Wearable gesture recognition glove based on single-channel linear conductive hydrogel sensor

Abstract

The invention discloses a wearable gesture recognition glove based on a single-channel linear conductive hydrogel sensor. The wearable gesture recognition glove comprises an elastic knitted glove base; the linear sensor comprises a silica gel catheter, conductive hydrogel filled in the silica gel catheter, and copper electrodes packaged at two ends of the silica gel catheter; the linear sensors are distributed in a single-channel U-shaped topological structure along finger back key joints and palm flexor muscle lines; the silica gel catheter is fixedly connected with the elastic knitted glove base after being packaged by a hot melt adhesive or polydimethylsiloxane. Through structural optimization of the single-channel linear sensor, modulus matching packaging of the ultra-thin silicone tube, hot melt adhesive physical anchoring spinning integration and strain amplification signal enhancement design, the intelligent glove can simplify the structure and improve the wearing experience, and meanwhile, accurate recognition of common gestures can be achieved.

Description

Technical Field

[0001] The present application relates to the intersection of flexible electronics, materials engineering and wearable sensing technology, and more specifically, to a wearable gesture recognition glove based on a single-channel linear conductive hydrogel sensor. Background Art

[0002] In recent years, conductive hydrogels have become a research hotspot in the field of flexible wearable sensors due to their high stretchability, biocompatibility, and adjustable conductivity. Existing technologies primarily utilize multi-channel arrays or thin-film structures to monitor human motion. For example, multiple electrodes are used to capture the angle of finger bending, or graphene / carbon nanotube composite hydrogels are used to enhance sensitivity. However, these technologies still have drawbacks in sensor structure design, packaging processes, and wearable integration stability, as shown below: (1) Multi-channel sensors lead to system complexity and insufficient wearable adaptability Existing technologies generally rely on multi-channel arrays (such as planar electrodes or distributed pressure sensors) to improve gesture recognition accuracy. For example, conductive hydrogel sensors fabricated using a sandwich structure (conductive layer + adhesive layer + elastic layer) achieve multi-dimensional signal acquisition. However, such designs require complex wiring (such as connecting independent channels), which increases device size and reduces wearer comfort. Experiments have shown that when multi-channel sensors are integrated into textile substrates, they are susceptible to signal drift due to mechanical deformation (such as repeated bending of fingers) and require high-performance algorithms to process multi-dimensional data. Furthermore, the high power consumption of multi-channel systems (e.g., microcontrollers need to process multiple signals in real time) limits their application in lightweight wearable devices.

[0003] (2) Traditional packaging technology restricts sensor performance and reliability The packaging technology of conductive hydrogels directly affects their stability and service life. In the existing technology, hydrogel packaging strategies can be divided into two categories: one is direct exposure or simple adhesion. Most studies use direct exposure of hydrogel sensors or fixation with a thin layer of adhesive. However, when the hydrogel is in direct contact with the external environment, changes in humidity and temperature will significantly affect its conductive properties. At the same time, it is easy to cause structural damage due to friction or collision during wearing, shortening the service life. The other is to use traditional rigid / semi-rigid packaging. Through a multi-layer structure, PDMS and other materials are used to wrap the hydrogel to enhance protection. However, the multi-layer packaging structure (such as insulating layer + conductive layer) increases the complexity of the process and is more expensive.

[0004] (3) Textile integration technology faces challenges in interface stability and signal noise Stably bonding flexible sensors to textiles, such as gloves, is a key challenge in practical applications. Existing technologies often rely on chemical adhesion or sewing, but these methods often mismatch the mechanical properties of the sensor interface and the textile, leading to adhesion failure under load. Furthermore, the interface between the hydrogel and textile fibers is susceptible to failure due to humidity or dynamic deformation, creating a critical challenge in balancing sensor performance and wearer comfort. Summary of the Invention

[0005] To solve the above problems, the present application provides a wearable gesture recognition glove based on a single-channel linear conductive hydrogel sensor, aiming to solve the performance bottlenecks in the field of flexible wearable gesture recognition caused by the complexity of multi-channel sensor systems, insufficient stability of hydrogel packaging, low reliability of textile integration and poor single-channel signal sensitivity.

[0006] A first aspect of an embodiment of the present invention provides a wearable gesture recognition glove based on a single-channel linear conductive hydrogel sensor, comprising: elastic knit glove base; At least one linear sensor, the linear sensor comprising a silicone tube, a conductive hydrogel filled in the silicone tube, and copper electrodes encapsulated at both ends of the silicone tube; The linear sensors are distributed in a single-channel "U"-shaped topology along the key joints on the back of the fingers and the flexor lines of the palm; The silicone catheter is fixedly connected to the elastic knitted glove base after being encapsulated by hot melt adhesive or polydimethylsiloxane.

