A multi-channel body surface micro-gesture interaction system and method based on conductive fabric

Abstract

This invention discloses a multi-channel micro-gesture interaction system and method based on conductive fabric, belonging to the field of human-computer interaction and flexible sensing signal processing technology. The system includes several conductive fabric channels and an interaction control module. The interaction control module is equipped with several acquisition interfaces corresponding to the conductive fabric channels. During the working phase, the interaction control module acquires several real-time resistance data corresponding to the conductive fabric channels and mutual resistance state data between any two conductive fabric channels based on the acquisition interfaces. Based on preset baseline resistance data, several real-time resistance data, mutual resistance state data, and preset multi-layer gesture parsing logic, it outputs multi-channel parsing data and obtains gesture interaction control commands based on the multi-channel parsing data. The system then controls the controlled device to perform the current interaction action based on the gesture interaction control commands, achieving wearable interaction with low complexity, low computing power, and low cost.

Description

Technical Field

[0001] This invention relates to the field of human-computer interaction and flexible sensing signal processing technology, and in particular to a multi-channel micro-gesture interaction system and method based on conductive fabric. Background Technology

[0002] With the development of flexible electronics and wearable devices, body gesture interaction has become an important research direction in the field of human-computer interaction. Users expect to directly control drones, AR / VR devices, smart home appliances, and other controlled devices in wearable scenarios through natural hand movements (such as pressing, bending, and pinching), without needing a handheld remote control or relying on visual recognition. An ideal body interaction system should meet the following requirements: it should be able to support parallel input of multiple gesture modalities on the same wearable part; the signal parsing process should be simple and reliable; it should be able to run in real time on low-computing-power embedded platforms; and it should not require additional non-woven sensors to maintain the lightweight, low-power, and comfortable nature of the wearable device. However, existing flexible electronics and wearable devices face a common technical challenge when processing multimodal gestures: when simultaneously detecting pressing and bending gestures on the same conductive fabric, the resistance signals output by the sensors often exhibit indistinguishable characteristics. Specifically, most flexible piezoresistive materials show a decreasing resistance trend when subjected to pressure or bending deformation, causing the resistance signals of the two gestures to overlap and become confused with each other. The magnitude of the resistance change alone cannot reliably distinguish whether the user is pressing or bending. Furthermore, the robustness of the amplitude distinction is easily affected by factors such as individual force differences, wearing tightness, and ambient temperature, resulting in blurred judgment boundaries and a high misjudgment rate.

[0003] To overcome this obstacle, existing technologies have mainly developed three types of solutions. The first type is the spatial isolation solution, which involves deploying sensors that respond to only a single modality in different areas of the fabric. When multimodal input is required, multiple sets of dedicated sensing areas must be deployed, leading to a linear increase in system complexity, reduced channel utilization, and increased complexity in wearable wiring. The second type is the multi-sensor fusion solution, which uses heterogeneous devices such as inertial measurement units (IMUs), pressure films, and capacitance matrices superimposed on the fabric to assist in recognizing gesture modalities. Although this solves the differentiation problem to some extent, it significantly increases the system integration complexity, the feeling of foreign objects in the wearer, and power consumption, thus losing the lightweight advantage of purely fabric-based wearables. The third category is aliasing signal classification schemes based on machine learning. These schemes use deep learning or traditional machine learning algorithms to regress the gesture category from time-domain or frequency-domain features. This approach relies on a large amount of labeled data for training, and the model is sensitive to individual differences (retraining or transfer learning is usually required when changing users). The inference process requires high computing power, making it difficult to run in real time on ultra-low-power microcontrollers. Furthermore, the model's decision-making process is a black box, lacking interpretability, which limits its application in safety-critical scenarios (such as drone control and medical assistance). However, all of the above schemes sacrifice modal discrimination capabilities by increasing hardware redundancy or computational complexity, resulting in high system costs and power consumption, and affecting the wearability and real-time performance of the device. Summary of the Invention

[0004] This invention provides a multi-channel micro-gesture interaction system and method based on conductive fabric, which can solve the technical problem of signal confusion caused by the same-direction response of pressing and bending in the prior art, and the need to rely on multi-sensor fusion or complex algorithms to distinguish gestures, and realize wearable interaction with low complexity, low computing power and low cost.

[0005] This invention provides a multi-channel micro-gesture interaction system based on conductive fabric, comprising several conductive fabric channels and an interaction control module, wherein: The interactive control module is provided with several acquisition interfaces corresponding to several of the conductive fabric channels. For any conductive fabric channel and the corresponding acquisition interface of the conductive fabric channel, the conductive fabric channel includes conductive fabric, a first electrode and a second electrode. The first end of the first electrode is electrically connected to the first end of the conductive fabric, the first end of the second electrode is electrically connected to the second end of the conductive fabric, the second end of the first electrode is electrically connected to the first acquisition end of the acquisition interface, and the second end of the second electrode is electrically connected to the second acquisition end of the acquisition interface. The interactive control module is used to acquire, during the working phase, several real-time resistance data corresponding to several conductive fabric channels and mutual resistance status data between any two conductive fabric channels based on several acquisition interfaces. Based on preset baseline resistance data, several real-time resistance data, mutual resistance status data and preset multi-layer gesture parsing logic, it outputs multi-channel parsing data and acquires gesture interaction control commands based on the multi-channel parsing data, so as to control the controlled device to perform the current interactive action based on the gesture interaction control commands.

[0006] The above solution sets up several conductive fabric channels and acquisition interfaces on the interactive control module. Each channel consists of conductive fabric and electrodes at both ends. During the working phase, it can simultaneously acquire real-time resistance data of each channel and mutual resistance data between any two channels. Based on a preset baseline resistance and multi-layer gesture parsing logic, it outputs multi-channel parsing data to generate control commands. By connecting the conductive fabric and modular acquisition with electrodes, it can acquire resistance signals of various gestures such as pressing, bending, and pinching in parallel without the need for additional inertial sensors or machine learning models. This enables the differentiation of different gesture signals and solves the technical problem of signal confusion caused by the same-direction response of pressing and bending in existing technologies, which requires the reliance on multi-sensor fusion or complex algorithms to distinguish gestures. This achieves wearable interaction with low complexity, low computing power, and low cost.

[0007] Furthermore, during the calibration phase, the interactive control module is also used to acquire several baseline resistances corresponding to several conductive fabric channels based on several acquisition interfaces, and to use several baseline resistances as the preset baseline resistance data.

[0008] The above solution automatically collects and stores the baseline resistance of each channel during the calibration phase, eliminating the influence of individual user differences (such as finger strength, joint flexibility, etc.) and wearing tightness on subsequent gesture judgment, making the system adaptive to different users and different wearing states.

[0009] This invention provides a multi-channel micro-gesture interaction system based on conductive fabric. Several conductive fabric channels are electrically connected to the acquisition interface of the interaction control module. Each channel consists of conductive fabric with positive resistance when pressed and negative resistance when bent, along with electrodes at both ends. The module acquires the baseline resistance of each channel during the calibration phase and, during operation, acquires the real-time resistance of each channel and the mutual resistance between any two channels in parallel. Combined with a preset multi-layer gesture parsing logic, it outputs multi-channel parsing data and generates control commands. Without the need for additional inertial sensors, capacitor matrices, or other heterogeneous devices, and without relying on machine learning models, it can simultaneously recognize three gesture modalities—press, bend, and pinch—on the same fabric using only conductive fabric electrodes and simple resistance signals. This achieves a low-complexity, low-computing-power, and low-cost wearable interactive hardware platform, solving the technical problems of high system integration complexity, strong wearing discomfort, high power consumption, and limited real-time performance caused by existing systems that are forced to adopt spatial isolation, multi-sensor fusion, or machine learning solutions due to gesture signal confusion.

[0010] This invention also provides a multi-channel micro-gesture interaction method based on conductive fabric, applied to the aforementioned multi-channel micro-gesture interaction system based on conductive fabric, which includes several conductive fabric channels and an interaction control module; the method uses the interaction control module as the execution entity, including: During the work phase: Acquire several real-time resistance data corresponding to several conductive fabric channels and mutual resistance status data between any two conductive fabric channels. Based on preset baseline resistance data, several real-time resistance data, mutual resistance status data, and preset multi-layer gesture parsing logic, multi-channel parsing data is output. Gesture interaction control commands are obtained based on the multi-channel parsing data, and the controlled device is controlled to perform the current interaction action based on the gesture interaction control commands.

[0011] The above solution can complete the conversion from resistance signal to control command in real time on a low-computing-power microcontroller by acquiring the real-time resistance and mutual resistance status of each channel and combining the preset baseline resistance and multi-layer parsing logic. It does not require complex mathematical calculations or external auxiliary sensors, and can distinguish different gesture signals. This solves the technical problem of existing technologies where the same-direction response of pressing and bending leads to signal confusion, and the need to rely on multi-sensor fusion or complex algorithms to distinguish gestures.

