Passive capacitance pen, electronic writing screen and capacitance non-contact touch identification method

Abstract

The invention discloses a passive capacitance pen, an electronic writing screen and a capacitance non-contact touch identification method, and relates to the field of touch equipment. The passive capacitance pen comprises a pen holder, a pen point and a tail plug; the pen holder, the pen point and the tail plug are all insulating parts; the pen point is connected with one end of the pen holder; the tail plug is connected with the other end of the pen holder; a groove is formed in the end, close to the pen point, of the pen holder, and a capacitive sensing component is contained in the groove. The touch equipment is high in anti-interference capability, simple in structure, low in cost, replaceable in part, free of charging and higher in positioning precision and sensitivity.

Description

Technical Field

[0001] The present invention relates to the technical field of touch devices, and in particular to a passive capacitive pen, an electronic writing screen, and a capacitive contactless touch recognition method. Background Art

[0002] In the related art, most touch devices use resistive sensor technology, passive capacitance technology or active capacitance technology. Among them, the resistive sensor technology uses a resistive stylus to apply pressure to a resistive film made of transparent conductive material covering the surface of the screen. The resistance value between the pen tip and the screen will change, and the touch position is determined by detecting this resistance change. However, this principle of relying on pressure as the detection of touch position and writing trajectory makes it very easy to be interfered with by external factors when writing, such as hand pressure, object squeezing and collision, etc., causing writing errors. Passive capacitance technology uses a passive capacitive stylus on a capacitive touch screen. By imitating the touch effect of a finger, the pen tip is a conductive material, and when it comes into contact with the touch screen, it can cause the capacitance of the touch screen matrix to change. However, the conductive materials involved in this method will have the problem of fast wear (conductive silicone material) or easy scratching of the screen (metal alloy material), and this method needs to detect capacitance changes at different frequencies to distinguish the signal source. At this time, high-density sensor support is required, and the cost is relatively high. Active capacitive technology uses a stylus with a built-in power supply and communication module to interact with the capacitive screen by actively sending signals to achieve accurate writing, drawing and other operations. However, this method requires charging and cannot have unlimited battery life.

[0003] Therefore, there is an urgent need for a capacitive contactless touch recognition technology that has strong anti-interference ability, simple structure, low cost, replaceable parts, and does not require charging. Summary of the Invention

[0004] In view of this, the present invention provides a passive capacitive pen, an electronic writing screen, and a capacitive contactless touch recognition method to solve the technical problems existing in the related art.

[0005] In a first aspect, the present invention provides a passive capacitive stylus, comprising: a pen body, a pen tip and a tail plug; the pen body, pen tip and tail plug are all insulating components; the pen tip is connected to one end of the pen body; the tail plug is connected to the other end of the pen body; the pen body is provided with a groove at one end near the pen tip, and a capacitive sensing component is accommodated in the groove.

[0006] In an optional embodiment, the pen body and the pen tip are integrally formed or detachably connected, and the tail plug is integrally formed or detachably connected to the pen body.

[0007] In an optional embodiment, the groove is provided with a smooth transition angle at an edge of one end close to the tail plug.

[0008] In an optional embodiment, the groove is shaped like a cylindrical structure that is recessed along the circumference of the pen shaft; the capacitive sensing component is shaped like a cylindrical structure that matches the shape of the groove; and the capacitive sensing component is sleeved on the groove of the pen shaft.

[0009] In an optional embodiment, an insulating casing is further included; the insulating casing is sleeved on at least a portion of the capacitive sensing component.

[0010] In a second aspect, the present invention further provides an electronic writing screen for use with a passive capacitive stylus, comprising:

[0011] A signal generator, used to generate a pulse signal;

[0012] A capacitive contactless sensing array is electrically connected to the signal generator and is used to detect a position change of the non-contact capacitive stylus after receiving the pulse signal, and to generate a feedback signal according to the position change of the non-contact capacitive stylus;

[0013] a detection module, electrically connected to the capacitive contactless sensing array, configured to receive and analyze the feedback signal and generate an analog signal including a resonant frequency offset;

[0014] An A / D converter, electrically connected to the detection module, for converting the analog signal into a digital signal;

[0015] a processor, electrically connected to the A / D converter, configured to receive the digital signal, analyze and process the digital signal using a machine learning algorithm, and output position information and angle information of the passive capacitive stylus;

[0016] The display module is used to display the writing track of the passive capacitive pen in real time according to the position information and angle information of the passive capacitive pen.

[0017] In an optional embodiment, the capacitive contactless sensing array includes multiple sensing sub-modules; each sensing sub-module is electrically connected to the detection module and the signal generator respectively; each sensing sub-module includes multiple capacitive contactless sensing array units, each capacitive contactless sensing array unit includes a capacitor and an inductor, one end of the capacitor and one end of the inductor are connected in series, the other end of the capacitor is grounded, and the other end of the inductor is electrically connected to the detection module and the signal generator respectively through a microstrip line.

