Gesture recognition sensing glove, preparation method thereof and distance typing system

Abstract

The invention discloses a gesture recognition sensing glove, a preparation method thereof and a distance typing system. The method comprises the following steps: taking an elastic insulating yarn as a core yarn, taking a conductive yarn as a wrapping yarn, and uniformly and spirally winding the conductive yarn on the surface of the elastic insulating yarn through a wrapping process to form an elastic strain sensing yarn; then, elastic strain sensing yarns are sewn on the hand back of each finger of the duck web structure glove to cover the whole finger to form on-finger strain sensing yarns so as to obtain longitudinal tensile strain resistance signals generated by bending or straightening the fingers; elastic strain sensing yarns are sewed on interfinger areas of the glove with the duck web structure to form interfinger strain sensing yarns so as to obtain transverse tensile strain information generated by opening or closing fingers. The gesture recognition sensing glove and the distance typing system provided by the invention have high sensitivity and high stability.

Description

Technical Field

[0001] This invention belongs to the field of gesture recognition, specifically a gesture recognition sensing glove, its preparation method, and a remote typing system. Background Technology

[0002] With the development of human-computer interaction technology, gesture recognition, as a natural and intuitive interaction method, has been widely used in fields such as smart wearable devices, virtual reality, augmented reality, and smart homes. Among them, gesture recognition technology based on sensor gloves has become a development hotspot due to its high accuracy and stability.

[0003] Currently, common sensing gloves mainly rely on visual sensors (e.g., the document with application number 202110252206.3), inertial measurement units (e.g., the document with application number 201710128589.7), or resistance strain sensors (e.g., the document with application number 201911124166.3) to sense finger movement information. However, optical sensors are greatly affected by ambient light and are prone to occlusion problems; although IMU sensors can provide three-dimensional motion information, they have cumulative errors, which will reduce recognition accuracy with long-term use; the traditional resistance strain sensor has a relatively complex fabrication process, making gesture recognition complicated and lacking universality, making it difficult to meet the needs of industrial development.

[0004] In recent years, advancements in flexible electronics technology have provided new solutions for gesture recognition. Flexible sensing yarns, due to their excellent mechanical flexibility and high sensitivity, have become ideal materials for wearable sensors. However, existing flexible sensing gloves mostly integrate sensors at the finger joints, failing to achieve real-time three-dimensional spatial monitoring of all fingers, including finger opening and closing. Current gloves primarily focus on the bending information of the finger joints, making it difficult to accurately capture the movement trajectory of the fingers in space, thus affecting the accuracy of gesture recognition. Furthermore, challenges remain regarding gesture complexity and training costs. For example, most gesture recognition gloves rely on custom gestures created by developers, hindering the universality and widespread adoption of gesture operations.

[0005] Therefore, developing a gesture recognition glove that integrates flexible sensing yarn and flexible sensor to achieve a high-precision, low-cost, and low-power universal air typing system will greatly expand its application potential in the field of intelligent interaction. Summary of the Invention

[0006] To address the shortcomings of existing technologies, this invention provides a gesture recognition sensing glove, its preparation method, and a remote typing system.

[0007] The technical solution of the present invention to solve the aforementioned technical problem is to provide a method for preparing a gesture recognition sensing glove, characterized in that the method includes the following steps: Step 1: Prepare elastic strain sensing yarn: Use elastic insulating yarn as core yarn and conductive yarn as wrapping yarn. Through the wrapping process, the conductive yarn is evenly spirally wound on the surface of the elastic insulating yarn to form an elastic strain sensing yarn with a regular stacked structure. Step 2: Sew the elastic strain sensing yarn obtained in Step 1 onto the back of each finger of the duck web structure glove, covering the entire finger to form a finger strain sensing yarn, so as to obtain the longitudinal tensile strain resistance signal generated by bending or straightening the finger. The elastic strain sensing yarn obtained in step 1 is sewn onto the finger area of ​​the duck web structure glove to form the finger strain sensing yarn, so as to obtain the lateral tensile strain information generated by opening or closing the fingers.