[0012] Furthermore, it also includes: during the calibration phase, acquiring several baseline resistances corresponding to several of the conductive fabric channels respectively, and using the several baseline resistances as the preset baseline resistance data.

[0013] The above solution automatically collects and stores the baseline resistance of each channel during the calibration phase, eliminating the influence of individual user differences and wearing tightness on subsequent gesture decisions, and improving adaptability to different users and different wearing conditions.

[0014] Furthermore, the preset multi-layer gesture parsing logic includes modal polarity judgment, event trigger judgment, and pinch judgment. The output of multi-channel parsing data based on preset baseline resistance data, several sets of real-time resistance data, mutual resistance state data, and the preset multi-layer gesture parsing logic includes: The real-time resistance and baseline resistance corresponding to any of the conductive fabric channels are obtained from the preset baseline resistance data and several of the real-time resistance data. Based on the real-time resistance and the baseline resistance, modal polarity determination and event trigger determination are performed respectively to obtain the trigger event result corresponding to any of the conductive fabric channels; The mutual resistance state between any two conductive fabric channels is obtained from the mutual resistance state data, and the kneading judgment is performed based on the mutual resistance state to obtain the kneading event judgment result between any two conductive fabric channels. Obtain the results of several trigger events corresponding to all the conductive fabric channels and the results of several kneading events between all the conductive fabric channels; The results of the triggering event and the judgment result of the pinching event are output as multi-channel parsed data.

[0015] The above solution processes the parsing logic in layers and independently, and uses polarity to separate pressing and bending, and mutual resistance change to detect pinching to achieve unambiguous parallel recognition output of various gestures. This solves the technical problem that multimodal signals on the same conductive fabric channel cannot be distinguished, and that existing technologies can only separate gestures through spatial isolation or machine learning.

[0016] Further, the step of determining the modal polarity and event triggering based on the real-time resistance and baseline resistance to obtain the triggering event result corresponding to any of the conductive fabric channels includes: The relative rate of change of resistance is calculated based on the real-time resistance and the baseline resistance. The current candidate event is obtained by determining the modal polarity based on the sign of the relative rate of change of resistance. The valid event judgment result is obtained by judging the event trigger based on the absolute value of the relative rate of change of resistance, the preset trigger threshold and the current candidate event; The trigger event result is obtained based on the current candidate event and the valid event judgment result.

[0017] The above scheme decouples modality discrimination from event triggering, distinguishing modes solely by the direction of resistance change, and thus determining what type of event should be triggered.

[0018] Furthermore, the process of determining the modal polarity based on the sign of the relative rate of change of resistance to obtain the current candidate event includes: Modal polarity is determined based on the sign of the relative rate of change of resistance. When the sign meets the preset polarity judgment condition, the current candidate event is determined to be a pressing event. When the sign does not meet the preset polarity judgment condition, the current candidate event is determined to be a bending event.

[0019] The above solution utilizes the physical properties of conductive fabrics, which exhibit positive resistance when pressed and negative resistance when bent. It achieves modal separation by using the positive and negative signs of the rate of change, without requiring any auxiliary sensors or machine learning training. This solves the fundamental technical problem that traditional unidirectional responsive fabrics cannot distinguish between pressing and bending by using signs.

[0020] Furthermore, the preset trigger threshold includes a preset press trigger threshold, and the process of obtaining a valid event judgment result based on the absolute value of the relative change rate of resistance, the preset trigger threshold, and the current candidate event includes: When the current candidate event is determined to be a press event: The event triggering judgment is based on the absolute value of the relative change rate of resistance and the preset press triggering threshold. When the absolute value is greater than the preset press triggering threshold, the duration acquisition action is triggered to obtain the press duration. When the duration of the press exceeds a preset time threshold, the result of the valid event judgment is output as a valid trigger event.

[0021] The above solution introduces dual judgment conditions of amplitude and time for the discrete command channel corresponding to the press event. It filters out brief touch or bounce signals based on insufficient duration, thus solving the technical problem of unintentional light touch or signal jitter causing the discrete command to be falsely triggered during the press action.

[0022] Furthermore, the preset trigger threshold includes a preset bending trigger threshold, and the process of obtaining a valid event judgment result by judging the event trigger based on the absolute value of the relative rate of change of resistance, the preset trigger threshold, and the current candidate event further includes: When the current candidate event is determined to be a bending event, the event triggering judgment is performed based on the absolute value of the relative change rate of resistance and the preset bending triggering threshold. When the absolute value is greater than the preset bending triggering threshold, the valid event judgment result is output as a valid triggering event. When the absolute value is not greater than the preset bending triggering threshold, the valid event judgment result is output as an invalid triggering event.

[0023] The above solution uses a single amplitude condition for the continuous control channel corresponding to bending events, without the need for a time threshold to ensure real-time response. At the same time, it uses a high preset bending trigger threshold to filter out hand tremors or unintentional slight bending, thus solving the technical problem of frequent false triggering of continuous control commands due to unintentional bending in daily activities.

[0024] Furthermore, after obtaining gesture interaction control instructions based on the multi-channel parsed data, and controlling the controlled device to perform the current interaction action based on the gesture interaction control instructions, the method further includes: Real-time acquisition of the current relative rate of change of resistance; When the preset termination condition is met based on the current relative rate of change of resistance, the preset trigger threshold, and the preset fallback ratio, the event is determined to end, and a stop interaction control command is output to control the controlled device to stop executing the current interaction action.

[0025] After generating control commands to control the controlled device, the above solution introduces a hysteresis anti-jitter mechanism. After an event is triggered, the absolute value of the rate of change needs to fall back to a certain percentage below the trigger threshold before the event can be considered over. This solves the technical problem of repeated triggering or inability to stably terminate commands caused by the signal jittering around the threshold.

[0026] The present invention offers the following advantages: It utilizes the reverse polarity relationship between the positive resistance of the conductive fabric during pressing and the negative resistance during bending, naturally separating modes through the sign of the rate of change. It optimizes the anti-misclick performance of discrete commands and the anti-jitter performance of continuous control by setting separate pressing and bending thresholds. Time gating prevents accidental triggering of discrete commands through brief touches, and a hysteresis mechanism prevents repeated triggering due to signal jitter. This method involves only subtraction, division, and comparison operations, and can run in real-time on low-computing-power microcontrollers. Simultaneously, it fundamentally solves the technical problems of indistinguishable signals caused by the unidirectional pressing and bending responses of mainstream piezoresistive fabrics, and the reliance of existing solutions on amplitude differentiation or machine learning. It achieves unambiguous, low-latency, and highly robust parallel recognition of three gesture modes—pressing, bending, and pinching—on the same conductive fabric channel, and supports multiple interaction paradigms from discrete binary commands to continuous proportional control. It can be widely applied in scenarios such as drones, AR / VR, smart home appliances, and assistive technologies. Attached Figure Description

[0027] To more clearly illustrate the technical solution of this application, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0028] Figure 1This is a schematic diagram of a multi-channel micro-gesture interaction system based on conductive fabric provided in this embodiment; Figure 2 This is a schematic diagram of a voltage divider circuit provided in this embodiment; Figure 3 This is a schematic diagram of a multi-channel micro-gesture interaction method based on conductive fabric provided in this embodiment; Figure 4 This is an experimental characterization diagram of the bending negative piezoresistive response of the conductive fabric provided in this embodiment, wherein, Figure 4 (a) in the figure is the time series curve of the fabric resistance changing with time under multiple consecutive bending-stretching cycles. Figure 4 (b) in the figure shows the detailed curve of the fabric resistance changing with time during a single bending-stretching event. Detailed Implementation

[0029] To make the objectives, technical solutions, and advantages of this application clearer, the technical solutions of this application will be clearly and completely described below with reference to the accompanying drawings of the embodiments. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0030] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains; the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the application; the terms “comprising” and “having”, and any variations thereof, in the specification, claims, and foregoing description of the drawings are intended to cover non-exclusive inclusion.

[0031] In the description of the embodiments of this application, technical terms such as "first" and "second" are used only to distinguish different objects and should not be construed as indicating or implying relative importance or implicitly specifying the number, specific order, or primary and secondary relationship of the indicated technical features. In the description of the embodiments of this application, "multiple" means two or more, unless otherwise explicitly defined.

[0032] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0033] In the description of the embodiments in this application, the term "and / or" is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, and B existing alone. Additionally, the character " / " in this document generally indicates that the preceding and following related objects have an "or" relationship.

[0034] In the description of the embodiments of this application, the term "multiple" refers to two or more (including two), similarly, "multiple sets" refers to two or more (including two sets), and "multiple pieces" refers to two or more (including two pieces).