[0018] In an optional embodiment, the resonant frequency of each capacitive contactless sensing array unit in the sensing submodule is different;

[0019] The signal generator and the electronic switch inside the detection module are both electrically connected to each sensing submodule in the capacitive contactless sensing array in a time-division multiplexing connection mode.

[0020] In a third aspect, the present invention further provides a capacitive contactless touch recognition method applied to an electronic writing screen, comprising:

[0021] Obtaining the digital signal output by the A / D converter corresponding to when the passive capacitive pen writes on the electronic writing screen;

[0022] The digital signal is input into a pre-trained CNN-LSTM hybrid network to obtain position information and angle information of the passive capacitive stylus; the CNN-LSTM hybrid network includes a CNN network and an LSTM network cascaded with the CNN network.

[0023] In an optional embodiment, the pre-training process of the CNN-LSTM hybrid network is:

[0024] Acquire a first sample data set; the first sample data set includes multiple first sample data subsets; the target first sample data subset includes a frequency offset value and a label value of a target passive capacitive stylus at a target position and a target angle at a target time; the label value is the actual landing point coordinates and the actual pen holder tilt angle of the target passive capacitive stylus at the target position and the target angle at the target time; the target time includes the current time and any time before the current time;

[0025] Preprocessing the target first sample data subset; the preprocessing includes normalization, filtering and imaging;

[0026] Input the preprocessed target first sample data subset into the CNN network to obtain the predicted landing point coordinates and the predicted pen tilt angle of the target passive capacitive stylus at the target position and target angle at the target time;

[0027] When the predicted landing point coordinates and the predicted pen tilt angle of the target passive capacitive stylus at the target position and target angle at the target moment meet the first expected value, a pre-trained CNN network is obtained;

[0028] Filtering the predicted landing point coordinates and predicted pen shaft tilt angles of a plurality of passive capacitive styli on different continuous writing trajectories from the first sample data set to obtain a second sample data set; the second sample data set includes a plurality of second sample data subsets; the target second sample data subset includes the predicted landing point coordinates and predicted pen shaft tilt angles of a target passive capacitive stylus on a target writing trajectory at a time before a current time;

[0029] Input the target second sample data subset into the LSTM network to obtain the predicted landing point coordinates and predicted pen tilt angle of the target passive capacitive pen on the target writing trajectory at the current moment;

[0030] When the predicted landing point coordinates and the predicted pen tilt angle of the target passive capacitive pen on the target writing trajectory at the current moment meet the second expected value, a pre-trained LSTM network is obtained;

[0031] The pre-trained CNN network and the pre-trained LSTM network are cascaded to obtain a pre-trained CNN-LSTM hybrid network.

[0032] The embodiments of the present invention have the following beneficial effects:

[0033] The passive capacitive pen of the present invention has a simple structure. It only requires a groove on the pen body to accommodate the capacitive sensing component. It does not require charging, and the data processing part is transferred to the electronic writing screen. At the same time, the pen tip can be freely replaced according to usage, which is low-cost. At the same time, the use of a capacitive non-contact capacitance sensing array as the sensing element of the passive capacitive pen, combined with the machine learning algorithm built into the processor, not only makes the processor more sensitive only to the conductor material of the capacitive sensing component, but also has better anti-interference ability to capacitance changes caused by other unrelated conductors (such as fingers), and also improves positioning accuracy and sensitivity. At the same time, by collecting the positioning position and pen body angle of a large number of written notes, the expression of the writing trajectory is further optimized, reducing the possibility of jagged trajectories. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] In order to more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the specific embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0035] Figure 1 is a schematic structural diagram of a passive capacitive stylus according to an embodiment of the present invention;

[0036] Figure 2 is a schematic structural diagram of a passive capacitive stylus according to an embodiment of the present invention without the capacitive sensing component and the insulating casing;

[0037] Figure 3 This is a schematic structural diagram of a passive capacitive stylus according to an embodiment of the present invention, after omitting the capacitive sensing component and the insulating housing and adding a smooth transition angle;

[0038] Figure 4 is a schematic structural diagram of a passive capacitive stylus according to an embodiment of the present invention without a capacitive sensing component;

[0039] Figure 5 is a structural framework diagram of an electronic writing screen according to an embodiment of the present invention;

[0040] Figure 6 is a schematic structural diagram of a capacitive contactless sensor according to an embodiment of the present invention;

[0041] Figure 7 is a circuit schematic diagram of a capacitive contactless sensor according to an embodiment of the present invention;

[0042] Figure 8 is a positional arrangement diagram of the sensing submodule and the capacitive contactless sensing array unit according to an embodiment of the present invention;

[0043] Figure 9 is a circuit schematic diagram of an induction submodule according to an embodiment of the present invention;

[0044] Figure 10 is a schematic structural diagram of time division multiplexing involving two electronic switches inside a signal generator and a detection module according to an embodiment of the present invention;

[0045] Figure 11 FIG. 4 is a flow chart of a capacitive contactless touch recognition method according to an embodiment of the present invention.