[0007] In an optional embodiment, the preparation of the conductive hydrogel comprises the following steps: The carbomer solution is added with triethanolamine and a conductive filler, stirred to form a conductive hydrogel, and then injected into the silicone catheter. The conductive filler is carbon nanotubes or sodium chloride.

[0008] In an optional embodiment, the inner diameter of the silicone catheter is 1.2 mm and the wall thickness is 0.5 mm.

[0009] In an optional embodiment, the path layout of the linear sensor includes extending from the metacarpophalangeal joint to the distal interphalangeal joint; Among them, the linear sensory organ pathways for the ring finger and little finger adopt a differentiated layout, including extending along the metacarpophalangeal joint to the proximal interphalangeal joint.

[0010] In an optional embodiment, the silicone catheter is pre-stretched during packaging.

[0011] In an optional embodiment, a signal processing module connected to the linear sensor is further included, and the signal processing module performs gesture classification and recognition on the single-channel resistance signal through a machine learning algorithm, and the machine learning algorithm includes at least one of a random forest classifier, a decision tree classifier, a convolutional neural network, or a long short-term memory network.

[0012] In an optional implementation, the signal processing module implements gesture classification and recognition through the following steps: Based on the collected single-channel resistance timing signal, a dataset containing gesture category labels is constructed; Expand the data set and divide the expanded data set into training set and test set; The training set is input into the random forest classifier to establish a gesture classification and recognition model, and the accuracy of the gesture classification and recognition model is verified.

[0013] In the disclosed embodiments, this application achieves a systemic breakthrough through structural optimization of a single-channel linear sensor, modularly matched packaging with ultra-thin silicone tubes, physical hot-melt adhesive anchoring to textile integration, and strain amplification signal enhancement. Linear hydrogel sensors are distributed along key finger joints, and directional deformation through thin tube packaging replaces multi-channel redundant designs, reducing system complexity. An ultra-thin silicone tube package with a wall thickness of approximately 0.5 mm is designed to minimize environmental interference while maintaining stretchability. A hot-melt adhesive curing process is used to directly anchor the packaged sensor to the textile. Combined with a localized strain amplification layout, this creates a strong mapping relationship between the single-channel resistance signal and specific gestures, enabling high-precision recognition (accuracy >95%) without the need for complex algorithms. The result is a lightweight, highly reliable, and low-cost smart glove sensing system. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following briefly introduces the drawings required for use in the description of the embodiments of the present application. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0015] Figure 1 Schematic diagram of a wearable gesture recognition glove based on a single-channel linear conductive hydrogel sensor proposed in one embodiment of the present application; Figure 2 This is a schematic diagram of the linear sensor structure proposed in one embodiment of the present application; Figure 3 This is a linear sensor layout design diagram proposed in one embodiment of the present application; Figure 4 (a) is a diagram showing the synthesis path of a conductive hydrogel linear sensor according to an embodiment of the present application; Figure 4 (b) is a schematic diagram of the microstructure of the carbomer hydrogel proposed in one embodiment of the present application; Figure 5 (a) is a diagram showing the change in gesture resistance according to an embodiment of the present application; Figure 5 (b) is a diagram of classification and recognition of four gestures using a random forest algorithm proposed in an embodiment of the present application; Figure 5 (c) is a diagram of classifying and recognizing five types of gestures using a random forest algorithm proposed in an embodiment of the present application. DETAILED DESCRIPTION

[0016] The following will be combined with the drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are part of the embodiments of this application, not all of them. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0017] Please refer to Figure 1 , Figure 1 This is a flow chart of a wearable gesture recognition glove based on a single-channel linear conductive hydrogel sensor proposed in one embodiment of the present application. Figure 1 As shown, a wearable gesture recognition glove based on a single-channel linear conductive hydrogel sensor is characterized by comprising: elastic knit glove base; At least one linear sensor, the linear sensor comprising a silicone tube, a conductive hydrogel filled in the silicone tube, and copper electrodes encapsulated at both ends of the silicone tube; In this embodiment, please refer to Figure 2 , Figure 2 This is a schematic diagram of the linear sensor structure proposed in one embodiment of the present application. Figure 2 As shown, the linear sensor is made of conductive hydrogel, silicone catheter (inner and outer diameter: 1.2x2.2mm), and copper electrodes.