[0035] In the description of the embodiments of this application, unless otherwise expressly specified and limited, technical terms such as "installation," "connection," "joining," and "fixing" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral part; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; they can refer to the internal communication of two components or the interaction between two components. For those skilled in the art, the specific meaning of the above terms in the embodiments of this application can be understood according to the specific circumstances.

[0036] Example 1: This embodiment provides a multi-channel body surface micro-gesture interaction system based on conductive fabric, including several conductive fabric channels and an interaction control module, wherein: The interactive control module is provided with several acquisition interfaces corresponding to several of the conductive fabric channels. For any conductive fabric channel and the corresponding acquisition interface of the conductive fabric channel, the conductive fabric channel includes conductive fabric, a first electrode and a second electrode. The first end of the first electrode is electrically connected to the first end of the conductive fabric, the first end of the second electrode is electrically connected to the second end of the conductive fabric, the second end of the first electrode is electrically connected to the first acquisition end of the acquisition interface, and the second end of the second electrode is electrically connected to the second acquisition end of the acquisition interface. The interactive control module is used to acquire, during the working phase, several real-time resistance data corresponding to several conductive fabric channels and mutual resistance status data between any two conductive fabric channels based on several acquisition interfaces. Based on preset baseline resistance data, several real-time resistance data, mutual resistance status data and preset multi-layer gesture parsing logic, it outputs multi-channel parsing data and acquires gesture interaction control commands based on the multi-channel parsing data, so as to control the controlled device to perform the current interactive action based on the gesture interaction control commands.

[0037] In this embodiment, the conductive fabric used is a functional fabric with a surface-deposited conductive composite coating. This conductive fabric exhibits both positive piezoresistive response under pressure and negative piezoresistive response under bending. Specifically, the polarity difference between the positive piezoresistive response under pressure and the negative piezoresistive response under bending stems from the different response mechanisms of the conductive coating's microstructure under the two deformation modes: Under normal compressive load, the fabric substrate undergoes local compression, generating in-plane tensile stress on the coating surface. This leads to the disruption of the overlapping network of conductive fillers (such as carbon nanotubes, graphene, or conductive polymers) within the coating surface, reducing the density of conductive pathways and macroscopically manifesting as increased resistance, i.e., the positive piezoresistive effect. Under bending deformation, the dominant stress on the coating is in-plane compressive stress. The conductive fillers are compressed and aggregated on the outside of the bend, increasing the overlap probability and decreasing the resistance, manifesting as the negative piezoresistive effect. This polarity difference between positive piezoresistive resistance under pressure and negative piezoresistive resistance under bending has been experimentally verified in a functional fabric system with a surface-deposited conductive composite coating (see the experimental verification results in Example 5 for details). Conductive fabrics possessing the aforementioned polarity differences can be prepared by surface deposition of conductive composite coatings onto conventional fiber substrates (such as cotton and polyester fiber fabrics). For example, carbon nanotube / elastomer composite slurries or conductive polymer composite slurries can be deposited onto the fabric surface using dip coating, spraying, or printing. The response polarity to pressing and bending can be controlled by adjusting the coating thickness, filler concentration, and substrate elastic modulus to meet the requirements of positive resistance to pressing and negative resistance to bending. The conductive fabric used in this embodiment was obtained through the above preparation method. In Example 5, its response polarity to pressing and bending was experimentally verified, demonstrating that it possesses the physical basis to support the gesture recognition logic of this scheme.

[0038] As a concrete and feasible source, the conductive fabric used in this embodiment can be a silk-based conductive functional fabric. Its preparation process includes: firstly, repeatedly impregnating and drying the degummed silk fabric with condensed tannin pretreatment using yam extract to form a chemically active interface layer on the silk fiber surface; then, premixing an iron-humic acid-fulvic acid ternary complex colloid (Fe-HA / FA colloid) with a poly(3,4-ethylenedioxythiophene)-polystyrene sulfonic acid (PEDOT:PSS) aqueous dispersion and coating it onto the pretreated silk fabric surface, followed by low-temperature drying. Experimental measurements show that the baseline resistance of this silk-based conductive functional fabric is in the hundreds of kiloohms range and exhibits high stability. The relative resistance change rate when pressed is +9.5% to +15.0% (positive piezoresistive response), and the relative resistance change rate when bent from 0° to 90° is approximately −25.1% (negative piezoresistive response). This forms the basis for the gesture recognition logic based on polarity discrimination in this embodiment.

[0039] In practical implementation, the number of conductive fabric channels can be determined according to the application scenario. The electrode pair, including the first and second electrodes, can take various forms such as copper foil tape, conductive silver paste, metal fasteners, sewn conductive threads, and conductive fabric bonding. The electrode shape is not limited to rectangles; it can be ring-shaped, interdigitated, etc.

[0040] In practical implementation, this embodiment provides a multi-channel micro-gesture interaction system based on conductive fabric. Taking a three-channel wearable interaction system in glove form as an example, it includes three conductive fabric channels, corresponding to three 5×2cm conductive fabric pieces. During application, these three pieces of conductive fabric are respectively attached to the backs of the index, middle, and ring fingers, with the functional coating facing outwards. Through a first electrode, a second electrode, and an interaction control module, three channels—index finger channel, middle finger channel, and ring finger channel—are formed, achieving real-time closed-loop control of the three channels. Each strip has a 1×2cm copper foil electrode attached to both ends, fixed with conductive adhesive, and a flexible wire leads to a data acquisition board on the back of the hand (i.e., the interaction control module in this embodiment). The core of the data acquisition board is an STM32F103 microcontroller (or an equivalent ARM Cortex-M microcontroller), equipped with three 12-bit ADC channels (analog-to-digital converter channels). The controlled device is a drone, and the data acquisition board communicates with the drone's main controller via a Bluetooth module.

[0041] In the specific implementation process, the three conductive fabric channels of this embodiment play different interactive roles. Specifically: the ring finger channel is used to respond to the pressing mode (i.e., the positive resistance response), and in the specific interaction process, it uses a press-to-trigger mode to control discrete commands (such as commands to control the switching of binary states such as takeoff / landing); the middle finger channel is used to respond to the bending mode (negative resistance response), with the natural extension posture of the finger as the calibration baseline. Bending reduces its resistance, and the relative rate of change of resistance is negative (judged as a bending event). When the absolute value of the relative rate of change of resistance exceeds the preset bending trigger threshold, it is mapped to a control command in one direction (such as a forward command). When the finger extends back to the baseline posture, it is considered a release and no reverse command is generated; the index finger channel is used to respond to the bending mode (negative resistance response), with the natural extension posture of the finger as the calibration baseline. Bending maps to a control command in another axis (such as a left turn command). When the finger extends back to the baseline posture, it is considered a release and no reverse command is generated.

[0042] When constructing several conductive fabric channels, N independent conductive regions (N≥1) are set on the conductive fabric with positive pressure piezoresistive response and negative bending piezoresistive response. Each conductive region (i.e., each piece of conductive fabric) is connected to a pair of electrodes (i.e., the first electrode and the second electrode) to form a conductive fabric channel. The positive pressure piezoresistive response refers to the increase in resistance when normal pressure is applied to the conductive fabric. The negative bending piezoresistive response refers to the decrease in resistance when the conductive fabric undergoes bending deformation.

[0043] In this embodiment, the interactive control module includes a resistance-to-voltage conversion circuit and a microcontroller. Three acquisition interfaces are set on the resistance-to-voltage conversion circuit to electrically connect the electrodes at both ends of three conductive fabric pieces, thereby enabling real-time acquisition of changes in the resistance of the conductive fabric and obtaining an analog voltage signal. This analog voltage signal is then transmitted to the analog-to-digital conversion module in the microcontroller to form voltage sampling data that is easily read and processed by the microcontroller. Based on this voltage sampling data, the microcontroller can then deduce several real-time resistance data and mutual resistance status data. This resistance-to-voltage conversion circuit includes, but is not limited to, voltage divider circuits and Wheatstone bridges.