[0046] Reference numerals:

[0047] 1. Pen body; 2. Pen tip; 3. Tail plug; 4. Groove; 5. Capacitive sensing component; 6. Smooth transition angle; 7. Insulating shell. DETAILED DESCRIPTION

[0048] To make the purpose, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without making creative efforts shall fall within the scope of protection of the present invention.

[0049] In the description of the present invention, it should be noted that the terms "upper," "lower," "left," "right," "inner," and "outer," etc., indicating directions or positional relationships, are based on the directions or positional relationships shown in the accompanying drawings and are intended solely to simplify the description of the present invention. They do not indicate or imply that the devices or components referred to must have a specific direction, be constructed, or operate in a specific direction. Therefore, they should not be construed as limitations on the present invention. Furthermore, the terms "first," "second," etc., etc., are used for descriptive purposes only and should not be construed as indicating or implying relative importance.

[0050] Figure 1 FIG. 1 is a schematic diagram showing the structure of a passive capacitive stylus according to an embodiment of the present invention. Figure 1 As shown, the passive capacitive stylus includes: a pen body 1, a pen tip 2 and a tail plug 3.

[0051] Specifically, the pen body 1, the pen tip 2 and the tail plug 3 are all insulating components; the pen tip 2 is connected to one end of the pen body 1; the tail plug 3 is connected to the other end of the pen body 1; the pen body 1 is provided with a groove 4 at one end near the pen tip 2, such as Figure 2 As shown, the capacitive sensing component 5 is accommodated in the groove 4 .

[0052] The material selection for the capacitive sensing component 5 is not limited here, as long as it can cause changes in the edge capacitance and mutual capacitance of the contactless capacitive sensor and has a certain charge capacity, such as cylindrical metal sheets, conductive fibers, conductive cloth, etc. Since the unique capacitance change curve of the contactless capacitive sensor can be generated by either a large or small size of the capacitive sensing component 5, the size of the capacitive sensing component 5 can be freely set according to actual needs and is not specifically limited here.

[0053] It should be noted that the shape and size of the pen holder 1 can be freely set according to actual needs and are not specifically limited here. The embodiment of the present invention only takes a cylindrical pen holder 1 as an example for demonstration.

[0054] Preferably, the tail plug 3 is in the shape of a hemispherical body; the pen tip 2 is in the shape of a combination of a truncated cone and a spherical top.

[0055] In an optional embodiment, the pen body 1 and the pen tip 2 are integrally formed or detachably connected, and the tail plug 3 is integrally formed or detachably connected to the pen body 1.

[0056] In an optional embodiment, as Figure 3 As shown, the groove 4 is provided with a smooth transition corner 6 at the edge of one end close to the tail plug 3 .

[0057] In an optional embodiment, as Figure 1 He Ru Figure 2 As shown, the shape of the groove 4 is a cylindrical structure that is recessed along the circumference of the pen shaft; the shape of the capacitive sensing component 5 is a cylindrical structure that matches the shape of the groove 4; the capacitive sensing component 5 is ring-shaped on the groove 4 of the pen shaft 1.

[0058] In an optional embodiment, as Figure 1 and Figure 4 As shown, the passive capacitive stylus further includes an insulating casing 7 ; the insulating casing 7 is sleeved on at least a portion of the capacitive sensing component 5 .

[0059] In the embodiment of the present invention, the insulating housing 7 is provided to separate the finger from the capacitive sensing component 5 , which not only prevents the finger from interfering with touch recognition, but also reduces the possibility of wear and damage to the capacitive sensing component 5 after long-term use.

[0060] Based on the aforementioned passive capacitive pen, an embodiment of the present invention further provides an electronic writing screen used in conjunction with the passive capacitive pen, such as Figure 5 As shown, it includes: a signal generator, a capacitive contactless sensing array, a detection module, an A / D conversion module, a processor, and a display module.

[0061] Specifically, the signal generator is used to generate a pulse signal; the capacitive contactless sensing array is electrically connected to the signal generator, and the capacitive contactless sensing array is used to detect the position change of the passive capacitive pen after receiving the pulse signal, and to generate a feedback signal according to the position change of the passive capacitive pen; the detection module is electrically connected to the capacitive contactless sensing array, and the detection module is used to receive and analyze the feedback signal, and generate an analog signal containing a resonant frequency offset; the A / D converter is electrically connected to the detection module, and the A / D converter is used to convert the analog signal into a digital signal; the processor is electrically connected to the A / D converter, and the processor is used to receive the digital signal, and to analyze and process the digital signal using a machine learning algorithm, and output the position information and angle information of the passive capacitive pen; the display module is used to display the writing trajectory of the passive capacitive pen in real time according to the position information and angle information of the passive capacitive pen.