[0008] Compared with the prior art, the beneficial effects of the present invention are as follows: (1) The gesture recognition sensing glove and air typing system provided by this invention have high sensitivity and high stability. The sensors of general gesture recognition gloves are only located on the back of the fingers, which can only capture the bending angle of the fingers (two-dimensional data) and cannot capture the three-dimensional spatial changes of the fingers. In this invention, the inventors have rationally arranged the flexible sensing yarn on the back and between the fingers of the glove. By using machine learning algorithms, the three-dimensional spatial position of the fingers can be accurately captured, thereby achieving accurate motion capture, effectively improving the accuracy of gesture recognition and reducing the false recognition rate.

[0009] (2) The present invention develops a low-power wearable data acquisition circuit to achieve low-power operation and improve the portability and continuous working capability of the system.

[0010] (3) This glove and recognition system are suitable for applications such as air typing, human-computer interaction, virtual reality / augmented reality, and intelligent medical rehabilitation. Compared with traditional gesture recognition systems, this invention has a simple structure, low manufacturing cost, is easy to mass-produce, and can be seamlessly integrated with existing electronic devices, further enhancing its practicality and market application value. In summary, the sensing glove and gesture recognition system of this invention have broad application prospects in the fields of intelligent interaction, wearable devices, and medical rehabilitation.

[0011] (4) A wrapped composite strain sensing yarn with a stacked structure is integrated into the knuckles and interdigital areas of the glove as a key functional unit in the gesture recognition glove to realize multi-degree-of-freedom gesture detection.

[0012] (5) When the core yarn is stretched when the finger is bent, the conductive yarn on its surface changes due to the change in the stacking structure, which causes the resistance to change, thereby realizing the perception of the gesture state. Attached Figure Description

[0013] Figure 1 This is a schematic diagram of the elastic strain sensing yarn of the present invention; Figure 2 This is a schematic diagram of a gesture recognition sensing glove according to an embodiment of the present invention, in a state where the palm is naturally open and the fingers are naturally spread. Figure 3 This is a schematic diagram of a gesture recognition sensing glove in another embodiment of the present invention, with the palm naturally open and the fingers naturally spread. Figure 4 This is a confusion matrix diagram showing the gesture recognition accuracy of Embodiment 1 of the present invention.

[0014] In the diagram, conductive yarn 1, elastic insulating yarn 2, interfinal strain sensing yarn 3, finger strain sensing yarn 4, and webbed glove 5 are shown. Detailed Implementation

[0015] Specific embodiments of the present invention are given below. These specific embodiments are only used to further illustrate the present invention in detail and do not limit the scope of protection of the present invention.

[0016] This invention provides a method for preparing a gesture recognition sensing glove (hereinafter referred to as the method), characterized in that the method includes the following steps: Step 1: Preparation of elastic strain sensing yarn: Using elastic insulating yarn 2 as core yarn and conductive yarn 1 as wrapping yarn, the conductive yarn 1 is evenly spirally wound on the surface of elastic insulating yarn 2 through a wrapping process to form an elastic strain sensing yarn (hereinafter referred to as elastic strain sensing yarn) with a regular stacked structure. Preferably, in step 1, the conductive yarn 1 is a highly conductive material, preferably silver-plated yarn, stainless steel yarn, graphene yarn, or carbon nanotube yarn; the elastic insulating yarn 2 is spandex filament, Ecoflex filament, or PDMS filament.

[0017] Preferably, in step 1, the linear density of the conductive yarn 1 is 20~70D; and the linear density of the elastic insulating yarn 2 is 420~1120D.

[0018] Preferably, in step 1, the wrapping process specifically involves: inserting the elastic insulating yarn 2 into the core yarn channel of the wrapping machine and fixing it between the yarn infeed and yarn take-up shafts, maintaining appropriate tension to ensure the stability of the wrapping process; wrapping the conductive yarn 1 around the rotating wrapping arm of the wrapping machine, and by adjusting the rotation speed and axial feed speed of the wrapping arm, uniformly spirally wrapping the conductive yarn 1 around the surface of the elastic insulating yarn 2 at a set wrapping angle, forming an elastic strain sensing yarn with a regularly stacked structure. The wrapping process is carried out in a wrapping machine with tension control and wrapping angle adjustment functions.