[0044] In one specific implementation, the resistor-to-voltage conversion circuit corresponding to each acquisition interface adopts, as follows: Figure 2 The voltage divider circuit shown consists of a 3.3V constant voltage source (i.e., the microcontroller's power supply voltage VCC) connected in series with a fixed voltage divider resistor Rf (470kΩ) and the corresponding conductive fabric channel R, then grounded. Specifically, the positive terminal of the constant voltage source is connected to one end of the fixed voltage divider resistor, and the other end (serving as the first acquisition segment) is connected to the first electrode of the conductive fabric channel. The second electrode of the conductive fabric channel is grounded (in this embodiment, the second acquisition terminal is used as the ground terminal). The microcontroller's ADC channel is connected at the midpoint between the fixed voltage divider resistor and the conductive fabric channel, measuring the midpoint voltage Vmid in real time. According to the voltage divider relationship, the real-time resistance R of the conductive fabric channel satisfies Vmid = VCC × R / (Rf + R), thus R = Rf × Vmid / (VCC − Vmid). In this embodiment, the fixed voltage divider resistor of 470kΩ is matched with the baseline resistance of the conductive fabric channel, which is in the hundreds of kilohms range. This allows the ADC midpoint voltage to operate in the middle range of the measurement range with higher resolution, thereby ensuring the accuracy of resistance measurement. When the baseline resistance of the conductive fabric is in a different range, the value of the fixed voltage divider resistor can be adjusted accordingly to make it be in the same range as the baseline resistance.

[0045] Optionally, during the calibration phase, the interactive control module is further configured to acquire several baseline resistances corresponding to several conductive fabric channels based on several acquisition interfaces, and use the several baseline resistances as the preset baseline resistance data.

[0046] In the specific implementation process, the calibration phase includes a baseline calibration phase after the multi-channel body surface micro-gesture interaction system based on conductive fabric is started and before entering the working phase. During the baseline calibration phase, the user wearing the multi-channel body surface micro-gesture interaction system based on conductive fabric provided in this embodiment needs to maintain a preset reference posture. Under this reference posture, each conductive fabric channel is continuously sampled and statistically analyzed as the baseline resistance of that channel. That is, several baseline resistances corresponding to several conductive fabric channels are obtained, forming preset baseline resistance data.

[0047] In the specific implementation process, in order to cope with the slow drift of the resistance of conductive fabric during long-term use, this embodiment further introduces an adaptive baseline tracking mechanism during the working phase.

[0048] During prolonged wear, the resistance of the conductive fabric may slowly drift due to changes in body temperature, sweat, fabric mechanical creep, or fluctuations in ambient temperature and humidity. If the baseline resistance from the initial calibration phase remains constant, the drifted idle resistance will cause the calculated relative rate of change of resistance to deviate, potentially leading to a shift in gesture judgment or false triggering. Therefore, this embodiment introduces the adaptive baseline tracking mechanism. This mechanism includes: continuously monitoring the real-time resistance data of each conductive fabric channel during operation; when the absolute value of the relative rate of change of resistance for all channels is lower than a preset idle threshold (which is much smaller than a basic preset trigger threshold) for a continuous period (e.g., 1 second), the system determines that it is currently in an idle state, meaning the user has not performed any conscious gesture. Then, in the idle state, the system slowly updates the baseline resistance of each channel using an exponentially weighted moving average, as shown in the following formula: ; In the formula, This indicates the baseline resistance of each conductive fabric channel after the update. This indicates the baseline resistance of each conductive fabric channel before the update. This represents the real-time resistance data measured at the current sampling moment; α represents the update coefficient, which in this embodiment is a very small positive number (e.g., α = 0.001), ensuring that the baseline resistance undergoes only minor adjustments each time. This update rate is much slower than the event response rate (e.g., the response time for event detection is tens of milliseconds, while complete baseline drift compensation may take tens of seconds to several minutes (depending on the sampling rate and drift amplitude). This much slower update rate ensures that when the user quickly performs pressing, bending, or pinching gestures, the baseline resistance remains almost unchanged, preventing the gesture signal from being mistakenly interpreted as drift and thus failing to trigger the corresponding interactive action.

[0049] During this process, when the interactive control module detects that the absolute value of the relative rate of change of resistance of any conductive fabric channel exceeds the idle threshold, it immediately pauses baseline updates, exits the idle state, and enters the gesture detection working state. Adaptive baseline tracking is only resumed after the signals of all channels stabilize again and remain below the idle threshold for a period of time.

[0050] Through the aforementioned adaptive baseline tracking mechanism, the system can automatically track resistance drift caused by environmental changes or long-term wear, ensuring that the baseline resistance always reflects the true state of hand-free relaxation. This eliminates the need for manual recalibration by the user, significantly improving the system's long-term stability and user experience.

[0051] In the specific implementation process, as the microcontroller deduces several real-time resistance data and cross-resistance status data based on the voltage sampling data, in order to suppress high-frequency noise and smooth the resistance signal, this embodiment performs digital filtering on the raw voltage data after sampling by the microcontroller. A moving average filter is applied to the microcontroller sampling value for each conductive fabric channel, with the filter window length set to 3 to 10 sampling points (5 sampling points in this embodiment). That is, the arithmetic mean of the five most recent consecutive sampling values ​​is calculated as the effective voltage value at the current moment, effectively attenuating random noise and power supply ripple while maintaining the transient response characteristics of the gesture. In addition to the moving average filter, this embodiment also sets a first-order low-pass filter (e.g., a cutoff frequency of 10Hz to filter out power frequency interference and high-frequency jitter) or a median filter (e.g., a window length of 5 to eliminate isolated pulse noise) according to the actual noise environment, significantly improving the signal-to-noise ratio, avoiding false triggering caused by noise, and with minimal computational load, making it suitable for real-time operation on low-power microcontrollers.

[0052] When controlling a controlled device based on gesture interaction control commands, the signal transmission method in this embodiment is not limited to Bluetooth. Other wired or wireless communication methods such as UWB (Ultra-Wideband), WiFi, ZigBee, and wired USB can also be used to adapt to the interface and latency requirements of different controlled devices. The signal processing can be completed on a local microcontroller (such as an STM32F103) for all gesture parsing and gesture interaction control command generation. Alternatively, the original resistance signal (i.e., the analog voltage signal) or the pre-processed feature data (i.e., the several real-time resistance data and mutual resistance state data) can be transmitted to a smartphone, tablet, or edge computing device for further parsing and command mapping. This distributed processing architecture reduces the power consumption and hardware cost of the wearable device while leveraging the stronger computing power of the terminal device to achieve more complex user calibration and data recording functions.

[0053] In one embodiment, such as Figure 1As shown, this embodiment provides a multi-channel micro-gesture interaction system based on conductive fabric, including: a conductive fabric sensing layer comprising several conductive fabric channels, using conductive fabric with positive piezoresistive response to pressure and negative piezoresistive response to bending, with at least one conductive area on the conductive fabric as a sensing unit for sensing resistance changes caused by pressing, bending, and pinching gestures; an electrode array composed of electrode pairs electrically connected to each conductive area for extracting electrical signals; and an interaction control module including a signal acquisition submodule, a gesture parsing submodule, and a command mapping submodule, wherein the signal acquisition submodule includes a resistor- Voltage conversion circuits (such as voltage dividers or Wheatstone bridges) and analog-to-digital converters are used to convert the resistance changes of each conductive fabric channel into digital signals in real time; the gesture parsing submodule is used to run baseline calibration and three-level gesture parsing logic (including modal polarity discrimination, event trigger determination, and pinch detection), and output multi-channel parsing data; the command mapping submodule is used to convert the parsing results into control commands according to the predefined mapping rules of the application scenario, and transmit them to downstream controlled devices via wired or wireless means (such as Bluetooth, WiFi, UWB, ZigBee, or USB) to execute corresponding interactive actions.

[0054] Example 2: This embodiment provides a multi-channel micro-gesture interaction method based on conductive fabric, applied to the aforementioned multi-channel micro-gesture interaction system based on conductive fabric. This system includes several conductive fabric channels and an interaction control module. The method uses the interaction control module as the execution entity, such as... Figure 3 As shown, it includes: During the work phase: S1. Obtain several real-time resistance data corresponding to several conductive fabric channels and mutual resistance status data between any two conductive fabric channels. S2. Output multi-channel parsing data based on preset baseline resistance data, several real-time resistance data, mutual resistance status data, and preset multi-layer gesture parsing logic; S3. Obtain gesture interaction control instructions based on the multi-channel parsing data, and control the controlled device to perform the current interaction action based on the gesture interaction control instructions.

[0055] Optionally, it also includes: during the calibration phase, acquiring several baseline resistances corresponding to several conductive fabric channels respectively, and using several baseline resistances as the preset baseline resistance data.

[0056] Optionally, the preset multi-layer gesture parsing logic includes modal polarity judgment, event trigger judgment, and pinch judgment, and step S2 includes: The real-time resistance and baseline resistance corresponding to any of the conductive fabric channels are obtained from the preset baseline resistance data and several of the real-time resistance data. Based on the real-time resistance and the baseline resistance, modal polarity determination and event trigger determination are performed respectively to obtain the trigger event result corresponding to any of the conductive fabric channels; The mutual resistance state between any two conductive fabric channels is obtained from the mutual resistance state data, and the kneading judgment is performed based on the mutual resistance state to obtain the kneading event judgment result between any two conductive fabric channels. Obtain the results of several trigger events corresponding to all the conductive fabric channels and the results of several kneading events between all the conductive fabric channels; The results of the triggering event and the judgment result of the pinching event are output as multi-channel parsed data.