[0062] Among them, the signal generator can serve as the RF source of the capacitive contactless sensing array, activating the capacitive contactless sensing array by generating a pulse signal (also called an excitation signal). After being excited, the capacitive contactless sensing array can be used to detect the position change of the non-contact capacitive pen and generate a corresponding feedback signal.

[0063] Before describing the electronic writing screen of an embodiment of the present invention, the working principle of the capacitive contactless sensor array is first described in detail:

[0064] The capacitive contactless sensing array adopts a parallel plate structure, which is composed of multiple capacitive contactless sensors and arranged in a matrix. Each capacitive contactless sensor is composed of a pair of electrode plates and a dielectric layer, such as Figure 6 As shown in the figure, fringe field effect is commonly used for contactless detection of conductive objects. The pulse signal generated by a signal generator applies a certain voltage between two electrode plates, creating a potential difference between them. Most of the electric field lines run parallel between the two plates, but a few spread outward from the high-potential electrode and eventually return to the low-potential electrode. This portion of the electric field that passes through the external space is called the fringe field.

[0065] When a conductor approaches, the fringe field around the two electrode plates changes. The presence of a conductive object changes the fringe field of the capacitive contactless sensor, thereby affecting the measured capacitance value. Specifically, Figure 7 As shown, in the non-contact mode, a mutual capacitance C is generated between the two electrode plates. m, and a fringe capacitance C is generated between the electrode and the adjacent conductor f When the conductor is away from the capacitive contactless sensor, C f Can be ignored; but when the conductors are close, C f will increase significantly.

[0066] According to Gauss's law, the electric flux through any closed surface is proportional to the total charge contained in the surface, that is:

[0067]

[0068] in, The electric field strength vector represents the magnitude and direction of the electric field in space, with units of volts per meter (V / m); is the area element vector, its direction is perpendicular to the closed surface, its size is the area element, and its unit is square meter (m 2 ); Q is the total charge enclosed by the closed surface, in coulombs (C); ε0 is the dielectric constant of vacuum, which is approximately 8.854×10 -12 F / m (farad per meter).

[0069] Therefore, when the conductor approaches, the electric field generated by the capacitive contactless sensor is shunted and weakened, resulting in a decrease in the amount of charge between the two electrode plates.

[0070] According to the definition of capacitance:

[0071]

[0072] Where C is the capacitance, which represents the ability of a capacitor to store charge, and is measured in farads (F); Q is the charge on a single electrode plate, and is measured in coulombs (C); and V is the potential difference (i.e., voltage) between the two electrode plates, and is measured in volts (V).

[0073] When the potential difference between the two electrode plates remains unchanged, the amount of charge on the electrode plates decreases, so that the total capacitance C is measured. total Reduce, C total＝ C m +C f .

[0074] Thus, by measuring the total capacitance C total The change in the distance or position relationship between the conductor and the capacitive contactless sensor can be further detected.

[0075] Preferably, the dielectric in capacitive contactless sensors can be made of flexible substrate materials commonly used in flexible proximity sensors, such as PDMS (polydimethylsiloxane) and PET (polyethylene terephthalate). For conformal and stretchable sensor substrates, PDMS is the preferred choice due to its inherent high tensile strength, excellent mechanical robustness, flexibility, and elasticity.

[0076] At the same time, since the charge capacity of the sensing component is a unique property of each material, different materials will produce different change curves of the capacitive sensing array unit when they are close to the capacitive sensing array unit.

[0077] The detection module can analyze the high-frequency carrier signal containing the resonant frequency offset information (ie, the feedback signal output by the capacitive contactless sensing array), detect its resonant offset, and generate an analog signal corresponding to the resonant offset.

[0078] It should be noted that the detection module can use a vector network analyzer to measure impedance or transmission coefficient, thereby indirectly determining its resonance offset. The detection module can also be integrated into a processor, such as the STM32, where Fourier transform of the signal can be performed to obtain the signal power at each frequency, thereby detecting its offset. The embodiments of the present invention do not require the specific implementation process of the detection module; any method that can achieve the purpose of detecting the resonance offset is acceptable and is not specifically limited here.

[0079] The function of the A / D converter is to convert analog signals into digital signals so that subsequent processors can recognize and process these digital signals.

[0080] The processor may include a programmable logic control component (such as a PLC or CPU), a memory, and electronic components connected to the editable logic control component, etc., which are well known to those skilled in the art and will not be described in detail here.

[0081] The display module may be configured as a display panel in the form of a liquid crystal display, an organic light emitting diode, etc.