[0019] Preferably, in step 1, the longitudinal elongation of the elastic strain sensing yarn is 100-300%, and the recovery rate is 97-100%.

[0020] Preferably, in step 1, during the winding process, the winding angle of the conductive yarn 1 is 90~120° and the winding density is 50~400 twists / cm, so as to achieve a sensitive response to tensile deformation.

[0021] Preferably, in step 1, after the winding is completed, the elastic strain sensing yarn is slowly wound to the take-up beam by the winding system, and the end of the elastic strain sensing yarn is sealed with hot melt adhesive or insulating tape to prevent the conductive yarn 1 from loosening and to facilitate the subsequent sewing process.

[0022] Preferably, in step 1, the final sensing yarn has good tensile adaptability and strain sensitivity, and can be used for subsequent sewing and integration into the glove knuckles or interdigital areas to construct a multi-channel resistive sensing network.

[0023] Step 2: Sew the elastic strain sensing yarn obtained in Step 1 onto the back of each finger of the duck web structure glove 5, covering the entire finger, to form the finger strain sensing yarn 4, so as to obtain the longitudinal tensile strain resistance signal generated by bending or straightening the finger. The elastic strain sensing yarn obtained in step 1 is sewn onto the interdigital area of ​​the duck web structure glove 5 to form interdigital strain sensing yarn 3, so as to obtain information on the lateral tensile strain generated when the fingers are opened or closed. Preferably, in step 2, the webbed glove 5 is an electrically insulating elastic five-finger glove with a webbed structure. The interfinal area is the webbed structure between two adjacent fingers.

[0024] Preferably, in step 2, the specific process of sewing the finger strain sensing yarn 4 is as follows: throughout the sewing process, the webbed glove 5 is placed naturally to ensure that the finger strain sensing yarn 4 has effective stretching space in actual use; sewing begins, and the elastic strain sensing yarn is evenly sewn along the length of the finger onto the back of each finger of the webbed glove 5 using a sewing needle, ensuring that it fits tightly against the finger surface while avoiding excessive pre-stress; after sewing, the two ends of the elastic strain sensing yarn are fixed to the base and tip of the finger of the webbed glove 5 respectively, so that it can undergo axial stretching deformation when the finger is bent, inducing a change in the stacked structure of the conductive yarn 1; the two ends of the conductive yarn 1 in the finger strain sensing yarn 4 are communicatively connected to the corresponding channels of the multi-channel data acquisition module of the air typing system; Preferably, in step 2, the specific process of sewing the interdigital strain sensing yarn 3 is as follows: throughout the sewing process, the webbed glove 5 is placed naturally to ensure that the interdigital strain sensing yarn 3 has effective stretching space in actual use; sewing begins, and the elastic strain sensing yarn is sewn onto the webbed structure between two adjacent fingers using a sewing needle; after sewing, the two ends of the elastic strain sensing yarn are fixed to the adjacent sides of the two adjacent fingers of the webbed glove 5 to ensure that the elastic strain sensing yarn can be fully stretched when the fingers are closed, thereby activating the stacked structure deformation response mechanism in the conductive yarn 1; the two ends of the conductive yarn 1 in the interdigital strain sensing yarn 3 are communicatively connected to the corresponding channels of the multi-channel data acquisition module of the air typing system; Preferably, in step 2, during the sewing of the interfinal strain sensing yarn 3, one end of the elastic strain sensing yarn is fixed to the side of a finger, then passes through the webbed structure of the webbed glove 5, and into the adjacent side of an adjacent finger; then passes through the webbed structure of the webbed glove 5, and back to the same side of the first finger; the above process is repeated to form an S-shaped interfinal strain sensing yarn 3 (e.g., ...). Figure 3 (As shown).

[0025] Preferably, in step 2, the resistance change of the interfinite strain sensing yarn 3 can be accurately detected by the data acquisition circuit, generally at least 10Ω, so that the detection error is less than 10%.