[0057] In the specific implementation process, when monitoring the mutual resistance state between any two of the conductive fabric channels, the mutual resistance state is normally an open circuit state. When the mutual resistance state changes abruptly from an open circuit state to a low resistance state, it is determined that a pinching event exists, and the pinching event judgment result is output.

[0058] In this embodiment, the pinching event corresponds to a discrete switch input, such as commands like "take off," "select," or "grab." For example, taking the glove-shaped three-channel wearable interactive system provided in Embodiment 1 as an example, when a continuous pinching or contact event is detected in the thumb channel (i.e., the channel formed by attaching an auxiliary conductive fabric piece to the thumb pad and connecting electrodes) and the ring finger channel, and the middle finger channel is determined to have a validly triggered bending event, it is identified as a "take off + forward" gesture. At this time, the corresponding current interactive action performed by the controlled device is: the drone continues to move forward.

[0059] Optionally, the step of determining the modal polarity and event triggering based on the real-time resistance and baseline resistance to obtain the triggering event result corresponding to any of the conductive fabric channels includes: The relative rate of change of resistance is calculated based on the real-time resistance and the baseline resistance. The current candidate event is obtained by determining the modal polarity based on the sign of the relative rate of change of resistance. The valid event judgment result is obtained by judging the event trigger based on the absolute value of the relative rate of change of resistance, the preset trigger threshold and the current candidate event; The trigger event result is obtained based on the current candidate event and the valid event judgment result.

[0060] Optionally, the step of determining the modal polarity based on the sign of the relative rate of change of resistance to obtain the current candidate event includes: Modal polarity is determined based on the sign of the relative rate of change of resistance. When the sign meets the preset polarity judgment condition, the current candidate event is determined to be a pressing event. When the sign does not meet the preset polarity judgment condition, the current candidate event is determined to be a bending event.

[0061] In this embodiment, the pressing event corresponds to the pressing mode described above, and the bending event corresponds to the bending mode described above. In the specific implementation process, when calculating the relative rate of change of resistance R for each channel's real-time resistance, the specific calculation process is as follows: ; in, The relative rate of change of resistance. For real-time resistance, This is the baseline resistance.

[0062] In determining modal polarity, according to The sign of the variable determines the modality type, i.e., when... When the value is greater than 0, the preset polarity judgment condition is met, and the current candidate event is determined to be a press event. When the value is ≤0, it is determined that the preset polarity judgment condition is not met, and the current candidate event is determined to be a bending event.

[0063] Optionally, the preset trigger threshold includes a preset press trigger threshold, and the process of obtaining a valid event judgment result based on the absolute value of the relative change rate of resistance, the preset trigger threshold, and the current candidate event includes: When the current candidate event is determined to be a press event: The event triggering judgment is based on the absolute value of the relative change rate of resistance and the preset press triggering threshold. When the absolute value is greater than the preset press triggering threshold, the duration acquisition action is triggered to obtain the press duration. When the duration of the press exceeds a preset time threshold, the result of the valid event judgment is output as a valid trigger event.

[0064] Building upon polarity determination, this embodiment optionally introduces lightweight temporal features as auxiliary criteria to further improve the recognition accuracy of complex gesture sequences. These lightweight temporal features include, but are not limited to, the rising edge slope and duration of the relative rate of change of resistance. These features do not rely on machine learning models and can be obtained through simple numerical calculations (such as differential or timing calculations). They can be logically combined with the polarity determination results (e.g., "pressing with a rising edge slope greater than a certain threshold" is determined as a rapid tap, "pressing with a duration greater than a certain threshold" is determined as a long press), thereby distinguishing more gesture variations. The pressing duration is one such lightweight temporal feature.

[0065] Optionally, the preset trigger threshold includes a preset bending trigger threshold, and the step of obtaining a valid event judgment result by judging the event trigger based on the absolute value of the relative rate of change of resistance, the preset trigger threshold, and the current candidate event further includes: When the current candidate event is determined to be a bending event, the event triggering judgment is performed based on the absolute value of the relative change rate of resistance and the preset bending triggering threshold. When the absolute value is greater than the preset bending triggering threshold, the valid event judgment result is output as a valid triggering event. When the absolute value is not greater than the preset bending triggering threshold, the valid event judgment result is output as an invalid triggering event.

[0066] In this embodiment, a tiered threshold structure is set when determining event triggering, specifically including setting a preset press trigger threshold. And direction control threshold (preset bending trigger threshold) , > After determining that the current candidate event is a press event, when When the valid event judgment result is determined, it is considered a valid trigger event, that is, the press event in the current candidate event is determined to be a valid trigger event, which can be used to trigger the corresponding interactive action in the future. After determining that the current candidate event is a bending event, when Since the ΔR / R0 of the bending event is negative, the comparison here uses its absolute value, i.e. The output is a valid event judgment result, that is, the bending event in the current candidate event is determined to be a valid trigger event, which can be used to trigger the corresponding interactive action in the future. A preset press trigger threshold is set. Used to determine whether an event has occurred, the preferred value range in this embodiment is: Direction control threshold For driving continuous direction control commands, the recommended value range is: This is to avoid accidental triggering caused by hand tremors or unintentional slight bending.

[0067] In this embodiment, to avoid false triggering by brief touch or bouncing signals, a time gating system for the press trigger mode is further implemented. Specifically, for discrete commands, this embodiment sets a preset time threshold. After determining that the current candidate event is a press event, when the absolute value of the relative change rate of resistance is greater than the preset press trigger threshold, continuous timing begins. Only when the obtained press duration exceeds the preset time threshold is the valid event judgment result output as a valid trigger event, which can then be used to trigger the corresponding interactive action.

[0068] Preferably, the preset time threshold set in this embodiment is 150-500ms.

[0069] In practice, the multi-channel parsing data includes the current candidate event, the valid event judgment result, and the pinch event judgment result. The current candidate event includes whether the event triggered by the corresponding conductive fabric channel is a pressing event or a bending event; the valid event judgment result includes whether the event triggered by the corresponding conductive fabric channel is a valid triggering event. If it is a valid triggering event, the valid event judgment result should also include the absolute value of the corresponding resistance change rate; the pinch event judgment result includes whether a pinch event was triggered between the two corresponding conductive fabric channels.

[0070] Optionally, after step S3, the following steps are also included: Real-time acquisition of the current relative rate of change of resistance; When the preset termination condition is met based on the current relative rate of change of resistance, the preset trigger threshold, and the preset fallback ratio, the event is determined to end, and a stop interaction control command is output to control the controlled device to stop executing the current interaction action.

[0071] This embodiment also introduces a hysteresis anti-jitter mechanism to prevent repeated triggering caused by signal jitter near the threshold. The hysteresis anti-jitter mechanism includes: after the event is triggered, that is, after the current interactive action is executed, a hysteresis interval is introduced, a preset fallback ratio is set, and when the current relative rate of change of resistance falls below the preset fallback ratio of the preset trigger threshold, it is determined that the preset termination condition is met, then the event is determined to end, and then a stop interactive control command is output to control the controlled device to stop executing the current interactive action.

[0072] Preferably, the preset fallback ratio in this embodiment ranges from 40% to 60%.

[0073] In the specific implementation process, after obtaining multi-channel parsing data, the event types and absolute values ​​of resistance change rates of each channel included in the multi-channel parsing data are mapped to gesture interaction control commands for downstream applications. The mapping rules are predefined or dynamically configured by the application scenario. Specifically, S3 includes: the multi-channel parsing data includes the trigger event results of each conductive fabric channel (i.e., whether the current candidate event is pressing or bending, whether it is effectively triggered, and the corresponding absolute value of the relative resistance change rate) and the judgment results of the pinch event between any two channels (whether there is a sudden change in mutual resistance). The interaction control module converts this event information into specific gesture interaction control commands according to the preset command mapping rules (which are predefined or dynamically configured by the application scenario). For example, in a three-channel drone control scenario in glove form: if the ring finger channel is determined to be a valid press trigger and the duration exceeds the time threshold, it is mapped to a takeoff / landing command; if the middle finger channel is determined to be a valid bend trigger (negative relative rate of change of resistance), it is mapped to a forward command, with the amplitude corresponding to the speed ratio; if the index finger channel is determined to be a valid bend trigger (negative relative rate of change of resistance), it is mapped to a left turn command; if a continuous pinch or contact event is detected simultaneously in the thumb and ring finger channels and the middle finger channel is validly bend-triggered, it is mapped to a composite command of "takeoff and forward". Finally, the interaction control module sends the generated gesture interaction control command to the controlled device via wired or wireless means, causing it to execute the corresponding current interaction action.