[0082] In an optional embodiment, the capacitive contactless sensing array includes a plurality of sensing submodules, such as Figure 8 As shown in the figure, the black dotted box represents a sensing submodule, and multiple sensing submodules are arranged in a matrix form. Each sensing submodule is electrically connected to the detection module and the signal generator respectively; each sensing submodule includes multiple capacitive contactless sensing array units, such as any one of f1, f2, f3, and f4. Each capacitive contactless sensing array unit includes a capacitor and an inductor, such as Figure 9As shown in the figure, the black dotted box represents a capacitive contactless sensing array unit. The four capacitive contactless sensing array units are all integrated in the microstrip line to form a sensing sub-module. One end of the capacitor and one end of the inductor are connected in series, the other end of the capacitor is grounded, and the other end of the inductor is electrically connected to the detection module and the signal generator through the microstrip line.

[0083] Specifically, such as Figure 9 As shown, port 1 is used as the input port and port 2 as the output port. The output of the signal generator (RF source) is connected to port 1, and the high-frequency carrier signal emitted by the RF source serves as the signal excitation source for each sensing submodule. Each sensing submodule is composed of multiple capacitive contactless sensing array units. Each capacitive contactless sensing array unit forms an independent resonant frequency channel with the RF source, so that the capacitance change of each channel on the sensing submodule can be detected through the resonant frequency offset, thereby converting the physical quantity (such as proximity distance) into a measurable electrical signal. Port 2 is connected to the detection circuit, and the feedback signal obtained from each channel is input into the detection circuit to analyze its frequency offset.

[0084] Each capacitive contactless sensing array unit includes a capacitor and an inductor. The capacitor and the inductor are connected in series to form an LC resonant circuit. The resonant frequency is calculated as follows:

[0085]

[0086] Where L is the inductance of the inductor, C is the capacitance of the capacitor, and fc is the resonant frequency.

[0087] When the inductance value L is fixed, the resonant frequency depends only on the capacitance value C. Multiple capacitive contactless sensing array units with different inductance values can operate at different resonant frequencies. For example, f1, f2, f3, and f4 can be tuned to four different resonant frequencies, respectively, so that the capacitive contactless sensing array can operate at different resonant frequencies. Multiple capacitive contactless sensing array units can be further integrated into a microstrip line to form an inductive submodule. Thus, multiple inductive submodules can form an array to work, and different inductive submodules are tuned to different resonant frequencies. Therefore, when the capacitance value of a certain inductive submodule changes, the corresponding resonant frequency will shift accordingly. Sensing can be performed by detecting the change in the resonant frequency, and positioning can be performed by changing the feedback signal of the resonant frequency.

[0088] The embodiments of the present invention further illustrate the differences between microstrip lines and ordinary conductors, mainly including the following two aspects:

[0089] First, in terms of high-frequency characteristics: ordinary wires only need to conduct current at low frequencies, while microstrip lines need to maintain signal integrity at high frequencies (such as impedance matching and reducing radiation loss).

[0090] Second, electromagnetic field control: the electromagnetic field of ordinary wires is easily affected by external interference, while microstrip lines are shielded by ground plates to enhance signal stability.

[0091] It should be noted that the present invention only uses one of the sensing submodules as an example for description. The other sensing submodules are similar and will not be described in detail here. The capacitors in the capacitive contactless sensing array unit of the present invention are all contactless capacitive sensors. For simplicity of description, they are referred to as capacitors.

[0092] In an optional embodiment, each capacitive contactless sensing array unit in the sensing submodule has a different resonant frequency. The signal generator and the electronic switch inside the detection module are both electrically connected to each sensing submodule in the capacitive contactless sensing array using a time division multiplexing connection method.

[0093] Specifically, such as Figure 10 As shown, the two electronic switches inside the signal generator and the detection module perform time division multiplexing according to regular timing changes, dividing the time into multiple time slots, so that the signal generator and detection circuit of each time slot are connected to different sensing sub-modules. For example, when it is necessary to connect the sensing sub-module 1, the two electronic switches are connected to the circuit of the sensing sub-module 1 at the same time. At this time, the high-frequency carrier signal generated by the signal generator can transmit the information of the sensing sub-module 1 to the detection module for comparison. Repeating the above steps can select any one sensing sub-module in the capacitive contactless sensing array, and at the same time, the coupling and crosstalk between the remaining sensing sub-modules can be shielded.

[0094] Furthermore, for each sensing submodule, four groups of capacitors and inductors are combined to form capacitive contactless sensing array units with four different resonant frequencies (f1, f2, f3, and f4). The resonant peaks of each channel are separated in the frequency domain, so changes in the corresponding capacitance only affect the transmission of the corresponding stopband and do not interfere with other frequency bands. Spectral isolation prevents multi-channel signal aliasing, similar to the "guard band" function of frequency division multiplexing, providing anti-interference capabilities. Furthermore, all signals of different frequency bands within the same sensing submodule can be transmitted and processed simultaneously, without the need for time slot switching, allowing for parallel transmission.

