CN111553307B - Gesture recognition system fusing bioelectrical impedance information and myoelectric information - Google Patents
Gesture recognition system fusing bioelectrical impedance information and myoelectric information Download PDFInfo
- Publication number
- CN111553307B CN111553307B CN202010390176.8A CN202010390176A CN111553307B CN 111553307 B CN111553307 B CN 111553307B CN 202010390176 A CN202010390176 A CN 202010390176A CN 111553307 B CN111553307 B CN 111553307B
- Authority
- CN
- China
- Prior art keywords
- module
- gesture recognition
- information
- signal
- data
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/20—Movements or behaviour, e.g. gesture recognition
- G06V40/28—Recognition of hand or arm movements, e.g. recognition of deaf sign language
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/22—Matching criteria, e.g. proximity measures
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/25—Fusion techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G08—SIGNALLING
- G08C—TRANSMISSION SYSTEMS FOR MEASURED VALUES, CONTROL OR SIMILAR SIGNALS
- G08C17/00—Arrangements for transmitting signals characterised by the use of a wireless electrical link
- G08C17/02—Arrangements for transmitting signals characterised by the use of a wireless electrical link using a radio link
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/12—Classification; Matching
- G06F2218/16—Classification; Matching by matching signal segments
Abstract
The invention discloses a gesture recognition system fusing bioelectrical impedance information and bioelectrical muscle information, which comprises: the system comprises a power supply module, an electrical impedance information acquisition module, a biological myoelectric signal acquisition module, a microcontroller module, a wireless transmission module and an upper computer; the host computer includes: the system comprises a data fusion module and a machine learning classification module; the impedance information acquisition module is responsible for acquiring impedance distribution information inside the wrist, the biological electromyographic signal acquisition module is responsible for acquiring the electromyographic information of the arm, and the two types of biological information are transmitted to the upper computer for data processing through the wireless transmission module under the coordination control of the microcontroller; on the upper computer, the two kinds of biological information are fused through a data fusion processing algorithm, and a machine learning classification algorithm is adopted for training and finishing gesture recognition. The invention can enhance the robustness and diversity of the gesture recognition method and improve the accuracy of the gesture recognition method by fusing the two biological signals.
Description
Technical Field
The invention relates to the field of gesture recognition in the field of human-computer interaction, in particular to a method for performing gesture recognition by fusing bioelectrical impedance signals and bioelectrical electrical signals.
Background
With the continuous development of computer technology, human-computer interaction is concerned more and more by researchers, while traditional human-computer interaction modes such as a mouse and a keyboard cannot meet the requirements, and new human-computer interaction modes pay more attention to humanization and perception interaction. The hand is the most flexible limb of the human body, rich meanings can be expressed through gesture actions, and especially for deaf-mutes, sign language is the most important mode for the deaf-mutes to communicate with the outside. Therefore, research on gesture recognition has been an important content in the field of human-computer interaction.
Gesture recognition using computer vision is the most common method. The camera is used for collecting images during hand movement, and the collected images are subjected to image denoising, image segmentation, image feature extraction and other operations, so that recognition is completed. However, the necessity of a camera limits the use environment of the method, so that the method has great limitation and is easily influenced by factors such as lighting conditions, shading and background environment. Data gloves, another common identification method, have high identification accuracy due to its complex multi-sensor measurement module. The hand motion recognition method is characterized in that sensors are arranged at joints of a hand to acquire information such as speed, direction and angle during hand motion, and then data analysis and processing are performed to complete recognition. The complex structure of this approach presents a significant drawback for everyday wear and use.
Disclosure of Invention
The invention aims to provide a gesture recognition system capable of fusing electrical impedance information and an electromyographic signal aiming at some defects of the existing gesture recognition mode so as to improve the accuracy of gesture recognition and the diversity and robustness of gesture recognition.