[0074] Optionally, outputting the trigger event result and the pinch event judgment result as multi-channel parsed data includes: Obtain the judgment results of several valid events corresponding to all conductive fabric channels; When it is determined that there are several effective triggering events corresponding to several adjacent conductive fabric channels in the judgment results of several effective events, the conductive fabric channel with the largest absolute value of resistance change rate is selected as the main channel from the several effective triggering events, and the remaining conductive fabric channels are selected as candidate weak channels. When the absolute value of the resistance change rate corresponding to the candidate weak channel is greater than or equal to the preset coupling ratio threshold of the absolute value of the resistance change rate corresponding to the main channel, the valid trigger event corresponding to the candidate weak channel is output. When the absolute value of the resistance change rate corresponding to the candidate weak channel is less than the preset coupling ratio threshold of the absolute value of the resistance change rate corresponding to the main channel, the valid trigger event corresponding to the candidate weak channel will not be output. The output is multi-channel parsed data based on the updated valid trigger events and pinch events.

[0075] In actual wearable use, when a user performs a bending or pressing motion with one finger, adjacent fingers may undergo unconscious slight deformation due to mechanical linkage or glove fabric, causing a slight change in the resistance signal of the corresponding channel, i.e., passive co-activation (crosstalk) phenomenon. For example, in the three-channel glove embodiment of the multi-channel micro-gesture interaction method based on conductive fabric provided in this embodiment, when the user forcefully bends the middle finger, the ring finger may be slightly moved. Although the absolute value of the resistance change rate of the bending channel is lower than that of the middle finger, it may exceed the basic trigger threshold, thus being misjudged as an independent bending event. To solve this type of crosstalk problem between channels, this embodiment further introduces an inter-channel decoupling strategy. This inter-channel decoupling strategy includes: in each sampling frame, comparing the amplitude of all channels that detect valid trigger events. The main channel is defined as the channel with the largest absolute value of the relative resistance change rate among all current valid trigger events, and the other simultaneously triggered adjacent channels are candidate weak channels. The ratio of the absolute value of the relative resistance change rate of each weak channel to the absolute value of the relative resistance change rate of the main channel is calculated. If the ratio is less than a preset coupling ratio threshold, it is determined that the triggering of the weak channel is not a conscious user action, but rather a crosstalk signal generated by the passive coordinated activation of the main channel action. This crosstalk is then suppressed, meaning the valid triggering event of the weak channel is not output to the instruction mapping module in the microcontroller. Conversely, if the amplitude ratio of a weak channel is greater than or equal to the preset coupling ratio threshold, it is considered that the user intentionally performed multi-channel gestures simultaneously, and the valid triggering event of that channel is retained.

[0076] Preferably, the preset coupling ratio threshold in this embodiment is 50%.

[0077] To balance the real-time performance of gesture recognition with system power consumption, this embodiment also optimizes the sampling rate. In normal operating mode, when performing step S1 according to the pre-set sampling rate, the sampling rate for each conductive fabric channel is recommended to be between 100Hz and 500Hz. For example, in a glove-shaped three-channel wearable interactive system, this embodiment preferably uses a sampling rate of 200Hz. This frequency is sufficient to fully capture the dynamic signals of pressing, bending, and pinching gestures, while ensuring that the decision delay for each frame is in the millisecond range, meeting the requirements of real-time interaction.

[0078] For battery-powered wearable devices (such as wristbands and cuffs), an event-driven sampling strategy can be further employed to reduce power consumption. During the idle state, each channel is intermittently sampled at a low frequency (e.g., 20Hz) to maintain only basic signal monitoring. When any channel detects that the absolute value of the relative rate of change of resistance exceeds a preset trigger threshold, it is determined that a gesture may have occurred. The sampling rate is then automatically switched to a higher frequency (e.g., 200Hz) in the operating mode to acquire the gesture signal at high resolution. After all channel signals stabilize again and remain below the idle threshold for a period of time (e.g., 0.5 seconds), the sampling rate is switched back to the low-frequency sampling mode. This event-driven sampling method can significantly reduce the average power consumption of the system when the user has no interaction, while rapidly increasing the sampling rate when a gesture occurs to ensure recognition accuracy. This is highly suitable for lightweight wearable applications with strict battery life requirements.

[0079] In its implementation, this embodiment employs a threshold adaptive strategy to provide users with personalized threshold calibration functionality. Upon first use, the system guides the user through a set of standard gestures, automatically calculating the optimal threshold based on the individual user's response amplitude to accommodate varying finger strength, joint flexibility, and fabric tightness. Specifically, upon entering calibration mode, the system guides the user via LED or voice to sequentially perform the following standard gestures: maintain a relaxed state for 5 seconds to acquire baseline resistance; press firmly 5 times, recording the relative rate of change of resistance in the pressing response; bend firmly 5 times, recording the relative rate of change of resistance in the bending response; and perform a pinch 5 times to confirm the connectivity of the pinch channel. The system automatically calculates a preset trigger threshold of 60%–80% of the minimum amplitude of the pressing and bending responses to ensure sufficient sensitivity and allowance for noise margin. The calibration results are stored in the microcontroller's non-volatile memory and loaded directly upon the next power-on.

[0080] This embodiment provides a multi-channel micro-gesture interaction system and method based on conductive fabric. Utilizing the reverse polarity relationship between the positive resistance of the conductive fabric when pressed and the negative resistance when bent, it achieves parallel recognition of three gesture modes—pressing, bending, and pinching—on the same fabric. The decision is made only by determining the sign of the resistance change rate and a threshold value, requiring no machine learning training or additional inertial sensors or other heterogeneous devices. Furthermore, it can run in real-time on low-computing-power platforms. The entire decision process consists only of subtraction, division, and comparison operations, and can operate in real-time on general-purpose, low-cost microcontrollers such as the STM32F103 at sampling rates from 100Hz to 500Hz, with negligible decision delay per frame. Regarding the press response, the baseline resistance is stable with minimal fluctuations. Experiments have verified that the relative rate of change of resistance remains stable within a certain range after multiple consecutive press events, with a short recovery time that supports natural tapping rates. In terms of the bending response, the resistance changes monotonically with the bending angle, exhibiting significant cyclic amplitude and good repeatability. Regarding the pinch response, the amplitude jumps of the open-circuit and low-resistance binary signals are much greater than those of the analog signals for pressing and bending, demonstrating the highest robustness in decision-making. The polarities of the two continuous modes (pressing and bending) are opposite, and their amplitudes are significantly higher than the baseline noise, ensuring the physical reliability of the polarity decision. This embodiment also uses a three-channel glove-shaped configuration as an example for testing in a UAV flight control scenario. Test results demonstrate a high success rate for takeoff and landing commands and a high success rate for four-directional control. A few failures are due to channel crosstalk caused by passive collaborative activation of adjacent fingers, which is an ergonomic optimization issue and does not affect the effectiveness of the decision-making method itself. Furthermore, the signal processing flow, threshold structure, and polarity decision logic of this embodiment do not depend on the manufacturing process or material composition of a specific conductive fabric. As long as the fabric exhibits a polarity relationship of positive resistance when pressed and negative resistance when bent, it is applicable. The wearing form is not limited to gloves; it can be adapted to wristbands, cuffs, neck straps, joint braces, and other forms. The three gesture modalities designed in this embodiment can be combined through different channel layouts and mapping rules to create a rich vocabulary of interaction words, supporting various interaction paradigms from simple binary commands to continuous proportional control. This allows for wide application in scenarios such as drone control, AR / VR input, smart home appliance control, and assistive technology interaction.

[0081] Example 3: This embodiment provides a wearable form alternative for a multi-channel body surface micro-gesture interaction system based on conductive fabric, including: Wristband design: Conductive fabric wraps around the wrist to detect wrist flexion (bending mode) and wrist pressure (pressing mode), suitable for scenarios such as eyes-free navigation and smartwatch assisted input; Cuff design: Conductive fabric is attached to the forearm to detect forearm pressure and elbow flexion, suitable for industrial control and rehabilitation assistance scenarios; Neckband / Neckline Design: Integrating conductive fabric into the neckband or neckline allows for hands-free interaction by detecting head tilt angle through neck flexion. Joint brace design: Conductive fabric is attached to the outside of joints such as elbows and knees to detect the flexion and extension angles and pressure movements of the joints. It is suitable for motion monitoring and rehabilitation feedback. Insole / sock sole morphology: Conductive fabric is embedded in the insole or sock sole to detect plantar pressure distribution and ankle flexion, which is suitable for gait analysis.

[0082] All of the above forms use the preset multi-layer gesture parsing logic (including modal polarity discrimination, event trigger determination and pinch detection) and signal processing method proposed in this invention, which can be achieved simply by adjusting the size, shape and channel layout of the conductive fabric according to the wearing part.