[0095] The stopband indicates that each capacitive contactless sensing array unit (an LC resonant circuit consisting of a capacitor and an inductor) has a specific resonant frequency. The resonant frequency of the unit corresponds to the center frequency of the stopband, and the signal will be blocked near this frequency (i.e., the stopband). Therefore, the resonant frequency can be inferred by determining the frequency of the stopband, and then the capacitance change value can be obtained.

[0096] For all the sensing sub-modules in the capacitive contactless sensing array, time division multiplexing can be used to select and read any one of the sensing sub-modules, while frequency division multiplexing can significantly improve the reading efficiency. The embodiments of the present invention combine time division multiplexing and frequency band isolation technology to reduce the mutual interference between different capacitive contactless sensing array units of the same sensing sub-module. Spectrum resources are fully utilized in both time and frequency dimensions, reducing the waste of idle frequency bands or time slots.

[0097] like Figure 11 As shown, an embodiment of the present invention further provides a capacitive contactless touch recognition method applied to an electronic writing screen, comprising:

[0098] S1, obtaining the digital signal output by the A / D converter corresponding to when the passive capacitive pen writes on the electronic writing screen;

[0099] S2. Input the digital signal into a pre-trained CNN-LSTM hybrid network to obtain the position information of the passive capacitive stylus; the CNN-LSTM hybrid network includes a CNN network and an LSTM network cascaded with the CNN network.

[0100] Specifically, the machine learning algorithm can use a CNN-LSTM hybrid network (convolutional neural network-long short-term memory network hybrid network) to predict the trajectory of the passive capacitive stylus and output the position and angle information of the passive capacitive stylus. The trained machine learning model is deployed in the recognition module of the processor.

[0101] It should be noted that in addition to using the CNN-LSTM hybrid network, the machine learning algorithm can also use decision trees, random forests, support vector machines, logistic regression, neural networks, etc., as long as it is a training model for images or time, and there is no specific limitation here.

[0102] In an optional embodiment, the training process of the CNN-LSTM hybrid network is:

[0103] Step a, obtaining a first sample data set; the first sample data set includes multiple first sample data subsets; the target first sample data subset includes the frequency offset value and label value of the target passive capacitive pen at the target position and target angle at the target moment; the label value is the actual landing point coordinates and the actual pen shaft tilt angle of the target passive capacitive pen at the target position and target angle at the target moment; the target moment includes the current moment and any moment before the current moment.

[0104] The first sample data set includes frequency offsets caused by various angles corresponding to any position on the display module by different passive capacitive styli at different times as input values for the CNN network. Furthermore, the actual landing point coordinates and actual pen holder tilt angles of different passive capacitive styli at various angles corresponding to any position on the display module at different times are used as labels. The frequency offsets, actual landing point coordinates, and actual pen holder tilt angles caused by the same passive capacitive styli at the same angle corresponding to the same position on the display module at the same time are input together into the CNN network for training.

[0105] Specifically, the raw data of the first sample data set comes from the A / D converter in the electronic writing screen. The A / D converter can convert the analog signal output by any sensing submodule into a digital signal that can be recognized by the processor. Each first sample data subset includes the following information:

[0106] The coordinates of the sensor module correspond to the row (X-axis) and column (Y-axis) positions of the capacitive contactless sensor array.

[0107] Frequency offset (or voltage / capacitance): refers to the change in the LC resonant stopband frequency caused by the electric field disturbance when a passive capacitive pen touches or approaches the electronic writing screen. This can be determined by comparing the electrical signal images when the passive capacitive pen is in contact with or close to the screen. The detection method for reflecting the frequency offset can be set according to actual conditions. For example, the frequency f is used as the x-axis, the impedance or transmission coefficient measured by a vector network analyzer is used as the y-axis, or the signal power at each frequency obtained by Fourier transforming the induced signal in the STM32 is used as the y-axis, etc., and no specific restrictions are given here.

[0108] Pen tip real coordinates (label): The actual landing point coordinates (x, y) of the pen tip and the inclination angles θ and φ of the pen shaft recorded by a high-precision optical positioning system (such as a laser tracker).

[0109] Step b: preprocessing the target first sample data subset; the preprocessing includes normalization, filtering and imaging.

[0110] Normalization: Map the original frequency offset value to the range of 0-255. The formula is:

[0111]

[0112] Filtering and denoising: Use sliding average filtering or Gaussian filtering to suppress high-frequency noise (such as environmental electromagnetic interference).

[0113] Image generation: First, a grid layout is performed, arranging the capacitive contactless sensing array in a grid. Each detection submodule corresponds to a pixel in the grid image. Next, a grayscale image is generated. The grayscale value of each pixel is determined by the normalized frequency offset of the corresponding sensing submodule. Highlight areas are characterized by large frequency offsets (tip contact or proximity) and grayscale values close to 255 (white). Low frequency offsets (no contact or long distance) and grayscale values close to 0 (black) are characterized by low light areas.