In order to achieve the purpose, the invention adopts the following technical scheme:
the invention relates to a gesture recognition system fusing bioelectrical impedance information and bioelectricity information, which is characterized by comprising the following components: the system comprises a power supply module, an electrical impedance information acquisition module, a biological electromyographic signal acquisition module, a microcontroller module, a wireless transmission module and an upper computer; the host computer includes: the system comprises a data fusion module and a machine learning classification module;
the electrical impedance information acquisition module consists of a signal excitation unit, a first measurement unit and a first acquisition electrode;
the first acquisition electrode is arranged on the wrist according to an array structure, and the signal excitation unit injects an excitation signal into the first acquisition electrode under the control of the microcontroller module; the first measuring unit acquires an impedance distribution signal on the first collecting electrode under the control of the microcontroller module and sends the impedance distribution signal to the upper computer through the wireless transmission module;
the biological electromyographic signal acquisition module consists of a second acquisition electrode and a second measurement unit;
the second collecting electrode is arranged on the arm according to an array structure, and the second measuring unit acquires the electromyographic signals on the second collecting electrode under the control of the microcontroller module and sends the electromyographic signals to the upper computer through the wireless transmission module;
the data fusion processing module is used for carrying out frequency matching on the received impedance distribution signal and the received electromyographic signal so as to obtain fused data;
the machine learning classification module comprises a deep cycle neural network model and a traditional machine learning model;
the deep circulation neural network model performs feature extraction and training on the fused data to obtain a trained dynamic model;
performing feature extraction and training on the fused data of the machine learning model to obtain a trained static model;
and the machine learning classification module judges whether the myoelectric information detection threshold is met or not according to the acquired real-time fusion data, if so, the dynamic model is called to complete gesture recognition, and otherwise, the static model is called to complete gesture recognition.
The gesture recognition system of the invention is also characterized in that the signal excitation unit of the electrical impedance information acquisition module adopts two injection modes with adjustable frequency, comprising: swept frequency injection and fixed frequency injection, and the frequency range is between 0.1khz and 100 khz.
And the number and the placement position of the electrodes of the first acquisition electrode are subjected to feedback adjustment according to the gesture recognition result, wherein the feedback adjustment is to compare the gesture recognition result with the fused data by using a feature selection method to obtain updated data sub-features, and obtain better number and placement position of the electrodes after feature mapping.
The electrodes of the first collecting electrode, which are injected with the excitation signal, can simultaneously acquire corresponding electrical signals by the first measuring unit.
The second measurement unit adopts a differential circuit to collect electric signals during muscle activity, and the second collecting electrodes are distributed on the surfaces of muscle groups at the joints of the arms.
On one hand, the data fusion module sets a threshold value according to the change range of the electromyographic signals so as to determine the electromyographic information of the active segment moment as a part of fused data; on the other hand, the electromyographic signals are down-sampled to the frequency segment of the impedance distribution signals, and the impedance distribution signals except the moment of the active segment of the electromyographic information are used as the other part of the fused data, so that the frequency matching is completed.
Compared with the prior art, the invention has the beneficial effects that:
1. according to the two biological information acquisition modules provided by the invention, on one hand, impedance distribution data at the wrist is acquired by using an electrical impedance imaging technology, and on the other hand, biological myoelectric information of arm muscles when different gestures are made is acquired by using a differential circuit, so that impedance information which is more sensitive to static gestures or myoelectric information which is more sensitive to dynamic gestures can be acquired simultaneously, and thus the diversity of gesture recognition is improved.
2. The data fusion module use state matching method provided by the invention combines the advantages of the two biological information in dynamic gesture recognition and static gesture recognition respectively, so that the obtained data has better representation capability, the static and dynamic gesture recognition is simultaneously satisfied, and the robustness of gesture recognition is improved.
3. The machine learning module provided by the invention uses two different learning models, namely a traditional machine learning model and a deep learning model, to perform gesture recognition, and can call different pre-training models in real time according to a threshold activity detection method, so that the accuracy and stability of gesture recognition are improved.
Drawings
FIG. 1 is a block diagram of a gesture recognition system according to the present invention;
FIG. 2 is a schematic block diagram of an electrical impedance measurement circuit of the present invention;
FIG. 3 is a schematic block diagram of the electromyography measurement circuit of the present invention;
FIG. 4 is a timing diagram of the synchronous acquisition of electrical impedance and electromyographic signals of the present invention;
FIG. 5 is a flow chart of data fusion according to the present invention;
FIG. 6 is a diagram of the recognition result of the present invention.