[0083] In one embodiment, when implementing dual-channel AR device control in a wristband form, two conductive fabric strips are respectively arranged on the flexor and dorsal sides of the wrist and connected to a miniature Bluetooth acquisition module; when the wrist is flexed, the fabric on the flexor side bends, and the relative rate of change of resistance is negative. A "confirm" command is triggered when the resistance is >20%; when the relative change rate of resistance on the back of the wrist is pressed, it is positive, and a "return" command is triggered when the relative change rate of resistance is >15%. This embodiment shows that this embodiment can be transferred to wrist scenarios, using polarity discrimination to analyze bending and pressing events in parallel on two channels.

[0084] In another embodiment, within the glove-shaped system, more complex gesture sequences can be identified by monitoring the temporal combination of multi-channel events: for example, simultaneous "index finger bending + middle finger bending" is identified as a "grabbing" gesture; "index finger bending" within 200ms after "ring finger pressing" is identified as a "pointing selection" gesture; and "middle finger bending" during a sustained "thumb-index finger pinch" is identified as a "zoom + rotation" gesture. Each basic event is still generated by polarity and threshold determination, and the recognition of gesture sequences only requires superimposing simple time windows and channel combination rules on them, without relying on machine learning.

[0085] In practical applications, the multi-channel micro-gesture interaction system and method based on conductive fabric provided in this embodiment are not limited to drone control scenarios and can be widely applied to various human-computer interaction fields. For example, in AR / VR device gesture input, users can interact naturally with the virtual environment by wearing gloves or wristbands made of conductive fabric, using gestures such as pressing, bending, and pinching, without needing to hold a controller; in smart home appliance control scenarios, a wristband or cuff-shaped system can be used to turn on / off, adjust, and switch modes of home appliances such as lights, air conditioners, and televisions, such as bending fingers to adjust volume or pressing to confirm selections; in surgical robot-assisted operation, surgical instruments can be remotely controlled via micro-gestures on the body surface in a sterile environment, avoiding the physical contact of traditional joysticks, improving operational flexibility and hygiene. Safety is paramount; in terms of assistive technology input, it can provide alternative interaction solutions for people with physical disabilities (such as users with limited finger movement), enabling them to complete tasks such as text input and environmental control using flexion and pressing movements of the wrist, neck, or foot; in addition, this system can also serve as a music / game interaction controller, mapping gestures to note triggers, sound effect controls, or action commands for game characters, enhancing the immersive experience; in sign language assisted translation applications, it uses multi-channel conductive fabric to detect combinations of flexion, pressing, and pinching of fingers and wrists, recognizes sign language words or letters, and translates them into speech or text output.

[0086] Example 4: Based on the above-described method embodiments, another embodiment of the present invention provides a computer-readable storage medium including a stored computer program, wherein, when the computer program is executed, it controls the device where the computer-readable storage medium is located to execute the multi-channel body surface micro-gesture interaction method based on conductive fabric described in any of the above-described method embodiments of the present invention.

[0087] The modules / units integrated in the device / terminal equipment, if implemented as software functional units and sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the above embodiments of the present invention can also be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include: any entity or device capable of carrying the computer program code, recording media, USB flash drives, portable hard drives, magnetic disks, optical disks, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signals, telecommunication signals, and software distribution media, etc.

[0088] Taking a three-channel UAV control system based on a glove as an example, the computer program's operation flow includes: During the calibration phase after power-on, the user is prompted to maintain a naturally extended finger posture. The system continuously samples the three conductive fabric channels (index, middle, and ring fingers) at a sampling rate of 200Hz. The voltage values ​​of the first 400 sampling points are then filtered using a moving average to calculate the resistance of each channel. The arithmetic mean of these values ​​is then calculated as the baseline resistance of the three channels. After calibration, an LED indicator signals that the system is ready. Subsequently, the system enters the working phase, maintaining a 200Hz sampling rate. Each frame sequentially performs the following processing: applying a 5-point moving average filter to the raw values ​​of the three channels to suppress high-frequency noise; calculating the real-time resistance of each channel from the filtered voltage values ​​using the voltage divider circuit formula; and calculating the relative rate of change of resistance for each channel.

[0089] For the ring finger channel (pre-configured as a pressing channel in this embodiment), with the natural extended posture of the finger as the calibration baseline, when the relative rate of change of resistance ΔR / R0>0 is detected (polarity is determined as a pressing event), and the absolute value is greater than the preset pressing trigger threshold T1, and continues to exceed the preset time threshold (200ms in this embodiment), a take-off command is triggered; after the drone takes off, if the same conditions are met again, a landing command is triggered.

[0090] For the middle finger and index finger channels (pre-configured as bending direction control channels in this embodiment), using the natural extension posture of the fingers as the calibration baseline, unidirectional proportional control is achieved using the negative polarity of the bending mode, which represents the relative rate of change of resistance: when the middle finger channel ΔR / R0 < 0 (polarity identified as a bending event) and its absolute value is greater than the preset bending trigger threshold T2 (i.e., ΔR / R0 < −T2), it is mapped as a forward command; when the index finger channel ΔR / R0 < 0 (polarity identified as a bending event) and its absolute value is greater than the preset bending trigger threshold T2 (i.e., ΔR / R0 < −T2), it is mapped as a left turn command; the bending channels are considered released when the fingers extend back to the baseline posture, and no reverse command is generated. The amplitude of the bending direction control command (i.e., the absolute value of the relative rate of change of resistance) can be further mapped to the forward speed or left turn rate of the UAV in the corresponding direction, achieving proportional control.

[0091] In addition, the system also introduces a hysteresis anti-jitter mechanism: after any event is triggered, the next event on the same channel is only allowed to be triggered when the absolute value of the relative rate of change of the resistance of the channel falls below 15%, thereby avoiding repeated triggering caused by signal jitter near the threshold boundary.

[0092] In the drone flight control test of the embodiment, the success rate of the take-off and landing command of the ring finger pressing channel was 10 out of 10. The overall success rate of the four-way control of the middle and index finger bending / extending channels was 8 out of 10. The two failures were caused by channel crosstalk due to the passive coordinated activation of adjacent fingers. Both failures could be improved by adjusting the tightness of the glove or by enabling the decoupling strategy between channels.

[0093] Example 5: This embodiment experimentally verifies the difference in polarity between the pressing and bending responses of a conductive fabric with a surface-deposited conductive composite coating. The experiment included: taking a 5×2cm piece of conductive fabric, spacing the electrodes at both ends 5cm apart, and connecting it to a multimeter. A pressing test was performed: the center of the fabric was pressed vertically with the fingertip. In 28 pressing events, the resistance showed a positive change, with the relative rate of change ΔR / R0 ranging from 9.5% to 15.0%. The baseline resistance R0 = 443.5 ± 0.8kΩ (coefficient of variation CV = 0.18%), and the power-law goodness of fit R² = 0.976. A bending test was performed: the fabric was attached to the back of the index finger and bent from 0° to 90°. In 23 valid events, the resistance showed a negative change, decreasing from approximately 220kΩ to 163.5kΩ, with ΔR / R0 approximately -25.1%. The results show that the response polarities of pressing and bending are significantly opposite, and the response amplitudes of both modes are much larger than the baseline noise (CV=0.18%), providing a sufficient physical basis for gesture recognition based on polarity discrimination.

[0094] This embodiment also verifies the signal characteristics of a pinching mode between two conductive fabrics with surface-deposited conductive composite coatings. This includes attaching conductive fabric pieces to the thumb and forefinger respectively; when pinched, the two conductive areas directly contact each other, closing the circuit. The mutual resistance abruptly changes from an open circuit state (megaohm level) to a low-resistance state (kiloohm level), and returns to an open circuit state when released. The amplitude jump of this binary signal (megaohm to kiloohm level) is much greater than the analog changes of pressing and bending (continuous changes within the kiloohm level), demonstrating high decision robustness and allowing it to be directly used as a discrete switch input. Furthermore, the mechanical action of the pinching mode differs from pressing and bending: pinching involves the interaction of two independent fabric areas, while pressing and bending act on a single fabric area.