[0114] Step c: input the preprocessed target first sample data subset into the CNN network to obtain the predicted landing point coordinates and the predicted pen tilt angle of the target passive capacitive pen at the target position and target angle at the target time.

[0115] Step d: When the predicted landing point coordinates and the predicted pen tilt angle of the target passive capacitive pen at the target position and target angle at the target moment meet the first expected value, a pre-trained CNN network is obtained.

[0116] Specifically, the mean square error (MSE), mean absolute error (MAE) or positioning accuracy can be used as the evaluation index of the CNN network. When the output result meets the expected value of any one of the indicators, the training is completed and a trained CNN network is obtained.

[0117] The mean square error (MSE) is the mean of the squared differences between the predicted coordinates (x', y') and the true coordinates (x, y); the mean absolute error (MAE) is the average absolute deviation of the coordinate prediction; and the positioning accuracy is the percentage of predicted coordinates that fall within the set error threshold (e.g., ±1 mm).

[0118] Step e: Filter out the predicted landing point coordinates and predicted pen shaft tilt angles of multiple passive capacitive pens on different continuous writing trajectories from the first sample data set to obtain a second sample data set; the second sample data set includes multiple second sample data subsets; the target second sample data subset includes the predicted landing point coordinates and predicted pen shaft tilt angle of the target passive capacitive pen on the target writing trajectory at the moment before the current moment.

[0119] Specifically, the second sample data set is screened out based on the first sample data set. The pen tip landing position, pen shaft angle, and writing time of different non-contact capacitive pens in their respective continuous writing trajectories (regardless of the length of the trajectories, including points, lines, etc.) are used as the second sample data set. The three data before a certain time node A are used as input, and the position of time node A at that time is used as output, so as to optimize the writing trajectory.

[0120] Step f: input the target second sample data subset into the LSTM network to obtain the predicted landing point coordinates and the predicted pen tilt angle of the target passive capacitive pen on the target writing trajectory at the current moment.

[0121] Step g: When the predicted landing point coordinates and the predicted pen shaft tilt angle of the target passive capacitive pen on the target writing trajectory at the current moment meet the second expected value, a pre-trained LSTM network is obtained.

[0122] Specifically, trajectory smoothness can be used as an evaluation indicator for the LSTM network. When the output result meets the expected value set in the indicator, the training ends and a trained LSTM network is obtained.

[0123] The smoothness of the trajectory is calculated by calculating the curvature change or the second-order derivative of the predicted trajectory to quantify the degree of reduction of the aliasing phenomenon.

[0124] It should be noted that the total update delay when the signal output from the capacitive contactless sensor array is input to the display module must meet the real-time requirements.

[0125] Step h: cascade the pre-trained CNN network and the pre-trained LSTM network to obtain a pre-trained CNN-LSTM hybrid network.

[0126] Finally, the pre-trained CNN-LSTM hybrid network is deployed in the processor, and the logical relationship between the processor's output signal and the display module is set.

[0127] In addition, the division and verification of the data set are explained:

[0128] The first sample data set is divided into training set, validation set, and test set in a ratio of 6:2:2 to ensure uniform distribution of different angles and positions.

[0129] Second sample dataset (trajectory model): Divide by temporal continuity, retaining complete trajectory segments to avoid truncation and leakage of temporal information. For example, the first 60% of time steps are used as the training set, the middle 20% as the validation set, and the last 20% as the test set.

[0130] K-fold cross validation (eg, K=5) is used on the first sample data set to verify the generalization ability of the model for position and angle.

[0131] Time Series Split is used for the first sample data set to ensure that the training set is always before the test set.

[0132] Since the CNN algorithm is sensitive to spatial image recognition and extracts the spatial features of the image (such as the shape of the pen tip and the gradient distribution of the contact area), the embodiment of the present invention uses the CNN algorithm to perform machine learning on the images of the first sample data set to train a positioning model. Secondly, the LSTM model is trained using the second sample data set to learn people's daily writing habits, optimize the trajectory expression, and reduce the possibility of jagged trajectories. By using a CNN-LSTM hybrid network, the landing point of the capacitive pen is located in space and the writing trajectory is optimized in time, which greatly reduces the error caused by external interference to the writing device. After machine learning, the processor is more sensitive to specific conductor materials and has better anti-interference ability to the capacitance changes caused by irrelevant conductors (such as fingers).

[0133] Although the embodiments of the present invention have been described with reference to the accompanying drawings, those skilled in the art may make various modifications and variations without departing from the spirit and scope of the present invention. Such modifications and variations are all within the scope defined by the appended claims.

Claims

1. A passive capacitive stylus, characterized in that: include: A pen body, a pen tip and a tail plug; the pen body, pen tip and tail plug are all insulating components; the pen tip is connected to one end of the pen body; the tail plug is connected to the other end of the pen body; the pen body is provided with a groove at one end close to the pen tip, and the groove accommodates a capacitive sensing component.