[0026] Preferably, in step 2, to achieve electrical connection with an external circuit, the two ends of the interdigital strain sensing yarn 3 and the finger strain sensing yarn 4, which are not sewn onto the webbed glove 5, need to be untwisted. This involves rotating the conductive yarn 1 to separate it from the elastic insulating yarn 2, and cutting off the excess elastic insulating yarn 2, leaving only the exposed conductive yarn 1 as the output electrode. This electrode is then connected to the corresponding channel of the multi-channel data acquisition module of the air typing system via welding, weaving, or connectors to acquire resistance signals of the strain state of each finger joint area. After this step, the entire glove will have multi-channel finger joint motion sensing capabilities, providing a basic sensing unit for the subsequent construction of a gesture recognition system.

[0027] The conductive yarn 1 is untwisted and separated from the elastic insulating yarn 2 by rotation, retaining the conductive yarn 1 as the output electrode, and removing excess elastic insulating yarn 2. The processed conductive yarn 1 can be connected to the input end of the data acquisition system through weaving, welding, or connectors to achieve independent acquisition of resistance signals for lateral deformation between fingers. After this sewing and connection step is completed, the glove can not only sense the degree of bending of each finger joint, but also monitor the opening and closing state between adjacent fingers in real time, further enriching the input dimensions of the gesture recognition system and improving recognition accuracy and the ability to distinguish complex gestures.

[0028] This invention also provides a method for preparing a gesture recognition sensor glove.

[0029] The present invention also provides a remote typing system (hereinafter referred to as the system) based on the gesture recognition sensing glove, characterized in that the system includes a gesture recognition sensing glove and a low-power data acquisition circuit; the circuit includes a multi-channel data acquisition module, a signal processing module, a wireless module and a power supply module, with an overall power consumption of about 5 milliwatts; The electrodes led out from the conductive yarn 1 of the gesture recognition sensing glove are respectively connected to the input ports of the multi-channel data acquisition module, ensuring that the strain sensing yarn 4 on each finger and the strain sensing yarn 3 between each finger are connected to the system as independent channels, realizing the synchronous acquisition and orderly transmission of multi-channel resistance signals; the output of the multi-channel data acquisition module is connected to the input of the signal processing module, and the output of the signal processing module is connected to the external host computer through the wireless module; the power supply module supplies power to the entire system.

[0030] Preferably, the wireless module is a low-power Bluetooth module; the signal processing module is an analog-to-digital converter (ADC).

[0031] Preferably, the multi-channel data acquisition module acquires the resistance signal of the gesture recognition sensor glove and transmits it to the signal processing module; the signal processing module converts the resistance signal into a digital signal and transmits it wirelessly to an external host computer in real time; then the host computer performs noise reduction processing on the raw data based on the system's preset filtering algorithm and inputs it into the gesture recognition model for classification and judgment, thereby realizing real-time analysis and recognition of hand movements.

[0032] Preferably, the power module uses a lithium battery or a thermoelectric fabric based on a variable cross-section thermoelectric unit; when a thermoelectric fabric based on a variable cross-section thermoelectric unit is used as the power module, the system also includes a power management module. During the connection process, it should be ensured that each electrode is firmly connected, with low contact resistance and clear channel numbering, so as to facilitate the subsequent training of the recognition model and channel mapping.

[0033] After all channels are connected, the wearer keeps their hand still and in a natural position. The system begins to collect the resistance values ​​of each sensing channel in the initial inactive state, serving as reference data for the system's baseline state. This process is operated through a software interface, automatically recording the initial resistance value of each channel and establishing a correspondence between channel numbers and fingers to form a preliminary channel calibration matrix. Based on this, the response characteristics of each elastic strain sensing yarn can be further calibrated under specific actions, such as slowly bending a finger, opening or closing a finger gap, etc. By recording the trend of resistance changes over time, training data and boundary conditions are provided for subsequent gesture classification models.