Detailed Description
In this embodiment, in order to improve robustness and accuracy of the gesture recognition method, a gesture recognition system that integrates bioelectrical impedance information and bioelectrical muscle information is designed, as shown in fig. 1, and includes: the system comprises a power supply module, an electrical impedance information acquisition module, a biological myoelectric signal acquisition module, a microcontroller module, a wireless transmission module and an upper computer; wherein, the host computer includes: the system comprises a data fusion module and a machine learning classification module;
the electrical impedance information acquisition module consists of a signal excitation unit, a first measurement unit and a first acquisition electrode;
the first acquisition electrode is arranged on the wrist according to an array structure, and the signal excitation unit injects excitation signals into the first acquisition electrode under the control of the microcontroller module; the first measuring unit acquires an impedance distribution signal on the first collecting electrode under the control of the microcontroller module and sends the impedance distribution signal to the upper computer through the wireless transmission module;
in specific implementation, the signal excitation unit of the electrical impedance information acquisition module adopts two injection modes with adjustable frequency, including: swept frequency injection and fixed frequency injection, and the frequency range is between 0.1khz and 100 khz. With reference to fig. 2, the electrical circuit of the electrical impedance information acquisition module comprises: the device comprises a microcontroller, a signal generator, a voltage-controlled constant current source, a multi-way switch, a collecting electrode, an amplifying and filtering circuit and an A/D collector. In the data acquisition process, a signal generator generates a voltage excitation signal with variable frequency and waveform, in the embodiment, a sine wave frequency of 40khz is used, the voltage-controlled constant current source converts the voltage excitation signal into an excitation current signal, the excitation current signal is injected into the wrist through different acquisition electrodes in sequence under the control of a multi-way switch (in the embodiment, 8-way analog switches are adopted), meanwhile, a response electric signal is measured, and the impedance signal of the wrist is obtained through signal amplification, filtering and A/D acquisition.
In this embodiment, the number and the placement position of the electrodes of the first collecting electrode are feedback-adjusted according to the result of the gesture recognition, the feedback adjustment is to compare the gesture recognition result with the fused data by using a feature selection method, a recursive feature elimination method is adopted, a training model is used for performing multiple rounds of training, after each round of training, the features of a plurality of weight coefficients are removed, next round of training is performed based on a new fused data subset, updated data sub-features are finally obtained, and better number and placement position of the electrodes are obtained after feature mapping.
It is to be noted that the electrodes of the first collecting electrode, which are injected with excitation signals, are simultaneously able to acquire corresponding electrical signals by the first measuring unit.
The principle of gesture recognition by using electromyographic signals is that when different gesture actions are performed, different stretching or contraction states of arm muscle groups can generate different action potentials, and the action potentials are collected, analyzed and processed to obtain corresponding gestures. The electromyographic signal is a bioelectrical current signal generated by contraction or stretching of arm muscles when different gestures are made, is a time-action electric sequence, and therefore can provide more useful information for dynamic gesture recognition.
The biological electromyographic signal acquisition module consists of a second acquisition electrode and a second measurement unit;
the second collecting electrode is arranged on the arm according to an array structure, and the second measuring unit acquires the electromyographic signals on the second collecting electrode under the control of the microcontroller module and sends the electromyographic signals to the upper computer through the wireless transmission module; as shown in fig. 3, the second measurement unit adopts a differential circuit to collect electric signals during muscle activity, noise is filtered through a filter amplification circuit, and then the electric signals are collected through A/D to obtain the electromyographic signals with the frequency of 1khz in the embodiment. And the second acquisition electrodes are distributed on the surface of the muscle group at the arm joint.
The gesture recognition based on the electrical impedance imaging technology is to recognize wrist impedance distribution states under different gestures by acquiring electrical impedance information inside a wrist during hand movement and then filtering, preprocessing data and the like. Since the impedance information reflects the distribution of the impedance inside the wrist in a certain gesture state, more useful information can be provided for static gesture recognition.
With reference to fig. 4, the timing sequence setting for synchronously acquiring the electrical impedance signal and the electromyographic signal comprises two switch square pulse signals of sampling start and sampling stop, and simultaneously, a 40khz sine wave excitation current is continuously applied for measuring the electrical impedance of the wrist; electromyographic signal data are stored in a cache according to the sampling frequency of 1khz, electrical impedance signal data are stored in the cache according to the sampling frequency of 100hz, and the two types of signal data are sent to an upper computer PC word by word through a Bluetooth wireless transmission module.