[0095] Figure 4 This is an experimental characterization diagram of the negative piezoresistive response of the conductive fabric under bending provided in this embodiment. The time-series curve of the fabric resistance changing with time under continuous bending-stretching cycles is shown below. Figure 4 As shown in (a) in the figure, the horizontal axis represents time (s). Figure 4 In (a), the blue curve represents resistance, and the gray or orange gradient vertical line represents the resistance change ΔR perevent from the valley to the valley recovery peak in a single event. The vertical axis is in kiloohms (kΩ), and the color scale on the right corresponds to the resistance change ΔR (in kΩ) in a single event. The initial rest resistance at the start of the test is 193kΩ. The light green shaded area represents the static recovery stage, and the light gray shaded area represents the sensor warm-up stage. Figure 4 The top right corner of (a) shows the statistical results, specifically the resistance swing ΔR = 51.1 ± 6.6 kΩ (approximately 23.5% of the peak recovery), and the number of statistical events in this study n = 23. According to... Figure 4 The statistical results in (a) show that, in multiple consecutive bending-stretching cycles, the fabric resistance significantly decreases during bending (0° to 90°) and recovers during stretching (90° to 0°), exhibiting a stable, repeatable, and distinguishable negative piezoresistive response. After excluding the initial sensing stability segment, 23 effective bending events were statistically analyzed. The average resistance swing from the trough to the recovery peak was 51.1 ± 6.6 kΩ, and the ratio of its standard deviation to the mean (coefficient of variation) was approximately 13%, indicating that the fabric can provide a stable and consistent response to repeated bending inputs.

[0096] Detailed curves of fabric resistance versus time during a single bending-stretching event are shown below. Figure 4 As shown in (b) in the figure, Figure 4The blue curve in (b) represents the change in resistance over time; the horizontal axis represents time (s), and the vertical axis represents resistance (kΩ). In (b), the orange-yellow shaded area represents the bending stage (0°~90°); the light green shaded area represents the stretching stage (90°~0°); and the light blue shaded area represents the recovery stage (held at 0°). In a single bending-stretching event, the fabric resistance decreases from approximately 220kΩ (pre-peak) to approximately 163.5kΩ when bent to approximately 90°, and then stretches back to the recovery peak resistance of approximately 218.4kΩ. A brief rise in resistance occurs during the bending process. The resistance change ΔR in a single event is approximately 54.9kΩ, corresponding to a relative rate of change of approximately 25.1%, exhibiting an overall continuous monotonic trajectory of "decline-recovery". The above results are consistent with the -25.1% negative resistance response obtained from the bending test in this embodiment, proving that the conductive fabric used has a stable and distinguishable continuous negative piezoresistive response in the bending mode, which can serve as an effective sensing signal source for bending direction control and provides experimental support for modal polarity discrimination based on the sign of the resistance change rate.

[0097] The above description represents the preferred embodiments of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of the present invention, and these improvements and modifications are also considered to be within the scope of protection of the present invention.

Claims

1. A multi-channel body surface micro-gesture interaction system based on conductive fabric, characterized in that, It includes several conductive fabric channels and interactive control modules, wherein: The interactive control module is provided with several acquisition interfaces corresponding to several of the conductive fabric channels. For any conductive fabric channel and the corresponding acquisition interface of the conductive fabric channel, the conductive fabric channel includes conductive fabric, a first electrode and a second electrode. The first end of the first electrode is electrically connected to the first end of the conductive fabric, the first end of the second electrode is electrically connected to the second end of the conductive fabric, the second end of the first electrode is electrically connected to the first acquisition end of the acquisition interface, and the second end of the second electrode is electrically connected to the second acquisition end of the acquisition interface. The interactive control module is used to acquire, during the working phase, several real-time resistance data corresponding to several conductive fabric channels and mutual resistance status data between any two conductive fabric channels based on several acquisition interfaces. Based on preset baseline resistance data, several real-time resistance data, mutual resistance status data and preset multi-layer gesture parsing logic, it outputs multi-channel parsing data and acquires gesture interaction control commands based on the multi-channel parsing data, so as to control the controlled device to perform the current interactive action based on the gesture interaction control commands.

2. The multi-channel body surface micro-gesture interaction system based on conductive fabric as described in claim 1, characterized in that, During the calibration phase, the interactive control module is also used to acquire several baseline resistances corresponding to several conductive fabric channels based on several acquisition interfaces, and to use several baseline resistances as the preset baseline resistance data.

3. A multi-channel micro-gesture interaction method based on conductive fabric, characterized in that, Applied to a multi-channel body surface micro-gesture interaction system based on conductive fabric as described in any one of claims 1 to 2, the system includes a plurality of conductive fabric channels and an interaction control module; the method uses the interaction control module as the execution subject, including: During the work phase: Acquire several real-time resistance data corresponding to several conductive fabric channels and mutual resistance status data between any two conductive fabric channels. Based on preset baseline resistance data, several real-time resistance data, mutual resistance status data, and preset multi-layer gesture parsing logic, multi-channel parsing data is output. Gesture interaction control commands are obtained based on the multi-channel parsing data, and the controlled device is controlled to perform the current interaction action based on the gesture interaction control commands.

4. The multi-channel micro-gesture interaction method based on conductive fabric as described in claim 3, characterized in that, Also includes: During the calibration phase, several baseline resistances corresponding to several conductive fabric channels are obtained, and these baseline resistances are used as the preset baseline resistance data.

5. The multi-channel micro-gesture interaction method based on conductive fabric as described in claim 3, characterized in that, The preset multi-layer gesture parsing logic includes modal polarity judgment, event trigger judgment, and pinch judgment. The multi-channel parsing data output based on preset baseline resistance data, several sets of real-time resistance data, mutual resistance status data, and the preset multi-layer gesture parsing logic includes: The real-time resistance and baseline resistance corresponding to any of the conductive fabric channels are obtained from the preset baseline resistance data and several of the real-time resistance data. Based on the real-time resistance and the baseline resistance, modal polarity determination and event trigger determination are performed respectively to obtain the trigger event result corresponding to any of the conductive fabric channels; The mutual resistance state between any two conductive fabric channels is obtained from the mutual resistance state data, and the kneading judgment is performed based on the mutual resistance state to obtain the kneading event judgment result between any two conductive fabric channels. Obtain the results of several trigger events corresponding to all the conductive fabric channels and the results of several kneading events between all the conductive fabric channels; The results of the triggering event and the judgment result of the pinching event are output as multi-channel parsed data.

6. The multi-channel micro-gesture interaction method based on conductive fabric as described in claim 5, characterized in that, The step of determining the modal polarity and event triggering based on the real-time resistance and baseline resistance to obtain the triggering event result corresponding to any of the conductive fabric channels includes: The relative rate of change of resistance is calculated based on the real-time resistance and the baseline resistance. The current candidate event is obtained by determining the modal polarity based on the sign of the relative rate of change of resistance. The valid event judgment result is obtained by judging the event trigger based on the absolute value of the relative rate of change of resistance, the preset trigger threshold and the current candidate event; The trigger event result is obtained based on the current candidate event and the valid event judgment result.

7. The multi-channel micro-gesture interaction method based on conductive fabric as described in claim 6, characterized in that, The process of determining the modal polarity based on the sign of the relative rate of change of resistance to obtain the current candidate event includes: Modal polarity is determined based on the sign of the relative rate of change of resistance. When the sign meets the preset polarity judgment condition, the current candidate event is determined to be a pressing event. When the sign does not meet the preset polarity judgment condition, the current candidate event is determined to be a bending event.

8. The multi-channel micro-gesture interaction method based on conductive fabric as described in claim 7, characterized in that, The preset trigger threshold includes a preset press trigger threshold. The process of determining a valid event based on the absolute value of the relative rate of change of resistance, the preset trigger threshold, and the current candidate event includes: When the current candidate event is determined to be a press event: The event triggering judgment is based on the absolute value of the relative change rate of resistance and the preset press triggering threshold. When the absolute value is greater than the preset press triggering threshold, the duration acquisition action is triggered to obtain the press duration. When the duration of the press exceeds a preset time threshold, the result of the valid event judgment is output as a valid trigger event.

9. The multi-channel micro-gesture interaction method based on conductive fabric as described in claim 7, characterized in that, The preset trigger threshold includes a preset bending trigger threshold. The process of determining a valid event based on the absolute value of the relative rate of change of resistance, the preset trigger threshold, and the current candidate event further includes: When the current candidate event is determined to be a bending event, the event triggering judgment is performed based on the absolute value of the relative change rate of resistance and the preset bending triggering threshold. When the absolute value is greater than the preset bending triggering threshold, the valid event judgment result is output as a valid triggering event. When the absolute value is not greater than the preset bending triggering threshold, the valid event judgment result is output as an invalid triggering event.

10. The multi-channel micro-gesture interaction method based on conductive fabric as described in claim 6, characterized in that, After obtaining gesture interaction control commands based on the multi-channel parsed data, and controlling the controlled device to perform the current interaction action based on the gesture interaction control commands, the method further includes: Real-time acquisition of the current relative rate of change of resistance; When the preset termination condition is met based on the current relative rate of change of resistance, the preset trigger threshold, and the preset fallback ratio, the event is determined to end, and a stop interaction control command is output to control the controlled device to stop executing the current interaction action.

Images