[0034] One welded side of the thermoelectric fabric faces the skin to absorb metabolic heat, while the front surface is in contact with the air or external environment, creating a temperature difference that drives the thermoelectric power generation effect. The electrodes of the thermoelectric fabric are connected to a power management module via flexible wires. This module has voltage regulation, filtering, and adjustment functions to ensure the output voltage remains stable within the range required for the acquisition board and processing module to operate (e.g., 3.8V or 5V). The module as a whole adopts a flexible encapsulation design, ensuring stable power supply performance even under dynamic conditions such as hand bending and compression. This self-powered solution significantly reduces the system's dependence on external batteries or power sources, improving the continuous operation capability of the wearable system and user convenience. Simultaneously, by combining the high thermal resistance design strategy of the thermoelectric unit, it maintains power supply efficiency without generating significant heat on the skin, further improving wearing comfort. After this step, the sensing system possesses preliminary energy self-sufficiency capability, laying the hardware foundation for truly passive gesture recognition.

[0035] Example 1: (1) Using a commercial multi-functional wrapping machine, first install 40D silver-plated yarn as the wrapping yarn on the rotating wrapping arm. Adjust the wrapping machine parameters to control the wrapping angle at approximately 90°, and set the wrapping density to 400 twists / cm to ensure the wrapping density and wrapping angle. Pass 560D spandex yarn as the core yarn through the core yarn guide rail, maintain a certain tension, and begin wrapping to ensure that the conductive silver-plated yarn is evenly and spirally wrapped around the surface of the elastic spandex yarn. Control the wrapping length within the required sensing yarn length range. After obtaining the composite yarn, measure the resistance in the range of 70~80Ω / cm.

[0036] (2) Using a sewing needle, sew the prepared elastic sensing yarn onto the two joints of each finger on the glove in a naturally extended state. The stitches should be even and close to the skin to avoid looseness or local pressure. Use glue to fix the two ends of the sensing yarn to the base and tip of the fingers on the glove to ensure that it is firmly fixed and does not affect the elastic deformation. The unsewn parts are left for subsequent electrode processing.

[0037] (3) The elastic sensing yarn is pre-stretched by 10% to give it a certain initial tension. Then, it is sewn to the web area between two adjacent fingers with a sewing needle. During the sewing process, the sensing yarn is spread evenly along the web direction and the sewing is tight so that it can generate lateral strain as the fingers open and close. The sewn ends are also fixed with glue. The unsewn free ends are separated to obtain conductive silver-plated yarn through a detwisting process. Excess spandex is cut off and the exposed conductive yarn 1 is sewn onto the glove as an electrode to facilitate subsequent electrical connection.

[0038] (4) Connect all the sewn elastic sensing yarn electrodes to the input terminal of the data acquisition board in sequence to ensure that the strain sensing yarn 4 on the finger and the strain sensing yarn 3 between the fingers of each finger are independent to avoid signal crosstalk. After the connection is completed, keep the wearer's hand still and collect the initial resistance value of each channel as the system reference for subsequent normalization processing and error correction of gesture action signals.

[0039] (5) A multi-unit thermoelectric power supply fabric is integrated in the palm area of ​​the glove. This fabric adopts a high thermal resistance design, optimizes the heat flow and interface heat transfer matching, absorbs heat from the human skin and converts it into stable electrical energy output. The thermoelectric fabric electrodes are connected to the power input terminals of the data acquisition board and signal processing module through flexible wires to ensure that the system can be self-powered and has stable power consumption, thereby improving the system's independence and portability.

[0040] (6) The data acquisition board has a multi-channel analog-to-digital conversion function, which can simultaneously acquire resistance signals from multiple sensor channels. The acquired raw data is transmitted to the computer in real time through a wireless communication module (such as BLE). The computer runs a dedicated data filtering algorithm to filter out noise, segment the time window, normalize the data, and reduce the dimensionality of the signal. Combined with a pre-trained gesture recognition model, it realizes high-precision, real-time gesture classification and air typing functions.

[0041] Depend on Figure 4 As can be seen, the gesture recognition sensor glove system can accurately predict English letters and spaces, with an average classification accuracy of 97.2%.

[0042] Example 2: (1) Using a commercial wrapping machine, 20D graphene fiber bundles are installed on the wrapping arm, the wrapping angle is adjusted to about 90°, and the wrapping density is set to 380 twists / cm, and wrapped on 420D spandex yarn. The core yarn tension is kept uniform to form a uniform wrapping structure. The resulting composite yarn has a unit length resistance in the range of 1500~1800Ω / cm, which meets the characteristics of graphene's conductivity.