The data fusion processing module is used for carrying out frequency matching on the received impedance distribution signal and the received electromyographic signal so as to obtain fused data; on one hand, the data fusion module sets a threshold value according to the change range of the electromyographic signals so as to determine the electromyographic information of the active segment moment as a part of fused data; on the other hand, the electromyographic signals are down-sampled to the frequency section of the impedance distribution signals, the impedance distribution signals except the moment of the active section of the electromyographic information are used as the other part of the fused data, specifically, the two kinds of data with different frequencies are subjected to frequency adjustment by methods such as up-sampling, down-sampling and interpolation sampling, so that the frequencies of the two kinds of data are kept consistent on the fused data, and the frequency matching is finished.
The machine learning classification module comprises a deep cycle neural network model of the first model and a traditional machine learning model of the second model;
the deep circulation neural network model performs feature extraction and training on the fused data to obtain a trained dynamic model;
the traditional machine learning model performs feature extraction and training on the fused data to obtain a trained static model; as shown in fig. 5, firstly, myoelectric information of muscle activity segments is determined according to threshold detection, and the myoelectric information is sent to a first model for training and learning, wherein the first model adopts a deep cyclic neural network with two layers of BiLSTM + Softmax (Added Fisher criterion), and can realize the automatic feature extraction and recognition classification functions of data from end to end. Then, when the inactive segment, namely the gesture state is kept, electromyographic information is down-sampled to impedance information frequency, namely 100hz, frequency matching fusion is carried out, then data with consistent frequency are spliced together and sent to a model II for training and learning, wherein the model II adopts a traditional machine learning algorithm such as an SVM (support vector machine), a Random Forest (RF) and the like.
And the machine learning classification module judges whether the myoelectric information detection threshold is met or not according to the acquired real-time fusion data, if so, the dynamic model is called to finish gesture recognition, and otherwise, the static model is called to finish gesture recognition.
Specific examples are as follows:
step 1, data acquisition, comprising: a total of 5 subjects were aged between 23 and 26 years. The initial setting has 8 static gestures and 7 dynamic gestures, for a total of 15 actions. Firstly, gesture actions of each person are collected in sequence, wherein the number of times of repetition of each action is 10, then corresponding impedance data and myoelectric data are obtained after the first round of all collection is completed, then the second round of collection is carried out, and the method is still repeated for five times to obtain an experimental data set. And one round of data is selected as a test set, and the other rounds of data are selected as a training set, so that the effect of five-round cross validation is achieved.
And 2. Step 2. Model training: the first model is responsible for training dynamic gesture data, a deep Recurrent Neural Network (RNN) model is adopted to train a data set, and the trained model is used as a second pre-training model. And the second model is responsible for training static gesture data, the traditional machine learning model Support Vector Machine (SVM), random Forest (RF), logistic Regression (LR) and Neural Network (NN) models are adopted for offline training, then the experimental effects of the 4 methods are compared, and the optimal model is selected as the first pre-training model.
And 3. Step 3. Real-time identification: and acquiring biological signals (myoelectricity + impedance) when the gesture is made in real time, calling the model to identify the biological signals when the myoelectricity information detection threshold is met, otherwise calling the model II to identify, and displaying the identification result on a screen. As shown in fig. 6, the recognition result thereof is displayed on the screen by the PC. Electrodes of the fusion system are attached to the wrist and the arm, and the recognition mode is opened, so that the system can accurately display the result on a PC through a schematic diagram in real time.
In conclusion, the method combines the advantages of two recognition modes of the electromyographic signals and the electrical impedance information, the method for recognizing the gesture based on the electromyographic signals not only gets rid of the limitation of the external use environment, but also has simple and convenient measurement, and the gesture recognition based on the electrical impedance imaging technology can provide more useful information for static gesture recognition, so that the two biological signals are fused together, various static gestures can be recognized accurately and quickly, meanwhile, the dynamic gestures can be recognized accurately, and the diversity and the accuracy of the gesture recognition method are improved; moreover, the signal is fused, so that the robustness of the signal is greatly improved, and the method can be better suitable for various environments.
