the signal purity of each coordinate point is analyzed and a phantom point signal strength discrimination threshold is set to obtain the signal strength discrimination benchmark. The spatial geometric relationship calculation is performed on the distance between touch points in the mixed coordinate data to identify abnormal coordinate combinations where the distance between adjacent touch points is less than a preset value; The signal strength of each coordinate point in the mixed coordinate data is compared with the signal strength discrimination benchmark to filter out coordinates with abnormal signal strength. Based on the intersection operation of the abnormal coordinate combination and the abnormal signal strength coordinate, the coordinate point that simultaneously violates the spatial constraints and signal strength requirements is determined as the phantom point, and the phantom point identification result is obtained. Remove the phantom point coordinates determined by the phantom point recognition result from the mixed coordinate data to obtain the touch positioning result.

9. The signal processing method for a touch-integrated display module according to claim 8, characterized in that, The process of determining the coordinate points that simultaneously violate both spatial constraints and signal strength requirements as phantom points through intersection operations based on the abnormal coordinate combinations and the abnormal signal strength coordinates, and obtaining phantom point identification results, includes: Each coordinate point in the abnormal coordinate combination is matched one by one with each point in the abnormal signal strength coordinate to obtain coordinate matching relationship data; Based on the coordinate matching relationship data, overlapping coordinate points that exist simultaneously in abnormal coordinate combinations and abnormal signal strength coordinates are selected to obtain a set of dual violation coordinates. For each coordinate point in the set of dual violation coordinates, calculate the degree of violation of spatial constraints and the degree of signal strength deviation, and establish phantom point confidence data; Based on the phantom point confidence data, coordinate points that simultaneously violate spatial constraints and signal strength requirements are identified as phantom points, and phantom point recognition results are generated.

10. A signal processing system for a touch-integrated display module, characterized in that, A method for performing signal processing of a touch-integrated display module as described in any one of claims 1-9, comprising: The signal acquisition module is used to acquire signals from the display driving circuit of the touch display module during the display frame interval to obtain mixed signal data; A common-mode noise separation module is used to separate the mixed signal data according to a common-mode noise reference to obtain touch signal data; The identification module is used to identify the differential noise components generated by the pixel driving voltage based on the touch signal data, and obtain the differential noise separation result and the multi-point touch signal; The extraction module is used to extract mixed coordinate data containing real touch points and phantom points based on the multi-touch signal; The verification module is used to perform spatial geometric constraint verification and signal strength verification on the phantom points in the mixed coordinate data based on the differential noise separation results, so as to obtain the touch positioning results.

Images