[0043] (2) Use a sewing needle to sew the elastic sensing yarn to the two joints of each finger on the glove, ensuring that the stitches fit snugly and do not affect the stretch elasticity. Secure the ends with glue to ensure a firm stitch, leaving some unsewn ends for later use.

[0044] (3) The elastic sensing yarn is pre-stretched by 7% and sewn to the webbed area between two adjacent fingers. When sewing, ensure that the sensing yarn is flat and wrinkle-free so as to adapt to the stretching deformation caused by the opening and closing of the fingers. The two ends are fixed with glue. The unsewn end is separated by untwisting the conductive graphene yarn, and the excess spandex core yarn is cut off. The exposed conductive yarn 1 is sewn onto the glove as an electrode to facilitate subsequent electrical connection.

[0045] (4) Connect all elastic sensing yarn electrodes sequentially to the input terminal of the data acquisition board to ensure that the strain sensing yarn 4 on each finger and the strain sensing yarn 3 between the fingers can independently acquire signals. Acquire the initial resistance value while the wearer is stationary to complete the benchmark calibration and ensure the accuracy and stability of the data acquisition.

[0046] (5) The data acquisition board has a wireless transmission function, which can transmit the acquired resistance signal to the computer in real time. First, the sensor data is denoised, segmented, normalized and dimensionality reduced by using a data filtering algorithm, and combined with the preset gesture recognition model to achieve high-precision gesture recognition.

[0047] Any aspects not covered in this invention are applicable to existing technologies.

Claims

1. A method for preparing a gesture recognition sensing glove, characterized in that, The method includes the following steps: Step 1: Prepare elastic strain sensing yarn: Use elastic insulating yarn (2) as core yarn and conductive yarn (1) as wrapping yarn. By wrapping process, the conductive yarn (1) is evenly spirally wound on the surface of elastic insulating yarn (2) to form an elastic strain sensing yarn with a regular stacking structure. Step 2: Sew the elastic strain sensing yarn obtained in Step 1 onto the back of each finger of the duck web structure glove (5), covering the entire finger to form a finger strain sensing yarn (4) to obtain the longitudinal tensile strain resistance signal generated by bending or straightening the finger. The elastic strain sensing yarn obtained in step 1 is sewn onto the interdigital area of ​​the webbed glove (5) to form interdigital strain sensing yarn (3) to obtain information on the lateral tensile strain generated when the fingers are opened or closed.

2. The method for preparing the gesture recognition sensing glove according to claim 1, characterized in that, In step 1, the conductive yarn (1) is a highly conductive material, preferably silver-plated yarn, stainless steel yarn, graphene yarn or carbon nanotube yarn; the elastic insulating yarn (2) is spandex filament, Ecoflex filament or PDMS filament; In step 1, the linear density of the conductive yarn (1) is 20~70D; the linear density of the elastic insulating yarn (2) is 420~1120D.

3. The method for preparing the gesture recognition sensing glove according to claim 1, characterized in that, In step 1, the wrapping process is as follows: the elastic insulating yarn (2) is inserted into the core yarn channel of the wrapping machine and fixed between the yarn feed shaft and the yarn take-up shaft, maintaining appropriate tension to ensure the stability of the wrapping process; the conductive yarn (1) is wrapped around the rotating wrapping arm of the wrapping machine, and by adjusting the rotation speed and axial feed speed of the wrapping arm, the conductive yarn (1) is evenly spirally wrapped around the surface of the elastic insulating yarn (2) at the set wrapping angle to form an elastic strain sensing yarn with a regular stacked structure; In step 1, the wrapping process is carried out in a wrapping machine with tension control and wrapping angle adjustment functions; In step 1, during the winding process, the winding angle of the conductive yarn (1) is 90~120° and the winding density is 50~400 twists / cm, so as to achieve a sensitive response to tensile deformation.

4. The method for preparing the gesture recognition sensing glove according to claim 1, characterized in that, In step 1, the longitudinal elongation of the elastic strain sensing yarn is 100-300%, and the recovery rate is 97-100%.