Claims (6)
1. A gesture recognition system fusing bioelectrical impedance information and bioelectrical electrical information is characterized by comprising: the system comprises a power supply module, an electrical impedance information acquisition module, a biological myoelectric signal acquisition module, a microcontroller module, a wireless transmission module and an upper computer; the host computer includes: the system comprises a data fusion module and a machine learning classification module;
the electrical impedance information acquisition module consists of a signal excitation unit, a first measurement unit and a first acquisition electrode;
the first acquisition electrode is arranged on the wrist according to an array structure, and the signal excitation unit injects excitation signals into the first acquisition electrode under the control of the microcontroller module; the first measuring unit acquires an impedance distribution signal on the first collecting electrode under the control of the microcontroller module and sends the impedance distribution signal to the upper computer through the wireless transmission module;
the biological electromyographic signal acquisition module consists of a second acquisition electrode and a second measurement unit;
the second collecting electrode is arranged on the arm according to an array structure, and the second measuring unit acquires the electromyographic signals on the second collecting electrode under the control of the microcontroller module and sends the electromyographic signals to the upper computer through the wireless transmission module;
the data fusion processing module is used for carrying out frequency matching on the received impedance distribution signal and the received electromyographic signal so as to obtain fused data;
the machine learning classification module comprises a deep cycle neural network model and a traditional machine learning model;
the deep circulation neural network model performs feature extraction and training on the fused data to obtain a trained dynamic model;
performing feature extraction and training on the fused data of the machine learning model to obtain a trained static model;
and the machine learning classification module judges whether the myoelectric information detection threshold is met or not according to the acquired real-time fusion data, if so, the dynamic model is called to finish gesture recognition, and otherwise, the static model is called to finish gesture recognition.
2. The gesture recognition system of claim 1: the electrical impedance information acquisition module is characterized in that a signal excitation unit of the electrical impedance information acquisition module adopts two injection modes with adjustable frequency, and the injection modes comprise: swept frequency injection and fixed frequency injection, and the frequency range is between 0.1khz and 100 khz.
3. The gesture recognition system of claim 1: the method is characterized in that the number and the placement positions of the electrodes of the first collecting electrode are subjected to feedback adjustment according to a gesture recognition result, the feedback adjustment is to compare the gesture recognition result with the fused data by using a feature selection method to obtain updated data sub-features, and the optimal number and placement positions of the electrodes are obtained after feature mapping.
4. The gesture recognition system of claim 1: the electrode of the excitation signal injected into the first collecting electrode can simultaneously obtain a corresponding electric signal by the first measuring unit.
5. The gesture recognition system of claim 1: the second measurement unit is characterized in that the second measurement unit adopts a differential circuit to collect electric signals generated during muscle activity, and the second collecting electrodes are distributed on the surfaces of muscle groups at joints of arms.
6. The gesture recognition system of claim 1: the method is characterized in that on one hand, a threshold value is set by the data fusion module according to the change range of the electromyographic signals, so that the electromyographic information of the active segment moment is determined to be used as a part of fused data; on the other hand, the electromyographic signals are down-sampled to the frequency section of the impedance distribution signals, and the impedance distribution signals except the active section time of the electromyographic signals are used as the other part of the fused data, so that the frequency matching is completed.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010390176.8A CN111553307B (en) | 2020-05-08 | 2020-05-08 | Gesture recognition system fusing bioelectrical impedance information and myoelectric information |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010390176.8A CN111553307B (en) | 2020-05-08 | 2020-05-08 | Gesture recognition system fusing bioelectrical impedance information and myoelectric information |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111553307A CN111553307A (en) | 2020-08-18 |
| CN111553307B true CN111553307B (en) | 2023-03-24 |
Family
ID=72000884
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010390176.8A Active CN111553307B (en) | 2020-05-08 | 2020-05-08 | Gesture recognition system fusing bioelectrical impedance information and myoelectric information |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111553307B (en) |