2. The passive capacitive stylus according to claim 1, characterized in that: The pen holder and the pen tip are integrally formed or detachably connected, and the tail plug is integrally formed or detachably connected to the pen holder.

3. The passive capacitive stylus according to claim 1, wherein: The groove is provided with a smooth transition angle at an edge of one end close to the tail plug.

4. The passive capacitive stylus according to claim 1, wherein: The shape of the groove is a cylindrical structure that is recessed along the circumference of the pen shaft; the shape of the capacitive sensing component is a cylindrical structure that matches the shape of the groove; the capacitive sensing component is sleeved on the groove of the pen shaft.

5. The passive capacitive stylus according to claim 4, characterized in that: It also includes an insulating shell; the insulating shell is sleeved on at least a portion of the capacitive sensing component.

6. An electronic writing screen used in conjunction with the passive capacitive stylus according to any one of claims 1 to 5, characterized in that: include: A signal generator, used to generate a pulse signal; A capacitive contactless sensing array is electrically connected to the signal generator and is used to detect a position change of the non-contact capacitive stylus after receiving the pulse signal, and to generate a feedback signal according to the position change of the non-contact capacitive stylus; a detection module, electrically connected to the capacitive contactless sensing array, configured to receive and analyze the feedback signal and generate an analog signal including a resonant frequency offset; An A / D converter, electrically connected to the detection module, for converting the analog signal into a digital signal; a processor, electrically connected to the A / D converter, configured to receive the digital signal, analyze and process the digital signal using a machine learning algorithm, and output position information and angle information of the passive capacitive stylus; The display module is used to display the writing track of the passive capacitive pen in real time according to the position information and angle information of the passive capacitive pen.

7. The electronic writing screen according to claim 6, characterized in that: The capacitive contactless sensing array includes multiple sensing submodules; each sensing submodule is electrically connected to a detection module and a signal generator respectively; each sensing submodule includes multiple capacitive contactless sensing array units, each capacitive contactless sensing array unit includes a capacitor and an inductor, one end of the capacitor and one end of the inductor are connected in series, the other end of the capacitor is grounded, and the other end of the inductor is electrically connected to the detection module and the signal generator respectively through a microstrip line.

8. The electronic writing screen according to claim 7, characterized in that: The resonant frequency of each capacitive contactless sensing array unit in the sensing submodule is different; The signal generator and the electronic switch inside the detection module are both electrically connected to each sensing submodule in the capacitive contactless sensing array in a time-division multiplexing connection mode.

9. A capacitive contactless touch recognition method, applied to the electronic writing screen according to any one of claims 6 to 8, characterized in that: include: Obtaining the digital signal output by the A / D converter corresponding to when the passive capacitive pen writes on the electronic writing screen; The digital signal is input into a pre-trained CNN-LSTM hybrid network to obtain position information and angle information of the passive capacitive stylus; the CNN-LSTM hybrid network includes a CNN network and an LSTM network cascaded with the CNN network.

10. The method according to claim 9, characterized in that The pre-training process of the CNN-LSTM hybrid network is: Acquire a first sample data set; the first sample data set includes multiple first sample data subsets; the target first sample data subset includes a frequency offset value and a label value of a target passive capacitive stylus at a target position and a target angle at a target time; the label value is the actual landing point coordinates and the actual pen holder tilt angle of the target passive capacitive stylus at the target position and the target angle at the target time; the target time includes the current time and any time before the current time; Preprocessing the target first sample data subset; the preprocessing includes normalization, filtering and imaging; Input the preprocessed target first sample data subset into the CNN network to obtain the predicted landing point coordinates and the predicted pen tilt angle of the target passive capacitive stylus at the target position and target angle at the target time; When the predicted landing point coordinates and the predicted pen tilt angle of the target passive capacitive stylus at the target position and target angle at the target moment meet the first expected value, a pre-trained CNN network is obtained; Filtering the predicted landing point coordinates and predicted pen shaft tilt angles of multiple passive capacitive pens on different continuous writing trajectories from the first sample data set to obtain a second sample data set; The second sample data set includes a plurality of second sample data subsets; The target second sample data subset includes the predicted landing point coordinates and the predicted pen shaft tilt angle of the target passive capacitive pen on the target writing trajectory at a time before the current time; Input the target second sample data subset into the LSTM network to obtain the predicted landing point coordinates and predicted pen tilt angle of the target passive capacitive pen on the target writing trajectory at the current moment; When the predicted landing point coordinates and the predicted pen tilt angle of the target passive capacitive pen on the target writing trajectory at the current moment meet the second expected value, a pre-trained LSTM network is obtained; The pre-trained CNN network and the pre-trained LSTM network are cascaded to obtain a pre-trained CNN-LSTM hybrid network.

Images