5. The method for preparing the gesture recognition sensing glove according to claim 1, characterized in that, In step 2, the duck web structure glove (5) is an electrically insulating elastic five-finger glove with a duck web structure, and the interfinal area is the duck web structure between two adjacent fingers.

6. The method for preparing the gesture recognition sensing glove according to claim 1, characterized in that, In step 2, the specific process of sewing the finger strain sensing yarn (4) is as follows: During the entire sewing process, the duck web structure glove (5) is placed naturally to ensure that the finger strain sensing yarn (4) has effective stretching space in actual use; sewing begins, and the elastic strain sensing yarn is evenly sewn along the length of the finger onto the back of each finger of the duck web structure glove (5) using a sewing needle to ensure that it fits tightly against the finger surface while avoiding excessive prestress; after sewing, the two ends of the elastic strain sensing yarn are fixed to the finger root and fingertip of the duck web structure glove (5) respectively, so that it can undergo axial stretching deformation when the finger is bent, inducing changes in the stacking structure of the conductive yarn (1); the two ends of the conductive yarn (1) in the finger strain sensing yarn (4) are communicatively connected to the corresponding channels of the multi-channel data acquisition module of the air typing system; In step 2, the specific process of sewing the interdigital strain sensing yarn (3) is as follows: During the entire sewing process, the webbed glove (5) is placed naturally to ensure that the interdigital strain sensing yarn (3) has effective stretching space in actual use; start sewing and use a sewing needle to sew the elastic strain sensing yarn onto the webbed structure between two adjacent fingers; after sewing, fix the two ends of the elastic strain sensing yarn to the adjacent sides of the two adjacent fingers of the webbed glove (5) respectively to ensure that the elastic strain sensing yarn can be fully stretched when the fingers are closed, thereby activating the stacked structure deformation response mechanism in the conductive yarn (1); the two ends of the conductive yarn (1) in the interdigital strain sensing yarn (3) are communicatively connected to the corresponding channels of the multi-channel data acquisition module of the air typing system.

7. The method for preparing the gesture recognition sensing glove according to claim 1 or 6, characterized in that, In step 2, during the sewing of the interfinal strain sensing yarn (3), one end of the elastic strain sensing yarn is fixed to the side of a finger, then passes through the web structure of the web-structured glove (5) and into the adjacent side of an adjacent finger; then passes through the web structure of the web-structured glove (5) and back to the same side of the first finger; repeat the above process to form an S-shaped interfinal strain sensing yarn (3). In step 2, the resistance change of the interdigital strain sensing yarn (3) is at least 10Ω.

8. A gesture recognition sensor glove prepared by the method of any one of claims 1-7.

9. A hands-free typing system based on the gesture recognition sensing glove of claim 8, characterized in that, The system includes a gesture recognition sensing glove and a low-power data acquisition circuit; the circuit includes a multi-channel data acquisition module, a signal processing module, a wireless module, and a power supply module. The electrodes led out from the conductive yarn (1) of the gesture recognition sensing glove are respectively connected to the input port of the multi-channel data acquisition module, ensuring that the strain sensing yarn (4) on each finger and the strain sensing yarn (3) between the fingers are connected to the system as independent channels, so as to realize the synchronous acquisition and orderly transmission of multi-channel resistance signals; the output of the multi-channel data acquisition module is connected to the input of the signal processing module, and the output of the signal processing module is connected to the external host computer through the wireless module; the power supply module supplies power to the entire system.

10. The air typing system according to claim 9, characterized in that, The wireless module uses a low-power Bluetooth module; the signal processing module uses an analog-to-digital converter module; the power module uses a lithium battery or a thermoelectric fabric based on a variable cross-section thermoelectric unit; when a thermoelectric fabric based on a variable cross-section thermoelectric unit is used as the power module, the system also includes a power management module. The multi-channel data acquisition module collects the resistance signal of the gesture recognition sensor glove and transmits it to the signal processing module. The signal processing module converts the resistance signal into a digital signal and transmits it wirelessly to an external host computer in real time. Then, the host computer performs noise reduction processing on the raw data based on the system's preset filtering algorithm and inputs it into the gesture recognition model for classification and judgment, thereby realizing real-time analysis and recognition of hand movements.

Images