[0018] A carbomer conductive hydrogel with a CNT mass fraction of 2wt% was filled into a silicone tube with an inner diameter of 1.2mm (wall thickness 0.5mm). Copper wire electrodes were embedded at both ends of the silicone tube, and the interface was encapsulated with melt glue or polydimethylsiloxane (PDMS) to ensure the stability of the electrode-hydrogel interface resistance during stretching.

[0019] The linear sensors are distributed in a single-channel "U"-shaped topology along the key joints on the back of the fingers and the flexor lines of the palm; In this embodiment, the key to gesture recognition using conductive hydrogel sensors lies in the acquisition of motion signals from individual interphalangeal joints. Conductive hydrogel sensors are typically designed as long rectangular shapes to cover the fingers, but this makes the sensor's cross-sectional dimensions difficult to control. Furthermore, hydrogel sensors often come into direct contact with the skin, leading to long acquisition times and other issues, including skin discomfort, sensor adhesion stability, and bulky system wiring.

[0020] This application addresses the above issues by proposing a single-channel "U"-shaped topology. A single-channel refers to a single, continuous sensing unit. A single-channel linear sensor, on the other hand, uses a single, continuous sensing unit to acquire one-dimensional physical signals (strain, resistance, etc.) and utilizes its linear response characteristics combined with algorithms to analyze signal changes at different locations or modes. The sensor utilizes a single-channel linear layout (not an array) along key finger joints (such as the proximal interphalangeal joints and metacarpophalangeal joints). This reduces system complexity by replacing multi-channel discrete sensing with linear continuous deformation.

[0021] The difficulty in developing a single-channel linear sensor lies in the limitations of traditional approaches and the information density constraints of single-dimensional signals. Existing technologies generally employ multi-channel array sensors (5-10 independent sensing units) to achieve high gesture recognition accuracy through spatially distributed, multi-dimensional signals. However, this design suffers from drawbacks such as dense wiring, high power consumption, high noise levels, and poor comfort. Secondly, a single-channel signal is considered insufficiently dimensional and contains little information. Consequently, academia and industry have long believed that single-channel, one-dimensional signals (such as resistance) cannot distinguish multi-degree-of-freedom gestures (such as fist clenching, index finger tapping, and thumb bending). Multimodal data (such as capacitance + pressure) or multi-channel collaboration is required. A significant challenge with gesture recognition using single-dimensional data is that different movements (such as finger bending degrees) can produce similar resistance curves, making them difficult to distinguish using traditional thresholding methods, leading to gesture recognition errors. Therefore, this application utilizes machine learning to deeply decouple time series signals, ultimately achieving higher gesture recognition accuracy.

[0022] The silicone catheter is fixedly connected to the elastic knitted glove base after being encapsulated by hot melt adhesive or polydimethylsiloxane.

[0023] In this embodiment, the wearable sensing glove consists of a linear sensor, a wire, a fixing part (3D printed), hot melt adhesive and an elastic knitted glove base.

[0024] Furthermore, the preparation of the conductive hydrogel comprises the following steps: Carbomer solution is added with triethanolamine and conductive filler, stirred to form conductive hydrogel, and then injected into the silicone catheter; the conductive filler is carbon nanotube or sodium chloride.

[0025] In this embodiment, conductive hydrogels face a dilemma between flexibility (high ductility) and mechanical strength (fatigue resistance): highly flexible hydrogels are susceptible to fracture due to mechanical friction or cyclic deformation, while highly tough hydrogels, due to their high modulus, restrict deformation freedom, resulting in decreased sensitivity. While conductive fillers can improve conductivity, excessive filling can compromise ductility. Furthermore, hydrogels are susceptible to water loss due to changes in ambient temperature and humidity, which directly impacts their sensing and mechanical properties. Traditional encapsulation materials (such as silicone and TPU) have different moduli from soft hydrogels, which can easily lead to interfacial delamination or inefficient strain transfer.

[0026] Therefore, this application aims to solve the above problems, please refer to Figure 4 (a), Figure 4 This is an embodiment of the present application Figure 4 (a) is a synthetic path diagram of a conductive hydrogel linear sensor proposed in one embodiment of the present application. Figure 4 As shown in (a), 2g of carbomer (CBM) was first added to 100ml of purified water and mechanically stirred at 70°C for 2h until the CBM was completely dissolved (Solution 1). Then, an appropriate amount of 10% triethanolamine (TEOA) and the corresponding mass fraction of CNT or NaCl were added to Solution 1 and planetary stirred for 5min to complete the preparation of CBM conductive hydrogel. The prepared CBM conductive hydrogel was then filled into a silicone tube (1.2x2.2mm in inner and outer diameter) using a syringe and the port was sealed with a copper electrode. Please refer to Figure 4 (b), Figure 4 (b) is a schematic diagram of the microstructure of the carbomer hydrogel proposed in one embodiment of the present application. Figure 4 As shown in (b), carbomer is a type of polyacrylic acid copolymer. When dissolved in deionized water, its molecules will undergo hydration and stretch to a certain extent. Triethanolamine (TEOA) is added for acid-base neutralization. When the solution of the reaction system approaches neutrality, polyacrylic acid will partially ionize hydrogen ions, thereby generating negatively charged polyacrylate ions. At this time, the negatively charged molecular chains begin to repel each other due to electrostatic repulsion, and the originally curled molecular structure gradually opens due to the electric repulsion, causing the system to thicken. At the same time, the polymer molecules expand rapidly and show viscosity, causing the viscosity of the system to increase sharply, prompting the CBM aqueous solution to be in a swollen state, and finally achieving the gelation process. The schematic diagram of the gelation is shown in the figure. Figure 4 (b)

[0027] Furthermore, the inner diameter of the silicone tube is 1.2 mm and the wall thickness is 0.5 mm.

[0028] Further, the path layout of the linear sensor includes extending from the metacarpophalangeal joint to the distal interphalangeal joint; Among them, the linear sensory organ pathways for the ring finger and little finger adopt a differentiated layout, including extending along the metacarpophalangeal joint to the proximal interphalangeal joint.

[0029] In this embodiment, please refer to Figure 3 , Figure 3 This is a layout design diagram of a linear sensor proposed in one embodiment of the present application. Figure 3 As shown in the figure, the ring finger and little finger have different layouts. The specific difference is that unlike the thumb, index finger, and middle finger, whose front fixed points are all located at the midpoint of the nail line, the front fixed points of the thumb and index finger are moved from the "midpoint of the nail line" to the "midpoint of the middle phalanx" without covering the distal interphalangeal joints. The purpose is to weaken the strain resistance change of the ring finger and little finger and enhance the difference in resistance change when the different interphalangeal joints of the five fingers are bent. Compared with other rigid sensors, the differentiated layout of linear sensors can significantly reduce the size and weight of wearable smart gloves.

[0030] Furthermore, the silicone catheter is pre-stretched during packaging.

[0031] In this embodiment, the linear sensor's flexibility and stretchability can reduce the sense of restraint and discomfort felt by users during wear. Through differentiated sensor layouts and pre-stretched packaging, the elastic retraction force of the silicone tube is leveraged to amplify deformation in the target direction.

[0032] Furthermore, it also includes a signal processing module connected to the linear sensor, and the signal processing module performs gesture classification and recognition on the single-channel resistance signal through a machine learning algorithm, and the machine learning algorithm includes at least one of a random forest classifier, a decision tree classifier, a convolutional neural network or a long short-term memory network.

[0033] In this embodiment, the key to gesture recognition using conductive hydrogel sensors lies in the acquisition of motion signals from individual interphalangeal joints. Conductive hydrogel sensors are typically designed as long rectangular shapes to cover the fingers, but this makes the sensor's cross-sectional dimensions difficult to control. Furthermore, hydrogel sensors often come into direct contact with the skin, leading to long acquisition times and other issues, including skin discomfort, sensor adhesion stability, and bulky system wiring.

[0034] Furthermore, the signal processing module implements gesture classification and recognition through the following steps: Based on the collected single-channel resistance timing signal, a dataset containing gesture category labels is constructed; Expand the data set and divide the expanded data set into training set and test set; The training set is input into the random forest classifier to establish a gesture classification and recognition model, and the accuracy of the gesture classification and recognition model is verified.

[0035] In this embodiment, please refer to Figure 5 , Figure 5 (a) is a diagram showing the change in gesture resistance according to an embodiment of the present application; Figure 5 (b) is a diagram of classification and recognition of four gestures using a random forest algorithm proposed in an embodiment of the present application; Figure 5 (c) is a diagram of classifying and identifying five types of gestures using a random forest algorithm proposed in one embodiment of the present application. Figure 5 As shown in (a), for gesture recognition, five representative gestures from everyday communication were selected. When combined with multimedia tools, people with speech impairments or hearing loss can wear the gloves to convey information such as location, direction, rejection, and common sentences. Gesture data was collected while wearing the gloves, and gesture categories were labeled with numbers from 1 to 5 to represent the typical sensor response under different gestures. In constructing the dataset, each gesture was collected 50 times, and each gesture was completed within 500 seconds. To address potential overfitting in the gesture classification and recognition model and improve its generalization, data augmentation was performed on the collected data. The data at each level was expanded by flipping, ultimately resulting in 100 data sets for each gesture. The gesture samples were then divided into training and test datasets in a 4:1 ratio.

[0036] The training sample data is used as input for the decision tree classifier (DF), Transformer (TF), convolutional neural network (CNN), long short-term memory network (LSTM) and random forest classifier (RF), and then a gesture classification and recognition model is established. The confusion matrix of the classification and recognition of the four gestures shows that the classification accuracy of the random forest algorithm is as high as 98.53%. Figure 5 (b) shows that the recognition accuracy of the five gestures is 86.7%. Figure 5 (c) shown.

[0037] In summary, the present invention not only simplifies the structure of the smart gloves and improves the wearing experience, but also enables accurate recognition of commonly used gestures. The low foreign body sensation and light weight of the smart gloves allow users to use them long-term in daily life, reducing their impact on their daily lives and work, and are expected to improve communication and work for people with speech impairments in their daily lives.

[0038] Although preferred embodiments of the present invention have been described, those skilled in the art may make additional changes and modifications to these embodiments once they become aware of the basic inventive concepts. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the embodiments of the present invention.

[0039] The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the various embodiments can be referenced to each other.

[0040] Although preferred embodiments of the present invention have been described, those skilled in the art may make additional changes and modifications to these embodiments once they become aware of the basic inventive concepts. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the embodiments of the present invention.

[0041] Finally, it should be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprises," "includes," or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article, or terminal device that includes a series of elements includes not only those elements, but also other elements not explicitly listed, or elements that are inherent to such process, method, article, or terminal device. In the absence of further restrictions, an element defined by the phrase "comprises a ..." does not exclude the presence of additional identical elements in the process, method, article, or terminal device that includes the element.

[0042] The above is a detailed introduction to the wearable gesture recognition glove based on a single-channel linear conductive hydrogel sensor provided by the present application. Specific examples are used in this article to illustrate the principles and implementation methods of the present application. The description of the above embodiments is only used to help understand the method and core idea of ​​the present application. At the same time, for those skilled in the art, according to the ideas of the present application, there may be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting the present application.

Claims

1. A wearable gesture recognition glove based on a single-channel linear conductive hydrogel sensor, characterized in that: include: elastic knit glove base; At least one linear sensor, the linear sensor comprising a silicone tube, a conductive hydrogel filled in the silicone tube, and copper electrodes encapsulated at both ends of the silicone tube; The linear sensors are distributed in a single-channel "U"-shaped topology along the key joints on the back of the fingers and the flexor lines of the palm; The silicone catheter is fixedly connected to the elastic knitted glove base after being encapsulated by hot melt adhesive or polydimethylsiloxane.

2. The wearable gesture recognition glove based on a single-channel linear conductive hydrogel sensor according to claim 1, characterized in that: The preparation of the conductive hydrogel comprises the following steps: Carbomer solution is added with triethanolamine and conductive filler, stirred to form conductive hydrogel, and then injected into the silicone catheter; the conductive filler is carbon nanotube or sodium chloride.

3. The wearable gesture recognition glove based on a single-channel linear conductive hydrogel sensor according to claim 1, characterized in that: The inner diameter of the silicone catheter is 1.2 mm and the wall thickness is 0.5 mm.

4. The wearable gesture recognition glove based on a single-channel linear conductive hydrogel sensor according to claim 1, characterized in that: The path layout of the linear sensor includes extending from the metacarpophalangeal joint to the distal interphalangeal joint; Among them, the linear sensory organ pathways for the ring finger and little finger adopt a differentiated layout, including extending along the metacarpophalangeal joint to the proximal interphalangeal joint.

5. The wearable gesture recognition glove based on a single-channel linear conductive hydrogel sensor according to claim 1, characterized in that: The silicone catheter is pre-stretched during packaging.

6. The wearable gesture recognition glove based on a single-channel linear conductive hydrogel sensor according to claim 1, characterized in that: It also includes a signal processing module connected to the linear sensor, which performs gesture classification and recognition on the single-channel resistance signal through a machine learning algorithm, and the machine learning algorithm includes at least one of a random forest classifier, a decision tree classifier, a convolutional neural network, or a long short-term memory network.

7. The wearable gesture recognition glove based on a single-channel linear conductive hydrogel sensor according to claim 6, characterized in that: The signal processing module implements gesture classification and recognition through the following steps: Based on the collected single-channel resistance timing signal, a dataset containing gesture category labels is constructed; Expand the data set and divide the expanded data set into training set and test set; The training set is input into the random forest classifier to establish a gesture classification and recognition model, and the accuracy of the gesture classification and recognition model is verified.

Images