CN105608432A - Instantaneous myoelectricity image based gesture identification method - Google Patents
Instantaneous myoelectricity image based gesture identification method Download PDFInfo
- Publication number
- CN105608432A CN105608432A CN201510973702.2A CN201510973702A CN105608432A CN 105608432 A CN105608432 A CN 105608432A CN 201510973702 A CN201510973702 A CN 201510973702A CN 105608432 A CN105608432 A CN 105608432A
- Authority
- CN
- China
- Prior art keywords
- instantaneous
- electromyographic signal
- gesture
- myoelectricity
- image
- 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.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/02—Preprocessing
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Artificial Intelligence (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Signal Processing (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Measurement And Recording Of Electrical Phenomena And Electrical Characteristics Of The Living Body (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses an instantaneous myoelectricity image based gesture identification method. During a training stage, firstly instantaneous myoelectricity signals acquired by array electrodes are preprocessed and arranged according to electrode positions to form an instantaneous myoelectricity image; and secondly an image classifier such as a deep convolutional neural network is trained by using the instantaneous myoelectricity image and a gesture tag corresponding to the instantaneous myoelectricity image to obtain network model parameters. During a test stage, firstly to-be-identified instantaneous myoelectricity signals acquired by the array electrodes are preprocessed and arranged according to the electrode positions to form the instantaneous myoelectricity image; and secondly the trained model parameters are substituted into the classifier to identify the gesture tag corresponding to the instantaneous myoelectricity signals. According to the instantaneous myoelectricity image based gesture identification method, a gesture can be quickly and accurately identified based on the instantaneous myoelectricity image and an image classification method. No literature for gesture identification by the instantaneous myoelectricity signals exists at home and abroad yet.
Description
Technical field
The invention belongs to the computer field that combines with bio signal, be specially and use with degree of depth convolution nerve netNetwork is the Image Classifier of representative, to the instantaneous electromyogram extracting from instantaneous electromyographic signal as corresponding handGesture is identified.
Background technology
Along with computer vision, touch the fast development of the new technologies such as mutual, perception calculating, perception user circleFace (perceptualuserinterface, PUI) becomes one of research emphasis of field of human-computer interaction.Perception user interface be the high interaction that is prototype of the interacting activity between a kind of person to person and people and real world,Multichannel user interface, its target is to make man-machine interaction become consistent next with people with the interaction of real worldReach directly perceived, naturally mutual boundary. In order to make computer can judge better and understand the mutual of the mankindIntention, " raw, mechanical, electrical integrated " is one of important trend of following man-machine interactive development, by specificSensing equipment by the cognition of organism or perceptual signal (as electromyographic signal) digitlization, and with other perceptionOr the signal of cognitive passage carries out integrated fusion, nature, complete various man-machine interaction tasks synergistically.
Research in the past is conventionally used machine learning method to carry out the gesture identification based on electromyographic signal, and given oneSection electromyographic signal, is used the grader training to classify, and obtains gesture label; This grader is to useThe electromyographic signal training gathering in advance obtains. Instantaneous myoelectricity letter is it is generally acknowledged in correlative study for a long timeNumber be full of random noise, thereby be not suitable for being directly used in identification gesture. Conventional method is generally large with a segment lengthBe unit in the electromyographic signal that equals 150 milliseconds, extract various features and form characteristic vector, re-use supportThe graders such as vector machine are classified to characteristic vector, thereby identify one section of gesture that electromyographic signal is corresponding.
Along with the appearance of array electromyographic electrode, the electric potential field that muscle activity forms at skin surface is in each momentSpatial distribution can go on record. The present invention uses array electrode to gather instantaneous electromyographic signal, according to electricityInstantaneous electromyographic signal is converted into electromyogram picture by the spatial arrangement of the utmost point, and use Image Classifier to electromyogram pictureClassify, thereby identify the represented gesture of instantaneous electromyographic signal. Electromyogram similarly is one of above-mentioned electric potential fieldPlant spatial sampling, contained muscle activity global characteristics at a time, (zero observation time delay) fastIdentify exactly gesture. There is no document both at home and abroad uses instantaneous electromyographic signal to carry out gesture identification.
Degree of depth convolutional neural networks is a kind of feedforward neural network, and its artificial neuron can respond a partSurrounding unit in coverage. This model is generally used for Images Classification.
Summary of the invention
The object of the invention is to for the deficiencies in the prior art, a kind of hand based on instantaneous electromyographic signal is providedGesture recognition methods.
The object of the invention is to be achieved through the following technical solutions: a kind of hand based on instantaneous electromyographic signalGesture recognition methods, the method comprises the following steps:
(1) gather instantaneous electromyographic signal by array electrode;
(2) the instantaneous electromyographic signal of pair array electrode collection is carried out pretreatment, generates instantaneous electromyogram picture, toolBody comprises following steps:
(2.1) the instantaneous electromyographic signal linear transformation to 0 in general ± 2.5mV interval is in 255 intervals;
(2.2) instantaneous electromyographic signal is arranged into instantaneous electromyogram picture according to the locus of electrode;
(3) train instantaneous myoelectricity Image Classifier, obtain the model parameter of grader;
(4) use the instantaneous myoelectricity Image Classifier identification gesture of training in step (3), specifically comprise asLower sub-step:
(4.1) the instantaneous electromyographic signal to be identified of pair array electrode collection is carried out the pre-place in step (2)Reason, obtains pending instantaneous electromyogram picture;
(4.2) model parameter training in step (3) being obtained is brought in instantaneous myoelectricity Image ClassifierIdentify pending gesture corresponding to instantaneous electromyogram picture.
Further, in step (2.2), by the instantaneous electromyographic signal value in each moment according to array electrodeLocus is arranged into matrix, forms instantaneous electromyogram picture; Instantaneous electromyogram similarly is single pass gray level image,Its height and width equate with line number and the columns of array electrode respectively.
Further, employing convolution degree of depth neutral net in step (3) (DeepConvolutionalNetwork,ConvNets) as instantaneous myoelectricity Image Classifier, specifically comprise following sub-step:
(3.1) use the VGGNet initialization of pre-training on coloured image to be used for from instantaneous electromyogram picture pointThe network model of class gesture;
(3.2) pretreatment of many people's that use gathers in advance instantaneous electromyographic signal in step (2), makesThe instantaneous electromyogram picture obtaining with pretreatment and corresponding gesture label training network model thereof, obtain network mouldShape parameter.
Further, in step (3.1), use colored natural image to train for the animal of classifying, car in advanceWait the VGGNet of 16 layers of natural forms, then use front 4 convolutional layers of this network to initialize with instantaneousIt is what input that electromyogram looks like, for the network model of the gesture of classifying; The wherein initialization side of first convolutional layerMethod is: by the weight matrix summation of RGB3 passage of the ground floor of the VGGNet of training in advance, and by resultAs the weight matrix of the ground floor of new network model.
The invention has the beneficial effects as follows: in the training stage, the first step, the instantaneous myoelectricity that pair array electrode gathersSignal carries out pretreatment, obtains instantaneous electromyogram picture; Second step, uses instantaneous electromyogram picture and correspondingGesture label training image grader, for example degree of depth convolutional neural networks, obtains network model parameter. SurveyingThe examination stage, the first step, the instantaneous electromyographic signal to be identified of pair array electrode collection is carried out pretreatment, obtainsInstantaneous electromyogram picture; Second step, is brought into the network model parameter training in network model and identifies everyGesture corresponding to instantaneous electromyogram picture. The present invention is based on instantaneous electromyogram picture and image classification method, can be fastSpeed is identified gesture exactly. Compared to conventional method, the present invention, can be with zero after user makes gestureObservation time delay is identified gesture rapidly.
Brief description of the drawings
Fig. 1 is the method for the invention flow chart;
Fig. 2 is the gesture collection that comprises 8 kinds of hand attitudes that the present invention tests use;
Fig. 3 is electrode patch schematic diagram used in the present invention;
Fig. 4 is degree of depth convolutional neural networks structure used in the present invention.
Detailed description of the invention
Below in conjunction with the drawings and specific embodiments, the present invention is described in further detail.
As shown in Figure 1, a kind of gesture identification method based on instantaneous electromyographic signal provided by the invention, comprisesFollowing steps:
(1) use array electrode to gather instantaneous electromyographic signal, below provide a kind of instantaneous electromyographic signal that gathersPreferred embodiment:
(1.1) tested sitting posture, right arm is horizontal forward to be stretched, and the centre of the palm is towards front-left.
(1.2) use alcohol to clean the skin of right forearm electrode to be pasted, treat that it dries.
(1.3) paste electrode (dot represents array electrode) according to Fig. 3, circle electrode (reference electrode) pastesMiddle outside wrist.
(1.4) connect collecting device, gather instantaneous electromyographic signal.
(2) the instantaneous electromyographic signal of pair array electrode collection is carried out pretreatment, generates instantaneous electromyogram picture, toolBody comprises following steps:
(2.1) the interval electromyographic signal linear transformation to 0 of general ± 2.5mV is in 255 intervals;
(2.2) the instantaneous electromyographic signal value in each moment is arranged into matrix according to the locus of array electrode,Form instantaneous electromyogram picture. Instantaneous electromyogram similarly is single pass gray level image, its height and width respectively withThe line number of array electrode and columns equate.
(3) train instantaneous myoelectricity Image Classifier, below taking convolution degree of depth neutral net as example:
Degree of depth study structure is a kind of multilayer perceptron containing many hidden layers. Degree of depth study is by combination low-level featureForm more abstract high level and represent attribute classification or feature, represent with the distributed nature of finding data. ThisInvention degree of depth study structure is from the amendment of VGGNet network structure, and this amendment makes the VGGNet canTo identify gesture corresponding to instantaneous electromyogram picture, this network structure comprises following Internet:
A) convolutional layer (Conv): in convolutional neural networks, every layer of convolutional layer is made up of some convolution unit, eachThe parameter of convolution unit all obtains by back-propagation algorithm optimization. The object of convolution algorithm is that extraction is defeatedThe different characteristic entering, ground floor convolutional layer may can only extract some rudimentary features as edge, lines and angleEtc. level, more multi-layered network can be from low-level features the more complicated feature of iterative extraction. The present invention uses 3x3Convolution mask.
B) ReLU activation primitive: ReLU can make a part neuronicly be output as 0, thereby makes network moreSparse, and reduced the relation of interdependence of parameter, reduce the over-fitting of network in training process.
C) full articulamentum (Fc): the hidden layer that the full articulamentum in CNN model is traditional neutral net. LastIndividual full articulamentum, i.e. output layer, relevant to concrete classification task, its output dimension equals hand to be identifiedGesture number.
D) Dropout layer: Dropout is mainly used to solve the mistake causing because of quantity of parameters in degree of depth studyFitting problems. Its main thought is: in training process, throw away at random a certain proportion of neuron and (compriseTheir connection), can stop so interneuronal undue common adaptation. Dropout ratio of the present invention is0.5。
E) Softmax layer: Softmax layer is softmax model to be applied to last training result returnReturn the layer of operation. Softmax model is that Logic Regression Models (is categorical measure to be sorted in many classification problemsBe greater than 2) on popularization.
Concrete network structure as shown in Figure 4. The wherein output dimension of this layer of numeral after Conv and Fc.
This step specifically comprises following sub-step:
(3.1) use colored natural image to train for 16 of the natural forms such as animal, vehicle of classifying in advanceThe VGGNet network of layer, then use front 4 convolutional layers of this network initialize taking the electromyogram of gray scale look like asInput, for the network model of the gesture of classifying. Wherein the initial method of first convolutional layer is: will be pre-The power summation of RGB3 passage of the ground floor of the VGGNet of training, and using result as new network modelThe weight matrix of ground floor.
(3.2) use the many people's that gather in advance instantaneous electromyographic signal to carry out the pretreatment in step (2),Use pretreatment to obtain instantaneous electromyogram picture and corresponding gesture label training network model thereof, obtain network mouldShape parameter.
(4) use the instantaneous myoelectricity Image Classifier identification gesture of training in step (3), specifically comprise asLower sub-step:
(4.1) the instantaneous electromyographic signal to be identified of pair array electrode collection is carried out the pre-place in step (2)Reason, obtains instantaneous electromyogram picture;
(4.2) network model parameter training in step (3) being obtained is brought into identification in network model and treatsGesture corresponding to instantaneous electromyogram picture of processing.
Embodiment
The present invention is based on instantaneous electromyographic signal identification gesture, according to Fig. 1, mainly comprise two parts: off-line instructionPractice part and ONLINE RECOGNITION part.
Off-line training part comprises:
A. paste electrode according to Fig. 3 for tester.
B. gather 8 kinds of gestures in Fig. 2. Every kind of gesture gathers 10 times, continues to have an effect 3 seconds at every turn, connectsBetween continuous having an effect for twice, have a rest 7 seconds.
C. the instantaneous electromyographic signal of pair array electrode collection is carried out pretreatment, the interval electromyographic signal of general ± 2.5mVLinear transformation to 0 is in 255 intervals. By the instantaneous electromyographic signal value in each moment according to the sky of array electrodeBetween position be arranged into matrix, form instantaneous electromyogram picture.
D. use the pre-training of colored natural image for 16 layers of the natural forms such as animal, vehicle of classifyingVGGNet network, then use front 4 convolutional layers of this network initialize taking the instantaneous electromyogram of gray scale look like asInput, for the network model of the gesture of classifying. Instantaneous electromyogram picture and the gesture label thereof of step c are pressedBe training set and checking collection according to the ratio cut partition of 7:3. Use training set to train this network mould by iteration optimizationType, and test on checking collection, until the accuracy rate on checking collection stops iteration while no longer rising.
ONLINE RECOGNITION part comprises:
A. paste electrode according to the step a of off-line training.
B. gather instantaneous electromyographic signal, generate instantaneous myoelectricity image sequence according to the step c of off-line training.
C. use every gesture corresponding to instantaneous electromyogram picture of network model identification training.
D. (optionally) agrees to rule ballot to the gesture label in nearest 150 milliseconds according to majority(MajorityVote), poll soprano, as this section of gesture that electromyographic signal is corresponding, knows thereby improve gestureOther accuracy rate.
Claims (4)
1. the gesture identification method based on instantaneous electromyographic signal, is characterized in that, comprises the following steps:
(1) gather instantaneous electromyographic signal by array electrode;
(2) the instantaneous electromyographic signal of pair array electrode collection is carried out pretreatment, generates instantaneous electromyogram picture, toolBody comprises following steps:
(2.1) the instantaneous electromyographic signal linear transformation to 0 in general ± 2.5mV interval is in 255 intervals;
(2.2) instantaneous electromyographic signal is arranged into instantaneous electromyogram picture according to the locus of electrode;
(3) train instantaneous myoelectricity Image Classifier, obtain the model parameter of grader;
(4) use the instantaneous myoelectricity Image Classifier identification gesture of training in step (3), specifically comprise asLower sub-step:
(4.1) the instantaneous electromyographic signal to be identified of pair array electrode collection is carried out the pre-place in step (2)Reason, obtains pending instantaneous electromyogram picture;
(4.2) model parameter training in step (3) being obtained is brought in instantaneous myoelectricity Image ClassifierIdentify pending gesture corresponding to instantaneous electromyogram picture.
2. a kind of gesture identification method based on instantaneous electromyographic signal according to claim 1, is characterized in that,In step (2.2), the instantaneous electromyographic signal value in each moment is arranged into according to the locus of array electrodeMatrix, forms instantaneous electromyogram picture; Instantaneous electromyogram similarly is single pass gray level image, its height and widthEquate with line number and the columns of array electrode respectively.
3. a kind of gesture identification method based on instantaneous electromyographic signal according to claim 1, is characterized in that,In step (3), adopt convolution degree of depth neutral net (DeepConvolutionalNetwork, ConvNets)As instantaneous myoelectricity Image Classifier, specifically comprise following sub-step:
(3.1) use the VGGNet initialization of pre-training on coloured image to be used for from instantaneous electromyogram picture pointThe network model of class gesture;
(3.2) pretreatment of many people's that use gathers in advance instantaneous electromyographic signal in step (2), makesThe instantaneous electromyogram picture obtaining with pretreatment and corresponding gesture label training network model thereof, obtain network mouldShape parameter.
4. a kind of gesture identification method based on instantaneous electromyographic signal according to claim 3, is characterized in that,In step (3.1), use the pre-training of colored natural image for the natural forms such as animal, vehicle of classifyingThe VGGNet of 16 layers, then uses front 4 convolutional layers initialization of this network to look like as input taking instantaneous electromyogram, for the network model of the gesture of classifying; Wherein the initial method of first convolutional layer is: will train in advanceThe weight matrix summation of RGB3 passage of ground floor of VGGNet, and using result as new network mouldThe weight matrix of the ground floor of type.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510973702.2A CN105608432B (en) | 2015-12-21 | 2015-12-21 | A kind of gesture identification method based on instantaneous myoelectricity image |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510973702.2A CN105608432B (en) | 2015-12-21 | 2015-12-21 | A kind of gesture identification method based on instantaneous myoelectricity image |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN105608432A true CN105608432A (en) | 2016-05-25 |
| CN105608432B CN105608432B (en) | 2019-02-22 |
Family
ID=55988358
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510973702.2A Active CN105608432B (en) | 2015-12-21 | 2015-12-21 | A kind of gesture identification method based on instantaneous myoelectricity image |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105608432B (en) |
Cited By (17)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106228116A (en) * | 2016-07-13 | 2016-12-14 | 华东理工大学 | A kind of grinding trembling recognition methods based on electric current instantaneous nonlinearity index |
| CN106236336A (en) * | 2016-08-15 | 2016-12-21 | 中国科学院重庆绿色智能技术研究院 | A kind of myoelectric limb gesture and dynamics control method |
| CN106293057A (en) * | 2016-07-20 | 2017-01-04 | 西安中科比奇创新科技有限责任公司 | Gesture identification method based on BP neutral net |
| CN106980367A (en) * | 2017-02-27 | 2017-07-25 | 浙江工业大学 | A kind of gesture identification method based on myoelectricity topographic map |
| CN107480697A (en) * | 2017-07-12 | 2017-12-15 | 中国科学院计算技术研究所 | A kind of myoelectricity gesture identification method and system |
| CN107657233A (en) * | 2017-09-28 | 2018-02-02 | 东华大学 | Static sign language real-time identification method based on modified single multi-target detection device |
| CN107885324A (en) * | 2017-09-28 | 2018-04-06 | 江南大学 | A kind of man-machine interaction method based on convolutional neural networks |
| CN108345873A (en) * | 2018-03-22 | 2018-07-31 | 哈尔滨工业大学 | A kind of multiple degrees of freedom body motion information analytic method based on multilayer convolutional neural networks |
| CN108453736A (en) * | 2018-03-22 | 2018-08-28 | 哈尔滨工业大学 | A kind of multiple degrees of freedom synchronization myoelectricity control method based on deep learning |
| CN108491077A (en) * | 2018-03-19 | 2018-09-04 | 浙江大学 | A kind of surface electromyogram signal gesture identification method for convolutional neural networks of being divided and ruled based on multithread |
| CN109662710A (en) * | 2018-12-06 | 2019-04-23 | 杭州电子科技大学 | A kind of EMG Feature Extraction based on convolutional neural networks |
| CN110333783A (en) * | 2019-07-10 | 2019-10-15 | 中国科学技术大学 | A kind of unrelated gesture processing method and system for robust myoelectric control |
| CN111103976A (en) * | 2019-12-05 | 2020-05-05 | 深圳职业技术学院 | Gesture recognition method and device and electronic equipment |
| CN111898526A (en) * | 2020-07-29 | 2020-11-06 | 南京邮电大学 | Myoelectric gesture recognition method based on multi-stream convolution neural network |
| WO2022027822A1 (en) * | 2020-08-03 | 2022-02-10 | 南京邮电大学 | Electromyographic signal-based intelligent gesture action generation method |
| CN114515146A (en) * | 2020-11-17 | 2022-05-20 | 北京机械设备研究所 | Intelligent gesture recognition method and system based on electrical measurement |
| US11705748B2 (en) | 2018-02-28 | 2023-07-18 | Russell Wade Chan | Wearable gesture recognition device for medical screening and associated operation method and system |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101371804A (en) * | 2008-07-16 | 2009-02-25 | 哈尔滨工业大学 | On-line recognizing method of hand gesture mode established based on sEMG |
| US20130232095A1 (en) * | 2008-06-26 | 2013-09-05 | Microsoft Corporation | Recognizing finger gestures from forearm emg signals |
-
2015
- 2015-12-21 CN CN201510973702.2A patent/CN105608432B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20130232095A1 (en) * | 2008-06-26 | 2013-09-05 | Microsoft Corporation | Recognizing finger gestures from forearm emg signals |
| CN101371804A (en) * | 2008-07-16 | 2009-02-25 | 哈尔滨工业大学 | On-line recognizing method of hand gesture mode established based on sEMG |
Non-Patent Citations (3)
| Title |
|---|
| CHRISTOPH AMMA 等: "Advancing Muscle-Computer Interfaces with High-Density Electromyography", 《ACM CONFERENCE ON HUMAN FACTORS IN COMPUTING SYSTEMS》 * |
| MONICA ROJAS-MARTÍNEZ 等: "High-density surface EMG maps from upper-arm and forearm muscles", 《JOURNAL OF NEUROENGINEERING AND REHABILITATION》 * |
| 朱明星 等: "基于高密度肌电电势图的正常吞咽过程可视化研究", 《集成技术》 * |
Cited By (24)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106228116A (en) * | 2016-07-13 | 2016-12-14 | 华东理工大学 | A kind of grinding trembling recognition methods based on electric current instantaneous nonlinearity index |
| CN106293057A (en) * | 2016-07-20 | 2017-01-04 | 西安中科比奇创新科技有限责任公司 | Gesture identification method based on BP neutral net |
| CN106236336A (en) * | 2016-08-15 | 2016-12-21 | 中国科学院重庆绿色智能技术研究院 | A kind of myoelectric limb gesture and dynamics control method |
| CN106980367A (en) * | 2017-02-27 | 2017-07-25 | 浙江工业大学 | A kind of gesture identification method based on myoelectricity topographic map |
| CN106980367B (en) * | 2017-02-27 | 2020-08-18 | 浙江工业大学 | Gesture recognition method based on electromyogram |
| CN107480697B (en) * | 2017-07-12 | 2020-04-03 | 中国科学院计算技术研究所 | Myoelectric gesture recognition method and system |
| CN107480697A (en) * | 2017-07-12 | 2017-12-15 | 中国科学院计算技术研究所 | A kind of myoelectricity gesture identification method and system |
| CN107885324A (en) * | 2017-09-28 | 2018-04-06 | 江南大学 | A kind of man-machine interaction method based on convolutional neural networks |
| CN107657233A (en) * | 2017-09-28 | 2018-02-02 | 东华大学 | Static sign language real-time identification method based on modified single multi-target detection device |
| CN107885324B (en) * | 2017-09-28 | 2020-07-28 | 江南大学 | Human-computer interaction method based on convolutional neural network |
| US11705748B2 (en) | 2018-02-28 | 2023-07-18 | Russell Wade Chan | Wearable gesture recognition device for medical screening and associated operation method and system |
| CN108491077B (en) * | 2018-03-19 | 2020-06-16 | 浙江大学 | Surface electromyographic signal gesture recognition method based on multi-stream divide-and-conquer convolutional neural network |
| CN108491077A (en) * | 2018-03-19 | 2018-09-04 | 浙江大学 | A kind of surface electromyogram signal gesture identification method for convolutional neural networks of being divided and ruled based on multithread |
| CN108345873A (en) * | 2018-03-22 | 2018-07-31 | 哈尔滨工业大学 | A kind of multiple degrees of freedom body motion information analytic method based on multilayer convolutional neural networks |
| CN108453736A (en) * | 2018-03-22 | 2018-08-28 | 哈尔滨工业大学 | A kind of multiple degrees of freedom synchronization myoelectricity control method based on deep learning |
| CN109662710A (en) * | 2018-12-06 | 2019-04-23 | 杭州电子科技大学 | A kind of EMG Feature Extraction based on convolutional neural networks |
| CN110333783A (en) * | 2019-07-10 | 2019-10-15 | 中国科学技术大学 | A kind of unrelated gesture processing method and system for robust myoelectric control |
| CN111103976A (en) * | 2019-12-05 | 2020-05-05 | 深圳职业技术学院 | Gesture recognition method and device and electronic equipment |
| CN111103976B (en) * | 2019-12-05 | 2023-05-02 | 深圳职业技术学院 | Gesture recognition method and device and electronic equipment |
| CN111898526A (en) * | 2020-07-29 | 2020-11-06 | 南京邮电大学 | Myoelectric gesture recognition method based on multi-stream convolution neural network |
| CN111898526B (en) * | 2020-07-29 | 2022-07-22 | 南京邮电大学 | Myoelectric gesture recognition method based on multi-stream convolution neural network |
| WO2022027822A1 (en) * | 2020-08-03 | 2022-02-10 | 南京邮电大学 | Electromyographic signal-based intelligent gesture action generation method |
| CN114515146A (en) * | 2020-11-17 | 2022-05-20 | 北京机械设备研究所 | Intelligent gesture recognition method and system based on electrical measurement |
| 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 |
Also Published As
| Publication number | Publication date |
|---|---|
| CN105608432B (en) | 2019-02-22 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105608432A (en) | Instantaneous myoelectricity image based gesture identification method | |
| CN105654037B (en) | A kind of electromyography signal gesture identification method based on deep learning and characteristic image | |
| CN108491077B (en) | Surface electromyographic signal gesture recognition method based on multi-stream divide-and-conquer convolutional neural network | |
| CN106982359B (en) | A kind of binocular video monitoring method, system and computer readable storage medium | |
| Mane et al. | Hand motion recognition from single channel surface EMG using wavelet & artificial neural network | |
| Cecotti et al. | Convolutional neural network with embedded Fourier transform for EEG classification | |
| CN106980367A (en) | A kind of gesture identification method based on myoelectricity topographic map | |
| CN106909220A (en) | A kind of sight line exchange method suitable for touch-control | |
| CN104463152B (en) | A kind of gesture identification method, system, terminal device and Wearable | |
| Cecotti | A time–frequency convolutional neural network for the offline classification of steady-state visual evoked potential responses | |
| CN104166548B (en) | Deep learning method based on Mental imagery eeg data | |
| CN106293057A (en) | Gesture identification method based on BP neutral net | |
| CN109664300A (en) | A kind of windy lattice calligraphic copying method of robot for feeling study based on power | |
| CN111553307A (en) | Gesture recognition system fusing bioelectrical impedance information and myoelectric information | |
| CN103426438A (en) | Method and system for analyzing baby crying | |
| CN109214395A (en) | A kind of new image representation method based on impulsive neural networks | |
| Shao et al. | Single-channel SEMG using wavelet deep belief networks for upper limb motion recognition | |
| CN110399846A (en) | A kind of gesture identification method based on multichannel electromyography signal correlation | |
| CN106980830A (en) | One kind is based on depth convolutional network from affiliation recognition methods and device | |
| Guo et al. | Research on optimization of static gesture recognition based on convolution neural network | |
| CN109730818A (en) | A kind of prosthetic hand control method based on deep learning | |
| Sun et al. | Gesture recognition algorithm based on multi‐scale feature fusion in RGB‐D images | |
| CN109662710A (en) | A kind of EMG Feature Extraction based on convolutional neural networks | |
| CN110502981A (en) | A kind of gesture identification method merged based on colour information and depth information | |
| CN109766559A (en) | A kind of Sign Language Recognition translation system and its recognition methods |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |