CN109800733B - Data processing method and device and electronic equipment - Google Patents
Data processing method and device and electronic equipment Download PDFInfo
- Publication number
- CN109800733B CN109800733B CN201910092534.4A CN201910092534A CN109800733B CN 109800733 B CN109800733 B CN 109800733B CN 201910092534 A CN201910092534 A CN 201910092534A CN 109800733 B CN109800733 B CN 109800733B
- Authority
- CN
- China
- Prior art keywords
- electromyographic signal
- test data
- signal test
- current
- 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
Abstract
The invention provides a data processing method, which comprises the following steps: sequentially inputting each electromyographic signal test data to a pre-constructed target classifier; when each electromyographic signal test data is input into a target classifier, acquiring a first test result corresponding to the current electromyographic signal test data, and judging whether the current electromyographic signal test data is the first electromyographic signal test data; if so, generating a voting queue and determining a second test result corresponding to the electromyographic signal test data; if not, matching a difference value corresponding to the current electromyographic signal test data with a preset threshold value to obtain a second test result; and updating the voting queue according to the first test result and the second test result, and voting the current electromyographic signal test data according to the updated voting queue to obtain an identification result, so that the accuracy of the identification result of the electromyographic signal test data is improved.
Description
Technical Field
The invention relates to the field of electromyographic signal identification, in particular to a data processing method and device and electronic equipment.
Background
In recent years, with the development of information technology, machine learning has been remarkably advanced, for example, in the field of surface electromyogram signal recognition based on deep learning, a surface electromyogram signal is a bioelectrical signal recorded from a muscle surface through electrode guidance during a neuromuscular system activity, can reflect an activity state of a neuromuscular to a certain extent, and can recognize a muscle activity state corresponding to different surface electromyogram signals through deep learning, thereby generating great assistance in the fields of rehabilitation medicine, sports medicine, biomechanics and the like.
The inventor of the present invention finds that, in the existing method for identifying surface electromyographic signals based on deep learning, a classification model is usually constructed, and the acquired electromyographic signals are input into the classification model to obtain the identification result of the electromyographic signals, however, because the electromyographic signals have instability and randomness, the identification result obtained by a classifier often has a large error, and therefore, how to solve the problem of inaccurate identification result caused by instability and randomness of the electromyographic signals becomes a problem to be solved by the technicians in the field.
Disclosure of Invention
The technical problem to be solved by the invention is to provide a data processing method, wherein pre-acquired electromyographic signal test data is input into a classifier to obtain a first test result, a second test result corresponding to the electromyographic signal test data is determined, a voting queue is updated according to the first test result and the second test result, an identification result is obtained according to the updated voting queue, and the accuracy of the identification result of the electromyographic signal test data is improved.
The invention also provides a data processing device for ensuring the realization and the application of the method in practice.
A method of data processing, comprising:
acquiring a plurality of pre-stored electromyographic signal test data, and sequentially inputting each electromyographic signal test data to a pre-constructed target classifier according to the division sequence of each electromyographic signal test data;
when each electromyographic signal test data is input into the target classifier, acquiring a first test result corresponding to the current electromyographic signal test data, and judging whether the current electromyographic signal test data is the first electromyographic signal test data or not; the current electromyographic signal test data is currently input into the target classifier;
if the current electromyographic signal test data is the first electromyographic signal test data, generating a voting queue and determining a second test result corresponding to the current electromyographic signal test data, wherein the voting queue is used for storing a first test result corresponding to each electromyographic signal test data;
if the current electromyographic signal test data is not the first electromyographic signal test data, matching a difference value corresponding to the current electromyographic signal test data with a preset threshold value to obtain a second test result corresponding to the current electromyographic signal test data, wherein the difference value represents the difference degree between the current electromyographic signal test data and the previous electromyographic signal test data;
and updating the voting queue according to a first test result and a second test result corresponding to the current electromyographic signal test data, and voting the current electromyographic signal test data according to the updated voting queue to obtain an identification result.
Optionally, the method for storing the electromyographic signal test data includes:
collecting electromyographic signals generated by the action executed by a user;
and dividing the electromyographic signals according to the time sequence of the electromyographic signal generation by adopting a sliding window technology to obtain a plurality of electromyographic signal test data, and sequentially storing the divided electromyographic signal test data.
Optionally, before storing the electromyographic signal test data, the method further includes:
analyzing myoelectric signal test data to be stored to obtain time domain characteristics corresponding to the myoelectric signal test data to be stored;
and constructing a two-dimensional electromyographic feature image corresponding to the electromyographic signal test data according to the time domain feature.
Optionally, the matching of the difference value corresponding to the current electromyographic signal test data with a preset threshold value to obtain a second test result corresponding to the current electromyographic signal test data includes:
determining a difference value corresponding to the current electromyographic signal test data according to the representation of the two-dimensional electromyographic feature image corresponding to the current electromyographic signal test data and the representation of the two-dimensional electromyographic feature image corresponding to the previous electromyographic signal test data of the current electromyographic signal test data;
and matching the difference value with a preset threshold value to obtain a second test result corresponding to the current electromyographic signal test data.
Optionally, in the method, if the current electromyographic signal test data is the first electromyographic signal test data, generating a voting queue, and determining a second test result corresponding to the electromyographic signal test data, where the voting queue includes:
if the current electromyographic signal test data is the first electromyographic signal test data, generating a voting queue;
and determining the current state of the voting queue, and determining a second test result corresponding to the current electromyographic signal test data as a non-action transition point according to the current state of the voting queue.
Optionally, the method for updating the voting queue according to the first test result and the second test result corresponding to the current electromyographic signal test data includes:
and when the second test result is a non-action conversion point, storing a first test result corresponding to the current electromyographic signal test data into the voting queue so as to update the voting queue.
Optionally, the matching of the difference value corresponding to the current electromyographic signal test data with a preset threshold value to obtain the second test result includes:
if the difference value is smaller than a preset first threshold value, determining a second test result corresponding to the current electromyographic signal test data as an action conversion point;
if the difference value is larger than a preset second threshold value, determining that a second test result corresponding to the current electromyographic signal test data is a non-action conversion point;
if the difference value is greater than or equal to a preset first threshold value and less than or equal to a preset second threshold value, obtaining a voting result of the current electromyographic signal measurement data according to the voting queue;
if the voting result is consistent with a first test result corresponding to the current electromyographic signal test data, determining that a second test result of the current electromyographic signal test data is a non-action conversion point;
and if the voting result is inconsistent with a first test result corresponding to the current electromyographic signal test data, determining a second test result of the current electromyographic signal test data as an action conversion point.
Optionally, the method for updating the voting queue according to the first test result and the second test result corresponding to the current electromyographic signal test data includes:
when a second test result corresponding to the current electromyographic signal test data is a non-action conversion point, storing a first test result corresponding to the current electromyographic signal test data into the voting queue to update the voting queue;
and when the second test result corresponding to the current electromyographic signal test data is an action conversion point, setting the queue state of the voting queue to be empty, and storing the first test result corresponding to the current electromyographic signal test data into the voting queue to update the voting queue.
A data processing apparatus comprising:
the system comprises an acquisition unit, a target classifier and a data processing unit, wherein the acquisition unit is used for acquiring a plurality of pre-stored electromyographic signal test data and sequentially inputting the electromyographic signal test data to the pre-constructed target classifier according to the division sequence of the electromyographic signal test data;
the judgment unit is used for acquiring a first test result corresponding to the current electromyographic signal test data when each electromyographic signal test data is input to the target classifier, and judging whether the current electromyographic signal test data is the first electromyographic signal test data or not; the current electromyographic signal test data is currently input into the target classifier;
the first determining unit is used for generating a voting queue and determining a second test result corresponding to the current electromyographic signal test data when the current electromyographic signal test data is the first electromyographic signal test data, wherein the voting queue is used for storing the first test result corresponding to each electromyographic signal test data;
the second determining unit is used for matching a difference value corresponding to the current electromyographic signal test data with a preset threshold value to obtain a second test result corresponding to the current electromyographic signal test data when the current electromyographic signal test data is not the first electromyographic signal test data, wherein the difference value represents the difference degree between the current electromyographic signal test data and the previous electromyographic signal test data;
and the identification unit is used for updating the voting queue according to a first test result and a second test result corresponding to the current electromyographic signal test data, and voting the current electromyographic signal test data according to the updated voting queue to obtain an identification result.
An electronic device comprising a memory, and one or more instructions, wherein the one or more instructions are stored in the memory and configured to be executed by the one or more processors to perform the data processing method described above.
Compared with the prior art, the invention has the following advantages:
the invention provides a data processing method, which comprises the following steps: acquiring a plurality of pre-stored electromyographic signal test data, and sequentially inputting each electromyographic signal test data to a pre-constructed target classifier according to the division sequence of each electromyographic signal test data; when each electromyographic signal test data is input into the target classifier, acquiring a first test result corresponding to the current electromyographic signal test data, and judging whether the current electromyographic signal test data is the first electromyographic signal test data or not; if the current electromyographic signal test data is the first electromyographic signal test data, generating a voting queue, and determining a second test result corresponding to the current electromyographic signal test data; if the current electromyographic signal test data is not the first electromyographic signal test data, matching a difference value corresponding to the current electromyographic signal test data with a preset threshold value to obtain a second test result; and updating the voting queue according to the first test result and the second test result corresponding to the current electromyographic signal test data, and voting the current electromyographic signal test data according to the updated voting queue to obtain an identification result, so that the accuracy of the identification result of the electromyographic signal test data is improved.
Drawings
In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings needed to be used in the description of the embodiments will be briefly introduced below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art to obtain other drawings based on these drawings without inventive labor.
FIG. 1 is a flow chart of a method of data processing according to the present invention;
FIG. 2 is a diagram illustrating a data processing method according to an embodiment of the present invention;
FIG. 3 is a schematic structural diagram of a data processing apparatus according to the present invention;
fig. 4 is a schematic structural diagram of an electronic device provided in the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
The invention is operational with numerous general purpose or special purpose computing device environments or configurations. For example: personal computers, server computers, hand-held or portable devices, tablet-type devices, multi-processor apparatus, distributed computing environments that include any of the above devices or equipment, and the like.
An embodiment of the present invention provides a data processing method, which may be applied to multiple system platforms, where an execution subject of the method may be a computer terminal or a processor of various mobile devices, and a flowchart of the method is shown in fig. 1, and specifically includes:
s101: the method comprises the steps of obtaining a plurality of pre-stored electromyographic signal test data, and sequentially inputting the electromyographic signal test data to a pre-constructed target classifier according to the division sequence of the electromyographic signal test data.
In the method provided by the embodiment of the invention, the target classifier is a classifier which is subjected to supervised learning, semi-supervised learning or unsupervised learning.
In the method provided by the embodiment of the present invention, the dividing sequence may be a time sequence or an arrangement sequence.
In the method provided by the embodiment of the present invention, the electromyographic signal test data may be any one of a plurality of electromyographic signal test data corresponding to a certain action performed by a user.
S102: when each electromyographic signal test data is input into the target classifier, acquiring a first test result corresponding to the current electromyographic signal test data, and judging whether the current electromyographic signal test data is the first electromyographic signal test data or not; the current electromyographic signal test data is currently input into the target classifier.
In the method provided by the embodiment of the invention, the electromyographic signal test data is input to the target classifier, so that the classification result of the electromyographic signal test data can be obtained, namely, the action type corresponding to the electromyographic signal test data is preliminarily identified by the target classifier, and the classification result is the first test result.
S103: if the current electromyographic signal test data is the first electromyographic signal test data, generating a voting queue and determining a second test result corresponding to the current electromyographic signal test data, wherein the voting queue is used for storing a first test result corresponding to each electromyographic signal test data.
In the method provided by the embodiment of the present invention, the voting queue may be used to store a first test result corresponding to the current electromyographic signal test data or a historical first test result, and the initial state of the voting queue is empty, that is, no data is stored; the historical first test result is a first test result of the myoelectric signal test data prior to the current myoelectric signal test data.
S104: and if the current electromyographic signal test data is not the first electromyographic signal test data, matching a difference value corresponding to the current electromyographic signal test data with a preset threshold value to obtain a second test result corresponding to the current electromyographic signal test data.
In the method provided by the embodiment of the present invention, the difference value represents a difference degree between the current electromyographic signal test data and the previous electromyographic signal test data of the current electromyographic signal test data.
S105: and updating the voting queue according to a first test result and a second test result corresponding to the current electromyographic signal test data, and voting the current electromyographic signal test data according to the updated voting queue to obtain an identification result.
In the method provided by the embodiment of the invention, after the voting queue is updated, voting is carried out according to the first test results respectively corresponding to one or more electromyographic signal test data stored in the voting queue, and the identification result is obtained according to the voting result.
In the method provided by the embodiment of the invention, the identification result is an action corresponding to the electromyographic signal test data.
According to the data processing method provided by the embodiment of the invention, a plurality of pre-stored electromyographic signal test data are acquired, and the electromyographic signal test data are sequentially input to a pre-constructed target classifier according to the division sequence of the electromyographic signal test data; when each electromyographic signal test data is input into the target classifier, acquiring a first test result corresponding to the current electromyographic signal test data, and judging whether the current electromyographic signal test data is the first electromyographic signal test data or not; if the current electromyographic signal test data is the first electromyographic signal test data, generating a voting queue, and determining a second test result corresponding to the current electromyographic signal test data; if the current electromyographic signal test data is not the first electromyographic signal test data, matching a difference value corresponding to the current electromyographic signal test data with a preset threshold value to obtain a second test result corresponding to the current electromyographic signal test data; and updating the voting queue according to the first test result and the second test result corresponding to the current electromyographic signal test data, and voting the current electromyographic signal test data according to the updated voting queue to obtain an identification result, so that the accuracy of the identification result of the electromyographic signal test data is improved.
In the data processing method provided in the embodiment of the present invention, the storing process of the electromyographic signal test data may include:
collecting electromyographic signals generated by the action executed by a user;
and dividing the electromyographic signals according to the time sequence of the electromyographic signal generation by adopting a sliding window technology to obtain a plurality of electromyographic signal test data, and sequentially storing the divided electromyographic signal test data.
In the method provided by the embodiment of the invention, the electromyographic signals generated by executing at least one action by the user are collected by the electromyographic signal collecting equipment, the user can repeatedly execute the same action for multiple times, different force can be adopted when the same action is executed each time, and the number of the row channels, the number of the column channels and the distance between the adjacent channels of the electromyographic signal collecting equipment are preset.
In the method provided by the embodiment of the invention, the action can be limb action of a human or an animal; optionally, the motion is a gesture motion of a human body.
Optionally, the action is at least one action randomly performed by a person, and one action corresponds to one electromyographic signal.
In the method provided by the embodiment of the invention, a sliding window technology is adopted to divide the collected continuous electromyographic signals corresponding to the single preset action executed by the user from the beginning to the end; and dividing the continuous electromyographic signals according to the time sequence of the generation of the continuous electromyographic signals to obtain a plurality of electromyographic signal test data by presetting the window length and the sliding increment of the sliding window.
In the data processing method provided in the embodiment of the present invention, before storing the electromyographic signal test data, optionally, the method further includes:
analyzing myoelectric signal test data to be stored to obtain time domain characteristics corresponding to the myoelectric signal test data to be stored;
and constructing a two-dimensional electromyographic feature image corresponding to the electromyographic signal test data according to the time domain feature.
In the method provided by the embodiment of the invention, the electromyographic signal test data is sequentially stored into a preset test data storage area according to a dividing sequence, and before or after the electromyographic signal test data is stored, the electromyographic signal test data is analyzed to obtain time domain characteristics; and constructing a two-dimensional electromyographic feature image corresponding to the electromyographic signal test data according to the time domain feature.
In the method provided by the embodiment of the invention, time domain characteristics are obtained by analyzing single electromyographic signal test data; the time domain features comprise a plurality of features related to the electromyographic signal test data, such as an average absolute value MAV, a slope symbol change rate SSC, a zero crossing rate ZC, a waveform length WL and the like.
In the method provided by the embodiment of the invention, the time domain characteristics of the single electromyographic signal test data are integrated into a two-dimensional electromyographic characteristic image according to the arrangement sequence of the channels of the acquisition equipment.
The step of matching a difference value corresponding to the current electromyographic signal test data with a preset threshold value to obtain a second test result corresponding to the current electromyographic signal test data includes:
determining a difference value corresponding to the current electromyographic signal test data according to the representation of the two-dimensional electromyographic feature image corresponding to the current electromyographic signal test data and the representation of the two-dimensional electromyographic feature image corresponding to the previous electromyographic signal test data of the current electromyographic signal test data;
and matching the difference value with a preset threshold value to obtain a second test result corresponding to the current electromyographic signal test data.
In the data processing method provided in the embodiment of the present invention, if the electromyographic signal test data of any one of the actions is selected for identification, the method for calculating the difference value specifically includes:
K(t,x)＝psnr3(S(t,x-1),S(t,x))
where(1≤t≤M,2≤x≤N(t))
k (t, x) represents a difference value of xth electromyographic signal test data of the tth preset action, S (t, x) represents a two-dimensional electromyographic feature image of the xth electromyographic signal test data of the tth preset action, psnr is a peak signal-to-noise ratio between the two-dimensional electromyographic feature images, N (t) represents the number of the xth electromyographic signal test data of the preset action, and M is the total number of the preset actions.
In the method provided in the embodiment of the present invention, optionally, if any plurality of test data corresponding to any action are simultaneously selected for identification, the method for calculating the difference value specifically includes:
K(x)＝psnr3(S(x-1),S(x))
where(1≤x≤N)
k (x) represents a difference value of the xth electromyographic signal test data, S (x) represents a two-dimensional electromyographic feature image of the xth electromyographic signal test data, psnr is a peak signal-to-noise ratio between the two-dimensional electromyographic feature images, and N represents the number of the electromyographic signal test data.
The difference value calculation method provided by the embodiment of the invention has the advantages of stronger robustness, simple calculation process and smaller time delay, is suitable for real-time myoelectric control, and is also suitable under the condition of no resting action continuous change.
In the data processing method provided in the embodiment of the present invention, the process of constructing the target classifier may include:
collecting myoelectric signals generated by a user executing a preset action; dividing the electromyographic signals according to the time sequence of the electromyographic signal generation by adopting a sliding window technology to obtain a plurality of electromyographic signal training data, and sequentially storing the divided electromyographic signal training data; storing the electromyographic signal test data into a preset training data storage area according to a dividing sequence; analyzing the electromyographic signal training data to obtain training data time domain characteristics; determining an initial classifier according to the training data time domain characteristics; and training the initial classifier according to the electromyographic signal training data to obtain a target classifier.
In the method provided by the implementation of the invention, the time domain characteristics of the training data of the electromyographic signals comprise the characteristics of an average absolute value MAV, a slope symbol change rate SSC, a zero crossing rate ZC, a waveform length WL and the like.
In the method provided by the embodiment of the invention, the type of an initial classifier is determined according to the characteristics of the mean absolute value MAV, the slope symbol change rate SSC, the zero crossing rate ZC, the waveform length WL and the like contained in the training data time domain characteristics of the electromyographic signal training data; the initial classifier can be one or more of a logistic regression classifier, an SVM classifier, a K-NN proximity classifier, a Bayesian classifier, and the like.
In the method provided by the embodiment of the invention, the initial classifier is trained according to the electromyographic signal training data, so that the accuracy of classifying the input electromyographic signal data reaches a preset standard, and the initial classifier reaching the standard is determined as a target classifier.
In the data processing method provided by the embodiment of the present invention, if the current electromyographic signal test data is the first electromyographic signal test data, generating a voting queue, and determining a second test result corresponding to the electromyographic signal test data, the method includes:
if the current electromyographic signal test data is the first electromyographic signal test data, generating a voting queue;
and determining the current state of the voting queue, and determining a second test result corresponding to the current electromyographic signal test data as a non-action transition point according to the current state of the voting queue.
In the method provided by the embodiment of the present invention, if the current status of the voting queue is empty, it is determined that a second test result corresponding to the electromyographic signal test data is a non-action transition point.
In the data method provided in the embodiment of the present invention, if the current electromyographic signal test data is the first electromyographic signal test data, the updating the voting queue according to the first test result and the second test result corresponding to the current electromyographic signal test data includes:
and when the second test result is a non-action conversion point, storing a first test result corresponding to the current electromyographic signal test data into the voting queue so as to update the voting queue.
In the data processing method provided in the embodiment of the present invention, the preset threshold includes a preset first threshold and a preset second threshold, and the matching the difference value corresponding to the current electromyographic signal test data with the preset threshold to obtain the second test result may include:
if the difference value is smaller than a preset first threshold value, determining a second test result corresponding to the current electromyographic signal test data as an action conversion point;
if the difference value is larger than a preset second threshold value, determining that a second test result corresponding to the current electromyographic signal test data is a non-action conversion point;
if the difference value is greater than or equal to a preset first threshold value and less than or equal to a preset second threshold value, obtaining a voting result of the electromyographic signal measurement data according to the voting queue;
if the voting result is consistent with a first test result corresponding to the current electromyographic signal test data, determining that a second test result of the current electromyographic signal test data is a non-action conversion point;
and if the voting result is inconsistent with a first test result corresponding to the current electromyographic signal test data, determining a second test result of the current electromyographic signal test data as an action conversion point.
And if the voting queue is empty, determining that a second test result of the current electromyographic signal test data is a non-action conversion point.
In the method provided in the embodiment of the present invention, the obtaining of the voting result of the electromyographic signal measurement data according to the voting queue may include:
if the number of historical first test results which are stored in the voting queue and are consistent with the first test result corresponding to the current electromyographic signal test data is greater than or equal to the number of historical first test results which are inconsistent with the first test result corresponding to the current electromyographic signal test data, determining that the voting result is consistent with the first test result corresponding to the current electromyographic signal test data;
and if the data quantity of the historical first test results which are stored in the voting queue and are consistent with the first test result corresponding to the current electromyographic signal test data is smaller than the quantity of the historical first test results which are inconsistent with the first test result corresponding to the current electromyographic signal test data, determining that the voting result is inconsistent with the first test result corresponding to the current electromyographic signal test data.
In the data processing method provided in the embodiment of the present invention, the updating the voting queue according to the first test result and the second test result corresponding to the current electromyographic signal test data may include:
when a second test result corresponding to the current electromyographic signal test data is a non-action conversion point, storing a first test result corresponding to the current electromyographic signal test data into the voting queue to update the voting queue;
and when the second test result corresponding to the current electromyographic signal test data is an action conversion point, setting the queue state of the voting queue to be empty, and storing the first test result corresponding to the current electromyographic signal test data into the voting queue to update the voting queue.
The data processing method provided by the embodiment of the invention can be applied to the recognition of surface electromyographic signals generated by the limb actions of people or animals, and the following description takes the gesture actions of recognizing people as an example, and specifically comprises the following steps:
the method comprises the steps of pre-selecting a plurality of gesture actions, wherein the gesture actions comprise 14 actions of extending a thumb, an index finger, a middle finger, a little finger, a thumb and an index finger together, an index finger and a middle finger together, a middle finger and a ring finger together, a ring finger and a little finger together, a thumb, an index finger and a middle finger together, a middle finger, a ring finger and a little finger together, a thumb, an index finger, a middle finger and a ring finger together, an index finger, a middle finger, a ring finger and a little finger together, and a thumb, an index finger, a middle finger, a ring finger and a little finger together.
The method comprises the steps of attaching an electrode array of an electromyographic signal acquisition device to a position of an action muscle activation area, wherein the electromyographic signal acquisition device selects a flexible high-density electrode array with 8 row channels, 6 column channels and 14mm adjacent channel spacing.
The user executes the 14 preset actions one by one, and in the process of executing each action, the electromyographic signals generated by the user executing each action are collected, wherein two different force ranges are selected to repeatedly execute one action, the first force is 70% -80% of the maximum random contractility, the second force is 30% -40% of the maximum random contractility, the first force and the second force are respectively used for executing each action for 4 times, isometric contraction with unchanged force is kept for 6 seconds each time, then 4 seconds are kept for rest, and different action intervals give sufficient rest time to a subject.
Dividing continuous electromyographic signals obtained by acquiring each preset action executed by a user by using a sliding window technology, and dividing the continuous electromyographic signals obtained by executing one preset action once by the user through one force each time, wherein the window length of a sliding window is set to be 128 milliseconds, and the sliding increment is set to be 64 milliseconds; dividing a continuous electromyographic signal flow into a certain number of analysis windows by a sliding window technology to obtain a plurality of sample data corresponding to a user executing a preset action by a force;
and performing multi-feature extraction on the electromyographic signals of each channel in each sample data to construct a two-dimensional electromyographic feature image.
The extracted multiple features comprise time domain features, wherein the time domain features comprise an average absolute value MAV, a slope symbol change rate SSC, a zero crossing rate ZC and a waveform length WL, then the features extracted from each sample are integrated into a two-dimensional myoelectricity feature image according to a channel arrangement sequence, the image length is 8 and is in direct proportion to the number of row channels, the image width is 6 and is in direct proportion to the number of column channels, and the color channel dimension is the same as the feature number.
Optionally, the sample data is divided into electromyographic signal training data and electromyographic signal test data according to the first force and the second force, and if the sample data corresponding to the action executed by the first force is determined as the electromyographic signal training data, the sample data corresponding to the action executed by the second force is determined as the electromyographic signal test data; and if the sample data corresponding to the action executed by the second force is determined as myoelectric signal training data, determining the sample data corresponding to the action executed by the first force as myoelectric signal test data, and storing the myoelectric signal training data and the myoelectric signal test data.
Determining a classifier according to time domain characteristics of sample data, training the classifier according to electromyographic signal training data, and determining the classifier meeting preset requirements as a target classifier, wherein the classifier meeting the preset requirements can be a classifier meeting a preset recognition rate or a classifier meeting a preset training amount.
Setting a first threshold and a second threshold according to the electromyographic signal training data, wherein the first threshold is TH1, and the second threshold is TH1+ TH 2; alternatively, TH1 has a value of-1E 5 and TH2 has a value of 0.35E 5.
The method comprises the steps of obtaining a plurality of pre-stored electromyographic signal test data, and sequentially inputting the electromyographic signal test data to a pre-constructed target classifier according to the division sequence of the electromyographic signal test data. The process of inputting the electromyographic signal test data to the target classifier, as shown in fig. 2, specifically includes:
s201: and when each electromyographic signal test data is input into the target classifier, acquiring a first test result corresponding to the current electromyographic signal test data.
In the process of executing S201, the first test result is an initial identification result obtained by the classifier according to the electromyographic signal test data.
S202: and judging whether the current electromyographic signal test data is the first electromyographic signal test data, if so, executing S203, and if not, executing S204.
S203: generating a voting queue and determining a second test result corresponding to the electromyographic signal test data.
S204: and matching a difference value corresponding to the current electromyographic signal test data with a preset threshold value to obtain the second test result.
S205: and judging whether the second test result is an action conversion point, if not, executing S206, and if so, executing S207.
S206: and storing the first test result into the voting queue to update the voting queue.
S207: and setting the queue state of the voting queue to be empty, and storing the first test result into the voting queue to update the voting queue.
S208: and voting the current electromyographic signal test data according to the updated voting queue to obtain an identification result.
Based on the method provided by the embodiment of the invention, the collected continuous electromyographic signals obtained by the user by performing forefinger stretching are divided into 5 electromyographic signal test data by adopting a sliding window technology and are sequentially input into a target classifier, assuming that the identification results of the 1 st electromyographic signal test data and the 2 nd electromyographic signal test data are both forefinger stretching and the first test results are both forefinger stretching, namely that two first test results representing the forefinger stretching exist in a voting queue, if the 3 rd electromyographic signal test data are input into the target classifier, the obtained first test result is the middle finger stretching, and the difference value of the 3 rd electromyographic signal test data is greater than a first threshold value, namely that the second test result of the 3 rd electromyographic signal test data is a non-action conversion point, and then the first test result of the 3 rd electromyographic signal test data is stored into the voting queue to realize updating voting queue, storing three first test results in the updated voting queue, wherein two test results represent that the index finger extends and one test result represents that the middle finger extends; voting the 3 rd electromyographic signal test data by the updated voting queue, wherein the number of the first test results representing the stretching of the index finger in the updated voting queue is larger than that representing the stretching of the middle finger, so that the identification result of the 3 rd electromyographic signal test data is finally obtained and is the stretching of the index finger.
If the first test result of the 3 rd electromyographic signal test data is 'index finger extension', the difference value is greater than or equal to the first threshold value and less than or equal to the second threshold value; two historical first test results representing index finger stretching exist in the current voting queue, so that the voting result is consistent with the first test result of the 3 rd electromyographic signal test data and is index finger stretching, the 3 rd electromyographic signal test data is judged to be a non-action transition point, the first test result of the 3 rd electromyographic signal test data is stored in the voting queue to update the voting queue, the 3 rd electromyographic signal test data is voted according to the updated voting queue, and the identification result of the 3 rd electromyographic signal test data is the index finger stretching.
In the method provided by the embodiment of the invention, in the process of determining the identification result through voting in the voting queue, if the historical first test result representing the stretching of the index finger stored in the voting queue is the same as the historical first test result representing the stretching of the middle finger, the voting result is determined according to the accuracy of the historical identification result or the voting result is determined according to the first test result of the current electromyographic signal test data.
In the method provided by the embodiment of the present invention, optionally, in the process of voting according to the voting queue to determine the second test result of the electromyographic signal test data, if the historical first test result representing "index finger stretching" stored in the voting queue is the same as the historical first test result representing "middle finger stretching", determining the current electromyographic signal test data to be an action conversion point or a non-action conversion point according to the first test result of the current electromyographic signal test data, if the first test result of the current electromyographic signal test data is the same as the first test result of the first electromyographic signal test data, and if the second test result of the current electromyographic signal test data is different from the first test result of the first electromyographic signal test data, the second test result of the current electromyographic signal test data is the action conversion point.
According to the data processing method provided by the embodiment of the invention, whether the action of the current electromyographic signal test data is changed in the acquisition process is determined by judging whether the second test result is an action conversion point, on the basis, whether the current electromyographic signal test data and the previous electromyographic signal test data correspond to the same action is judged, all the continuous electromyographic signal test data between two adjacent action conversion points are considered to be of the same action type, and the identification result of the electromyographic signal test data and the first test result corresponding to all the electromyographic signal test data from the previous action conversion point to the current electromyographic signal test data are subjected to majority voting every time of testing one electromyographic signal test data, so that the final identification result of the current electromyographic signal test data is obtained, and the method has strong robustness.
The above specific implementations and the derivation processes of the implementations are all within the scope of the present invention.
Corresponding to the method described in fig. 1, an embodiment of the present invention further provides a data processing apparatus, which is used for implementing the method in fig. 1 specifically, the data processing apparatus provided in the embodiment of the present invention may be applied to a computer terminal or various mobile devices, and a schematic structural diagram of the data processing apparatus is shown in fig. 3, and specifically includes:
an obtaining unit 301, configured to obtain a plurality of pre-stored electromyographic signal test data, and sequentially input each of the plurality of electromyographic signal test data to a pre-constructed target classifier according to a division order of each of the plurality of electromyographic signal test data;
a determining unit 302, configured to obtain a first test result corresponding to the current electromyographic signal test data when each electromyographic signal test data is input to the target classifier, and determine whether the current electromyographic signal test data is a first electromyographic signal test data; the current electromyographic signal test data is currently input into the target classifier;
a first determining unit 303, configured to generate a voting queue when the current electromyographic signal test data is a first electromyographic signal test data, and determine a second test result corresponding to the current electromyographic signal test data, where the voting queue is used to store a first test result corresponding to each electromyographic signal test data;
a second determining unit 304, configured to, when the current electromyographic signal test data is not the first electromyographic signal test data, match a difference value corresponding to the current electromyographic signal test data with a preset threshold to obtain a second test result corresponding to the current electromyographic signal test data, where the difference value represents a difference degree between the current electromyographic signal test data and a previous electromyographic signal test data;
the identifying unit 305 is configured to update the voting queue according to a first test result and a second test result corresponding to the current electromyographic signal test data, and vote for the current electromyographic signal test data according to the updated voting queue to obtain an identification result.
In the data processing device provided by the embodiment of the invention, a processor acquires a plurality of pre-stored electromyographic signal test data and sequentially inputs the electromyographic signal test data to a pre-constructed target classifier according to the division sequence of the electromyographic signal test data; when each electromyographic signal test data is input into the target classifier, acquiring a first test result corresponding to the current electromyographic signal test data, and judging whether the current electromyographic signal test data is the first electromyographic signal test data or not; if the current electromyographic signal test data is the first electromyographic signal test data, generating a voting queue and determining a second test result corresponding to the electromyographic signal test data; if the current electromyographic signal test data is not the first electromyographic signal test data, matching a difference value corresponding to the current electromyographic signal test data with a preset threshold value to obtain a second test result corresponding to the current electromyographic signal test data; and updating the voting queue according to the first test result and the second test result corresponding to the current electromyographic signal test data, and voting the current electromyographic signal test data according to the updated voting queue to obtain an identification result, so that the accuracy of the identification result of the electromyographic signal test data is improved.
In the data processing apparatus according to an embodiment of the present invention, before the storing of the electromyographic signal test data, the obtaining unit may include:
the first acquisition subunit is used for acquiring electromyographic signals generated by actions executed by a user;
the first storage subunit divides the electromyographic signals according to the time sequence of the electromyographic signal generation by adopting a sliding window technology to obtain a plurality of electromyographic signal test data, and stores the divided electromyographic signal test data in sequence.
In the data processing apparatus according to an embodiment of the present invention, before storing the electromyographic signal test data, the obtaining unit may further include:
the first analysis subunit is used for analyzing the electromyographic signal test data to be stored to obtain time domain characteristics corresponding to the electromyographic signal test data to be stored;
and the construction subunit is used for constructing a two-dimensional electromyographic characteristic image corresponding to the electromyographic signal test data according to the time domain characteristics.
In the data processing apparatus provided in the embodiment of the present invention, in the process of obtaining the second test result corresponding to the current electromyographic signal test data by matching the difference value corresponding to the current electromyographic signal test data with the preset threshold, the second determining unit may include:
the first determining subunit is configured to determine a difference value corresponding to the current electromyographic signal test data according to a representation of a two-dimensional electromyographic feature image corresponding to the current electromyographic signal test data and a representation of a two-dimensional electromyographic feature image corresponding to an electromyographic signal test data previous to the current electromyographic signal test data;
and the first matching subunit is used for matching the difference value with a preset threshold value to obtain a second test result corresponding to the current electromyographic signal test data.
In the data processing apparatus according to an embodiment of the present invention, the first determining unit, which generates a voting queue and determines a second test result corresponding to the electromyographic signal test data if the current electromyographic signal test data is a first electromyographic signal test data, includes:
the generation subunit is used for generating a voting queue when the current electromyographic signal test data is the first electromyographic signal test data;
and the third determining subunit is configured to determine a current state of the voting queue, and determine, according to the current state of the voting queue, that a second test result corresponding to the current electromyographic signal test data is a non-action transition point.
In the data processing apparatus according to an embodiment of the present invention, the identification unit that updates the voting queue according to the first test result and the second test result corresponding to the current electromyographic signal test data includes:
and the first updating subunit is configured to, when the second test result is a non-action transition point, store the first test result corresponding to the current electromyographic signal test data in the voting queue, so as to update the voting queue.
In the data processing apparatus provided in the embodiment of the present invention, the second determining unit that matches a difference value corresponding to the current electromyographic signal test data with a preset threshold to obtain the second test result includes:
a fourth determination subunit: the myoelectric signal testing device is used for determining a second testing result corresponding to the current myoelectric signal testing data as an action conversion point when the difference value is smaller than a preset first threshold value;
the fifth determining subunit is configured to determine, when the difference value is greater than a preset second threshold, that a second test result corresponding to the current electromyographic signal test data is a non-action transition point;
a sixth determining subunit, configured to, when the difference value is greater than or equal to a preset first threshold and less than or equal to a preset second threshold, obtain a voting result of the current electromyographic signal measurement data according to the voting queue;
a seventh determining subunit, configured to determine, when the voting result is consistent with the first test result corresponding to the current electromyographic signal test data, that a second test result of the current electromyographic signal test data is a non-action transition point;
and the eighth determining subunit is configured to determine, when the voting result is inconsistent with the first test result corresponding to the current electromyographic signal test data, that the second test result of the current electromyographic signal test data is an action conversion point.
In the data processing apparatus according to an embodiment of the present invention, the identification unit that updates the voting queue according to the first test result and the second test result corresponding to the current electromyographic signal test data includes:
the second updating subunit is configured to, when a second test result corresponding to the current electromyographic signal test data is a non-action transition point, store a first test result corresponding to the current electromyographic signal test data in the voting queue to update the voting queue;
and the third updating subunit is configured to set a queue state of the voting queue to be empty when the second test result corresponding to the current electromyographic signal test data is an action transition point, and store the first test result corresponding to the current electromyographic signal test data in the voting queue to update the voting queue.
The embodiment of the invention also provides a storage medium, which comprises a stored instruction, wherein when the instruction runs, the device where the storage medium is located is controlled to execute the data processing method.
An electronic device is provided in an embodiment of the present invention, and the structural diagram of the electronic device is shown in fig. 4, which specifically includes a memory 401 and one or more instructions 402, where the one or more instructions 402 are stored in the memory 401, and the one or more instructions 402 configured to be executed by one or more processors 403 include instructions for:
acquiring a plurality of pre-stored electromyographic signal test data, and sequentially inputting each electromyographic signal test data to a pre-constructed target classifier according to the division sequence of each electromyographic signal test data;
when each electromyographic signal test data is input into the target classifier, acquiring a first test result corresponding to the current electromyographic signal test data, and judging whether the current electromyographic signal test data is the first electromyographic signal test data or not; the current electromyographic signal test data is currently input into the target classifier;
if the current electromyographic signal test data is the first electromyographic signal test data, generating a voting queue and determining a second test result corresponding to the current electromyographic signal test data, wherein the voting queue is used for storing a first test result corresponding to each electromyographic signal test data;
if the current electromyographic signal test data is not the first electromyographic signal test data, matching a difference value corresponding to the current electromyographic signal test data with a preset threshold value to obtain a second test result corresponding to the current electromyographic signal test data, wherein the difference value represents the difference degree between the current electromyographic signal test data and the previous electromyographic signal test data;
and updating the voting queue according to a first test result and a second test result corresponding to the current electromyographic signal test data, and voting the current electromyographic signal test data according to the updated voting queue to obtain an identification result.
It should be noted that, in the present specification, the embodiments are all described in a progressive manner, each embodiment focuses on differences from other embodiments, and the same and similar parts among the embodiments may be referred to each other. For the device-like embodiment, since it is basically similar to the method embodiment, the description is simple, and for the relevant points, reference may be made to the partial description of the method embodiment.
Finally, it should also be noted that, herein, relational terms such as first and second, and the like may be used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Also, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other identical elements in a process, method, article, or apparatus that comprises the element.
For convenience of description, the above devices are described as being divided into various units by function, and are described separately. Of course, the functions of the units may be implemented in the same software and/or hardware or in a plurality of software and/or hardware when implementing the invention.
From the above description of the embodiments, it is clear to those skilled in the art that the present invention can be implemented by software plus necessary general hardware platform. Based on such understanding, the technical solutions of the present invention may be embodied in the form of a software product, which may be stored in a storage medium, such as ROM/RAM, magnetic disk, optical disk, etc., and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute the method according to the embodiments or some parts of the embodiments.
The data processing method and apparatus provided by the present invention are described in detail above, and a specific example is applied in the text to explain the principle and the implementation of the present invention, and the description of the above embodiment is only used to help understanding the method and the core idea of the present invention; meanwhile, for a person skilled in the art, according to the idea of the present invention, there may be variations in the specific embodiments and the application scope, and in summary, the content of the present specification should not be construed as a limitation to the present invention.
Claims (10)
1. A data processing method, comprising:
acquiring a plurality of pre-stored electromyographic signal test data, and sequentially inputting each electromyographic signal test data to a pre-constructed target classifier according to the division sequence of each electromyographic signal test data;
when each electromyographic signal test data is input into the target classifier, acquiring a first test result corresponding to the current electromyographic signal test data, and judging whether the current electromyographic signal test data is the first electromyographic signal test data or not; the current electromyographic signal test data is currently input into the target classifier;
if the current electromyographic signal test data is the first electromyographic signal test data, generating a voting queue and determining a second test result corresponding to the current electromyographic signal test data, wherein the voting queue is used for storing a first test result corresponding to each electromyographic signal test data;
if the current electromyographic signal test data is not the first electromyographic signal test data, matching a difference value corresponding to the current electromyographic signal test data with a preset threshold value to obtain a second test result corresponding to the current electromyographic signal test data, wherein the difference value represents the difference degree between the current electromyographic signal test data and the previous electromyographic signal test data;
and updating the voting queue according to a first test result and a second test result corresponding to the current electromyographic signal test data, and voting the current electromyographic signal test data according to the updated voting queue to obtain an identification result.
2. The method according to claim 1, wherein the storing process of the electromyographic signal test data comprises:
collecting electromyographic signals generated by the action executed by a user;
and dividing the electromyographic signals according to the time sequence of the electromyographic signal generation by adopting a sliding window technology to obtain a plurality of electromyographic signal test data, and sequentially storing the divided electromyographic signal test data.
3. The method of claim 2, wherein before storing the electromyographic signal test data, further comprising:
analyzing myoelectric signal test data to be stored to obtain time domain characteristics corresponding to the myoelectric signal test data to be stored;
and constructing a two-dimensional electromyographic feature image corresponding to the electromyographic signal test data to be stored according to the time domain feature.
4. The method according to claim 3, wherein the step of matching the difference value corresponding to the current electromyographic signal test data with a preset threshold value to obtain a second test result corresponding to the current electromyographic signal test data comprises:
determining a difference value corresponding to the current electromyographic signal test data according to the representation of the two-dimensional electromyographic feature image corresponding to the current electromyographic signal test data and the representation of the two-dimensional electromyographic feature image corresponding to the previous electromyographic signal test data of the current electromyographic signal test data;
and matching the difference value with a preset threshold value to obtain a second test result corresponding to the current electromyographic signal test data.
5. The method according to claim 1, wherein if the current electromyographic signal test data is the first electromyographic signal test data, generating a voting queue, and determining a second test result corresponding to the electromyographic signal test data, comprises:
if the current electromyographic signal test data is the first electromyographic signal test data, generating a voting queue;
and determining the current state of the voting queue, and determining a second test result corresponding to the current electromyographic signal test data as a non-action transition point according to the current state of the voting queue.
6. The method according to claim 5, wherein the updating the voting queue according to the first test result and the second test result corresponding to the current electromyographic signal test data comprises:
and when the second test result is a non-action conversion point, storing a first test result corresponding to the current electromyographic signal test data into the voting queue so as to update the voting queue.
7. The method according to claim 1, wherein the matching of the difference value corresponding to the current electromyographic signal test data with a preset threshold value to obtain the second test result comprises:
if the difference value is smaller than a preset first threshold value, determining a second test result corresponding to the current electromyographic signal test data as an action conversion point;
if the difference value is larger than a preset second threshold value, determining that a second test result corresponding to the current electromyographic signal test data is a non-action conversion point;
if the difference value is greater than or equal to a preset first threshold value and less than or equal to a preset second threshold value, obtaining a voting result of the current electromyographic signal measurement data according to the voting queue;
if the voting result is consistent with a first test result corresponding to the current electromyographic signal test data, determining that a second test result of the current electromyographic signal test data is a non-action conversion point;
and if the voting result is inconsistent with a first test result corresponding to the current electromyographic signal test data, determining a second test result of the current electromyographic signal test data as an action conversion point.
8. The method according to claim 7, wherein the updating the voting queue according to the first test result and the second test result corresponding to the current electromyographic signal test data comprises:
when a second test result corresponding to the current electromyographic signal test data is a non-action conversion point, storing a first test result corresponding to the current electromyographic signal test data into the voting queue to update the voting queue;
and when the second test result corresponding to the current electromyographic signal test data is an action conversion point, setting the queue state of the voting queue to be empty, and storing the first test result corresponding to the current electromyographic signal test data into the voting queue to update the voting queue.
9. A data processing apparatus, comprising:
the system comprises an acquisition unit, a target classifier and a data processing unit, wherein the acquisition unit is used for acquiring a plurality of pre-stored electromyographic signal test data and sequentially inputting the electromyographic signal test data to the pre-constructed target classifier according to the division sequence of the electromyographic signal test data;
the judgment unit is used for acquiring a first test result corresponding to the current electromyographic signal test data when each electromyographic signal test data is input to the target classifier, and judging whether the current electromyographic signal test data is the first electromyographic signal test data or not; the current electromyographic signal test data is currently input into the target classifier;
the first determining unit is used for generating a voting queue and determining a second test result corresponding to the current electromyographic signal test data when the current electromyographic signal test data is the first electromyographic signal test data, wherein the voting queue is used for storing the first test result corresponding to each electromyographic signal test data;
the second determining unit is used for matching a difference value corresponding to the current electromyographic signal test data with a preset threshold value to obtain a second test result corresponding to the current electromyographic signal test data when the current electromyographic signal test data is not the first electromyographic signal test data, wherein the difference value represents the difference degree between the current electromyographic signal test data and the previous electromyographic signal test data;
and the identification unit is used for updating the voting queue according to a first test result and a second test result corresponding to the current electromyographic signal test data, and voting the current electromyographic signal test data according to the updated voting queue to obtain an identification result.
10. An electronic device comprising a memory, and one or more instructions stored in the memory and configured to be executed by the one or more processors to perform the data processing method of any one of claims 1 to 8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910092534.4A CN109800733B (en) | 2019-01-30 | 2019-01-30 | Data processing method and device and electronic equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910092534.4A CN109800733B (en) | 2019-01-30 | 2019-01-30 | Data processing method and device and electronic equipment |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109800733A CN109800733A (en) | 2019-05-24 |
| CN109800733B true CN109800733B (en) | 2021-03-09 |
Family
ID=66560630
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910092534.4A Active CN109800733B (en) | 2019-01-30 | 2019-01-30 | Data processing method and device and electronic equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109800733B (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110403609B (en) * | 2019-09-03 | 2020-09-01 | 北京海益同展信息科技有限公司 | Motion speed analysis method and device and wearable equipment |
| CN111401166A (en) * | 2020-03-06 | 2020-07-10 | 中国科学技术大学 | Robust gesture recognition method based on electromyographic information decoding |
| CN113688802B (en) * | 2021-10-22 | 2022-04-01 | 季华实验室 | Gesture recognition method, device and equipment based on electromyographic signals and storage medium |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107679516A (en) * | 2017-10-25 | 2018-02-09 | 中国科学院合肥物质科学研究院 | Lower extremity movement recognition methods based on multiple dimensioned Gauss Markov random field model |
| CN108268844A (en) * | 2018-01-17 | 2018-07-10 | 上海术理智能科技有限公司 | Movement recognition method and device based on surface electromyogram signal |
| 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 |
| CN108703824A (en) * | 2018-03-15 | 2018-10-26 | 哈工大机器人（合肥）国际创新研究院 | A kind of bionic hand control system and control method based on myoelectricity bracelet |
| CN109213305A (en) * | 2017-06-29 | 2019-01-15 | 沈阳新松机器人自动化股份有限公司 | A kind of gesture identification method based on surface electromyogram signal |
Family Cites Families (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7733224B2 (en) * | 2006-06-30 | 2010-06-08 | Bao Tran | Mesh network personal emergency response appliance |
| US10921886B2 (en) * | 2012-06-14 | 2021-02-16 | Medibotics Llc | Circumferential array of electromyographic (EMG) sensors |
| US20150366504A1 (en) * | 2014-06-20 | 2015-12-24 | Medibotics Llc | Electromyographic Clothing |
| CN105022471A (en) * | 2014-04-23 | 2015-11-04 | 王建勤 | Device and method for carrying out gesture recognition based on pressure sensor array |
| CN105654037B (en) * | 2015-12-21 | 2019-05-21 | 浙江大学 | A kind of electromyography signal gesture identification method based on deep learning and characteristic image |
| CN107273798A (en) * | 2017-05-11 | 2017-10-20 | 华南理工大学 | A kind of gesture identification method based on surface electromyogram signal |
| CN107590432A (en) * | 2017-07-27 | 2018-01-16 | 北京联合大学 | A kind of gesture identification method based on circulating three-dimensional convolutional neural networks |
| CN108983973B (en) * | 2018-07-03 | 2021-01-26 | 东南大学 | Control method of humanoid smart myoelectric artificial hand based on gesture recognition |
-
2019
- 2019-01-30 CN CN201910092534.4A patent/CN109800733B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109213305A (en) * | 2017-06-29 | 2019-01-15 | 沈阳新松机器人自动化股份有限公司 | A kind of gesture identification method based on surface electromyogram signal |
| CN107679516A (en) * | 2017-10-25 | 2018-02-09 | 中国科学院合肥物质科学研究院 | Lower extremity movement recognition methods based on multiple dimensioned Gauss Markov random field model |
| CN108268844A (en) * | 2018-01-17 | 2018-07-10 | 上海术理智能科技有限公司 | Movement recognition method and device based on surface electromyogram signal |
| CN108703824A (en) * | 2018-03-15 | 2018-10-26 | 哈工大机器人（合肥）国际创新研究院 | A kind of bionic hand control system and control method based on myoelectricity bracelet |
| 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 |
Non-Patent Citations (1)
| Title |
|---|
| "Semi-supervised learning for surface EMG-based gesture recognition";Yu Du;《International Joint Conference on Artificial Intelligence》;20171231;全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109800733A (en) | 2019-05-24 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Lee et al. | American sign language recognition and training method with recurrent neural network | |
| Turner et al. | A generalized, likelihood-free method for posterior estimation | |
| CN109800733B (en) | Data processing method and device and electronic equipment | |
| KR102619981B1 (en) | Gesture classification apparatus and method using electromyogram signals | |
| US8892479B2 (en) | Recognizing finger gestures from forearm EMG signals | |
| CN109635772A (en) | A kind of dictation content corrects method and electronic equipment | |
| CN108135527A (en) | For the method, apparatus and system of bio-identification | |
| EP3836836B1 (en) | Real-time spike detection and identification | |
| CN110141232B (en) | Data enhancement method for robust electromyographic signal identification | |
| KR20120052610A (en) | Apparatus and method for recognizing motion using neural network learning algorithm | |
| Khodabandelou et al. | Attention-based gated recurrent unit for gesture recognition | |
| EP2879072A1 (en) | Biometric information registration device and method | |
| CN113722474A (en) | Text classification method, device, equipment and storage medium | |
| Wibawa et al. | Gesture recognition for Indonesian Sign Language Systems (ISLS) using multimodal sensor leap motion and myo armband controllers based-on naïve bayes classifier | |
| Jaramillo-Yanez et al. | Short-term hand gesture recognition using electromyography in the transient state, support vector machines, and discrete wavelet transform | |
| Sharma et al. | On the use of temporal and spectral central moments of forearm surface EMG for finger gesture classification | |
| Ponce-López et al. | Non-verbal communication analysis in victim–offender mediations | |
| JP2007280219A (en) | Motion pattern recognition device, motion pattern recognition method, and motion pattern recognition program | |
| Singh et al. | A reliable and efficient machine learning pipeline for american sign language gesture recognition using EMG sensors | |
| WO2021047376A1 (en) | Data processing method, data processing apparatus and related devices | |
| KR102363879B1 (en) | Method for predicting clinical functional assessment scale using feature values derived by upper limb movement of patients | |
| CN113516000A (en) | Method, device and equipment for processing waveform image and storage medium | |
| Ghaderi et al. | Kernel density estimation of electromyographic signals and ensemble learning for highly accurate classification of a large set of hand/wrist motions | |
| Fonseca et al. | Artificial neural networks applied to the classification of hand gestures using eletromyographic signals | |
| Jaber et al. | Elicitation hybrid spatial features from HD-sEMG signals for robust classification of gestures in real-time |
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 |