Families Citing this family (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112263235A (en) * | 2020-09-30 | 2021-01-26 | 浙江大学 | Gesture recognition system based on hybrid electrical tomography technology |
| CN112223242B (en) * | 2020-10-09 | 2022-03-01 | 河南理工大学 | Force feedback device for teleoperation system based on skin stimulation |
| CN112380943B (en) * | 2020-11-06 | 2022-12-06 | 北京航空航天大学 | Multi-position limb motion capture method based on electrical impedance |
| CN114515146B (en) * | 2020-11-17 | 2024-03-22 | 北京机械设备研究所 | Intelligent gesture recognition method and system based on electrical measurement |
| CN112966662A (en) * | 2021-03-31 | 2021-06-15 | 安徽大学 | Short-range capacitive dynamic gesture recognition system and method |
| CN113312994A (en) * | 2021-05-18 | 2021-08-27 | 中国科学院深圳先进技术研究院 | Gesture classification recognition method and application thereof |
| CN115114962A (en) * | 2022-07-19 | 2022-09-27 | 歌尔股份有限公司 | Control method and device based on surface electromyogram signal and wearable device |
| CN116304781B (en) * | 2022-09-09 | 2024-02-02 | 武汉中数医疗科技有限公司 | Thyroid sampling data identification method based on cyclic neural network |
| CN117238037B (en) * | 2023-11-13 | 2024-03-29 | 中国科学技术大学 | Dynamic action recognition method, device, equipment and storage medium |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109009718A (en) * | 2018-08-10 | 2018-12-18 | 中国科学院合肥物质科学研究院 | A method of based on electrical impedance technology combination gesture control wheelchair |
| CN110443309A (en) * | 2019-08-07 | 2019-11-12 | 浙江大学 | A kind of electromyography signal gesture identification method of combination cross-module state association relation model |
| CN110658915A (en) * | 2019-07-24 | 2020-01-07 | 浙江工业大学 | Electromyographic signal gesture recognition method based on double-current network |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US10061389B2 (en) * | 2014-06-03 | 2018-08-28 | Beijing TransBorder Information Technology Co., Ltd. | Gesture recognition system and gesture recognition method |
-
2020
- 2020-05-08 CN CN202010390176.8A patent/CN111553307B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109009718A (en) * | 2018-08-10 | 2018-12-18 | 中国科学院合肥物质科学研究院 | A method of based on electrical impedance technology combination gesture control wheelchair |
| CN110658915A (en) * | 2019-07-24 | 2020-01-07 | 浙江工业大学 | Electromyographic signal gesture recognition method based on double-current network |
| CN110443309A (en) * | 2019-08-07 | 2019-11-12 | 浙江大学 | A kind of electromyography signal gesture identification method of combination cross-module state association relation model |
Non-Patent Citations (1)
| Title |
|---|
| 一种基于融合多传感器信息的手语手势识别方法;阳平等;《航天医学与医学工程》;20120815(第04期);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111553307A (en) | 2020-08-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111553307B (en) | Gesture recognition system fusing bioelectrical impedance information and myoelectric information | |
| CN102499797B (en) | Artificial limb control method and system | |
| CN113288183B (en) | Silent voice recognition method based on facial neck surface myoelectricity | |
| CN105769173B (en) | A kind of cardioelectric monitor system with electrocardiosignal noise removal function | |
| CN101776981B (en) | Method for controlling mouse by jointing brain electricity and myoelectricity | |
| CN202288542U (en) | Artificial limb control device | |
| CN103699226A (en) | Tri-modal serial brain-computer interface method based on multi-information fusion | |
| CN113205074B (en) | Gesture recognition method fusing multi-mode signals of myoelectricity and micro-inertia measurement unit | |
| He et al. | Spatial information enhances myoelectric control performance with only two channels | |
| CN111317600A (en) | Artificial limb control method, device, system, equipment and storage medium | |
| CN107480716A (en) | A kind of combination EOG and video pan signal recognition method and system | |
| CN112732092A (en) | Surface electromyogram signal identification method based on double-view multi-scale convolution neural network | |
| WO2018103495A1 (en) | Eog-based method and apparatus for asynchronous character input | |
| Liu et al. | A cnn-transformer hybrid recognition approach for semg-based dynamic gesture prediction | |
| Park et al. | Korean finger number gesture recognition based on CNN using surface electromyography signals | |
| WO2022242133A1 (en) | Gesture classification and recognition method and application thereof | |
| CN112405539B (en) | Robot natural control method based on electromyographic signals and electroencephalogram error potentials | |
| CN110751060B (en) | Portable motion mode real-time identification system based on multi-source signals | |
| CN112101298A (en) | Gesture recognition system and method based on muscle electrical impedance signals | |
| CN116910464A (en) | Myoelectric signal prosthetic hand control system and method | |
| KR20100056254A (en) | Method and device for deducting pinch force, method and device for discriminating muscle to deduct pinch force | |
| CN112932508B (en) | Finger activity recognition system based on arm electromyography network | |
| CN212647500U (en) | Gesture recognition system based on muscle electrical impedance signal | |
| CN107765890A (en) | A kind of write content detection system and method based on acceleration transducer | |
| Chen et al. | Recognition of american sign language gestures based on electromyogram (emg) signal with xgboost machine learning |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |