Surface electromyography signal post-processing method and device

Abstract

An embodiment of the present invention provides a surface electromyography signal post-processing method and device, the method comprising: acquiring a first surface electromyography signal and a second surface electromyography signal, and determining a first effective potential segment in the first surface electromyography signal, and a second effective potential segment in the second surface electromyography signal; extracting multiple first features from the first effective potential segment, and extracting multiple second features from the second effective potential segment based on any time window of multiple time windows of different lengths; splicing multiple first features and multiple second features based on preset splicing rules to generate multiple feature vectors, and using a preset classifier to classify and identify each of the feature vectors; and determining a target classification recognition result corresponding to the first surface electromyography signal and the second surface electromyography signal from the classification recognition result.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent wearable devices, and in particular to a surface electromyography signal post-processing method and device. Background Art

[0002] In recent years, with the continuous development of industry and transportation, the number of amputees due to industrial production, construction projects, traffic accidents, and other reasons has been increasing year by year. To this end, intelligent wearable devices such as multi-degree-of-freedom myoelectric prostheses with bionic control functions have emerged. Multi-degree-of-freedom myoelectric prostheses with bionic control functions can, to a certain extent, help amputees better integrate into society and daily life, and therefore, multi-degree-of-freedom myoelectric prostheses with bionic control functions have gradually gained attention.

[0003] To control smart wearable devices such as multi-degree-of-freedom myoelectric prostheses, surface EMG signal processing is the most commonly used method for extracting control information. Surface EMG signals are signals of a certain length, collected by electrodes on the skin surface during muscle movement or at rest. They are characterized by low amplitude and susceptibility to interference. Common methods for processing EMG signals include threshold switching control, single-degree-of-freedom proportional control, encoding control, pattern control, and synchronous proportional control. Post-processing of surface EMG signals is often necessary to identify gestures corresponding to these signals and maintain a certain level of accuracy. Summary of the Invention

[0004] The purpose of the embodiments of the present invention is to provide a surface electromyography signal post-processing method and device to achieve the beneficial effect of identifying gestures corresponding to surface electromyography signals and improving the accuracy of gesture recognition. The specific technical solution is as follows:

[0005] In a first aspect of an embodiment of the present invention, a surface electromyography signal post-processing method is first provided, the method comprising:

[0006] Acquire a first surface electromyographic signal and a second surface electromyographic signal, and determine a first effective potential segment in the first surface electromyographic signal and a second effective potential segment in the second surface electromyographic signal;

[0007] extracting a plurality of first features from the first effective potential segment and a plurality of second features from the second effective potential segment based on any one of a plurality of time windows of different lengths;

[0008] splicing the plurality of first features and the plurality of second features based on a preset splicing rule to generate a plurality of feature vectors, and classifying and identifying each of the feature vectors using a preset classifier;

[0009] A target classification recognition result corresponding to the first surface electromyographic signal and the second surface electromyographic signal is determined from the classification recognition results.

[0010] In an optional embodiment, the extracting a plurality of first features from the first effective potential segment and the extracting a plurality of second features from the second effective potential segment based on any one of a plurality of time windows of different lengths includes:

[0011] splitting the first effective potential segment based on any one of a plurality of time windows of different lengths to obtain a plurality of first effective potential sub-segments;

[0012] For each of the first effective potential sub-segments, extracting a corresponding first feature;

[0013] splitting the second effective potential segment based on any one of a plurality of time windows of different lengths to obtain a plurality of second effective potential sub-segments;

[0014] For each of the second effective potential sub-segments, extracting a corresponding second feature;

[0015] The features include wavelength, number of zero crossings, number of slope sign changes, and average absolute value.

[0016] In an optional embodiment, determining the target classification recognition result corresponding to the first surface electromyography signal and the second surface electromyography signal from the classification recognition result includes:

[0017] For any classification recognition result, count the number of occurrences;

[0018] The classification recognition result with the largest number of occurrences is determined from the classification recognition results as the target classification recognition result corresponding to the first surface electromyography signal and the second surface electromyography signal.

[0019] In an optional embodiment, determining the first effective potential segment in the first surface electromyography signal and the second effective potential segment in the second surface electromyography signal includes:

[0020] Preprocessing the first surface electromyographic signal and the second surface electromyographic signal respectively to generate a corresponding first preprocessed surface electromyographic signal and a second preprocessed surface electromyographic signal, wherein the preprocessing includes removing power interference and bandpass filtering;

[0021] A first effective potential segment in the first preprocessed surface electromyography signal and a second effective potential segment in the second preprocessed surface electromyography signal are determined.

[0022] In an optional embodiment, determining the first effective potential segment in the first preprocessed surface electromyography signal and the second effective potential segment in the second preprocessed surface electromyography signal includes:

[0023] Correcting the first preprocessed surface electromyography signal and the second preprocessed surface electromyography signal respectively to obtain a corresponding first corrected surface electromyography signal and a second corrected surface electromyography signal;

[0024] Performing integration operations on the first corrected surface electromyography signal and the second corrected surface electromyography signal, respectively, to extract corresponding multiple first envelope signals and multiple second envelope signals;

[0025] determining a first effective potential segment in the first corrected surface electromyography signal based on a plurality of the first envelope signals;

[0026] A second effective potential segment in the second corrected surface electromyographic signal is determined based on a plurality of the second envelope signals.

[0027] In an optional embodiment, determining the first effective potential segment in the first corrected surface electromyography signal based on the plurality of first envelope signals includes:

[0028] Determining a first starting position and a first ending position of a first effective potential segment in the first corrected surface electromyography signal based on the plurality of first envelope signals;

[0029] Determining a first effective potential segment in the first corrected surface electromyography signal based on the first starting position and the first ending position;

[0030] The determining of a second effective potential segment in the second corrected surface electromyography signal based on the plurality of second envelope signals comprises:

[0031] determining a second starting position and a second ending position in the second corrected surface electromyography signal based on a plurality of the second envelope signals;

[0032] Based on the second starting position and the second ending position, a second effective potential segment in the second corrected surface electromyography signal is determined.

[0033] In an optional embodiment, the method further comprises:

[0034] extracting a plurality of third features from the first effective potential segment and a plurality of fourth features from the second effective potential segment based on a preset fixed time window;

[0035] Splicing the plurality of third features and the plurality of fourth features based on a preset splicing rule to generate a plurality of target feature vectors;

[0036] A preset classifier is used to classify and identify each target feature vector.

[0037] In a second aspect of the embodiments of the present invention, a surface electromyography signal post-processing device is further provided, the device comprising:

[0038] A signal acquisition module, configured to acquire a first surface electromyography signal and a second surface electromyography signal;

[0039] a potential segment determination module, configured to determine a first effective potential segment in the first surface electromyographic signal and a second effective potential segment in the second surface electromyographic signal;

[0040] a feature extraction module, configured to extract a plurality of first features from the first effective potential segment and a plurality of second features from the second effective potential segment based on any one of a plurality of time windows of different lengths;

[0041] a feature splicing module, configured to splice a plurality of the first features and a plurality of the second features based on a preset splicing rule to generate a plurality of feature vectors;

[0042] A separate identification module, configured to classify and identify each of the feature vectors using a preset classifier;

[0043] The classification determination module is used to determine the target classification recognition result corresponding to the first surface electromyography signal and the second surface electromyography signal from the classification recognition result.

[0044] In a third aspect of an embodiment of the present invention, there is further provided an intelligent wearable device, comprising a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other via the communication bus;

[0045] Memory for storing computer programs;

[0046] The processor is configured to implement any surface electromyography signal post-processing method described in the first aspect above when executing a program stored in the memory.

[0047] In the fourth aspect of the embodiment of the present invention, a storage medium is further provided, in which instructions are stored. When the storage medium is run on a computer, the computer executes the surface electromyography signal post-processing method described in any one of the first aspects above.

[0048] In a fifth aspect of the embodiments of the present invention, a computer program product comprising instructions is further provided, which, when executed on a computer, enables the computer to execute any of the surface electromyography signal post-processing methods described in the first aspect.

[0049] The technical solution provided by the embodiment of the present invention obtains a first surface electromyographic signal and a second surface electromyographic signal, and determines a first effective potential segment in the first surface electromyographic signal and a second effective potential segment in the second surface electromyographic signal, extracts a plurality of first features from the first effective potential segment based on any time window of a plurality of time windows of different lengths, and extracts a plurality of second features from the second effective potential segment, splices the plurality of first features and the plurality of second features based on a preset splicing rule to generate a plurality of feature vectors, classifies and identifies each feature vector using a preset classifier, and determines the target classification recognition result corresponding to the first surface electromyographic signal and the second surface electromyographic signal from the classification recognition result. In this way, the embodiment of the present invention extracts features from the effective potential segment through any time window of a plurality of time windows of different lengths, and uses a classifier to identify the splicing features, and determines the gesture action corresponding to the surface electromyographic signal from the classification recognition result, thereby achieving the beneficial effect of identifying the gesture action corresponding to the surface electromyographic signal and improving the accuracy of gesture action recognition. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the invention and, together with the description, serve to explain the principles of the invention.

[0051] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0052] Figure 1 Schematic diagram of an implementation flow of a surface electromyography signal processing method shown in an embodiment of the present invention;

[0053] Figure 2 1 is a schematic diagram of an implementation flow of determining a first effective potential segment and a second effective potential segment shown in an embodiment of the present invention;

[0054] Figure 3 Schematic diagram of a first effective potential segment in a first corrected surface electromyography signal shown in an embodiment of the present invention;

[0055] Figure 4 Schematic diagram of an implementation flow of a surface electromyography signal post-processing method shown in an embodiment of the present invention;

[0056] Figure 5 Schematic diagram of the structure of a surface electromyography signal post-processing device shown in an embodiment of the present invention;

[0057] Figure 6The figure is a schematic structural diagram of a smart wearable device shown in an embodiment of the present invention. DETAILED DESCRIPTION

[0058] To make the objectives, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts shall fall within the scope of protection of the present invention.

[0059] like Figure 1 FIG. 1 is a schematic diagram of an implementation flow of a surface electromyography signal processing method provided by an embodiment of the present invention. The method may specifically include the following steps:

[0060] S101, acquiring a first surface electromyographic signal and a second surface electromyographic signal, and determining a first effective potential segment in the first surface electromyographic signal and a second effective potential segment in the second surface electromyographic signal;

[0061] In an embodiment of the present invention, a gesture action set can be pre-set, which may include the following gesture actions: clenching fist, flexing wrist, extending wrist, extending palm, bending thumb, bending index finger, bending middle finger, bending ring finger, bending little finger, bending thumb, bending thumb and index finger combination bending, etc.

[0062] The user wears the electromyography acquisition module (armband or electromyography acquisition electrode, etc.) at the relevant position as required, and collects the first surface electromyography signal corresponding to the first position of the arm and the second surface electromyography signal corresponding to the second position for any gesture action in the gesture action set multiple times.

[0063] Specifically, the above-mentioned relevant positions may be flexor and extensor positions, and the first surface electromyographic signals corresponding to the extensor muscles and the second surface electromyographic signals corresponding to the flexor muscles in the forearm may be collected multiple times.

[0064] For example, for various gestures such as fist clenching, wrist flexion, wrist extension, palm extension, thumb flexion, index finger flexion, middle finger flexion, ring finger flexion, little finger flexion, thumb flexion, and thumb and index finger combined flexion, the fist clenching gesture is first collected six times, each time including the first surface electromyographic signal corresponding to the extensor muscle of the forearm and the second surface electromyographic signal corresponding to the flexor muscle, with an interval of 5 seconds each time. To avoid fatigue, after an interval of 30 seconds, the wrist flexion gesture is collected six times again, each time including the first surface electromyographic signal corresponding to the extensor muscle of the forearm and the second surface electromyographic signal corresponding to the flexor muscle, with a rest interval of 5 seconds, and so on.

[0065] The embodiment of the present invention can obtain the first surface electromyographic signal corresponding to the extensor muscle of the forearm and the second surface electromyographic signal corresponding to the flexor muscle collected at any time for any gesture action, which is convenient for subsequent processing.

[0066] The transmission mode of the first surface electromyography signal and the second surface electromyography signal may be wireless or wired, which is not limited in the embodiment of the present invention.

[0067] Since the surface electromyographic signal includes a resting potential segment and an action potential segment, and the action potential segment is generated when the muscle contracts, and the action potential segment is the effective potential segment, it is necessary to determine the first effective potential segment in the first surface electromyographic signal and the second effective potential segment in the second surface electromyographic signal.

[0068] Among them, Figure 2 As shown, the first effective potential segment in the first surface electromyographic signal and the second effective potential segment in the second surface electromyographic signal are determined specifically by the following method:

[0069] S201, preprocessing the first surface electromyographic signal and the second surface electromyographic signal respectively to generate a corresponding first preprocessed surface electromyographic signal and a second preprocessed surface electromyographic signal, wherein the preprocessing includes removing power interference and bandpass filtering;

[0070] For example, for a fist-clenching gesture, the first surface electromyographic signal corresponding to the extensor muscle of the forearm and the second surface electromyographic signal corresponding to the flexor muscle collected for the first time are firstly subjected to a notch filter to remove 50Hz power interference, and then are respectively subjected to 20-450Hz band-pass filtering to generate the corresponding first pre-processed surface electromyographic signal and the second pre-processed surface electromyographic signal.

[0071] S202, correcting the first preprocessed surface electromyography signal and the second preprocessed surface electromyography signal respectively to obtain a corresponding first corrected surface electromyography signal and a second corrected surface electromyography signal;

[0072] S203, performing integration operations on the first corrected surface electromyography signal and the second corrected surface electromyography signal, respectively, to extract corresponding multiple first envelope signals and multiple second envelope signals;

[0073] For any gesture action, the first surface electromyographic signal corresponding to the extensor muscle of the forearm and the second surface electromyographic signal corresponding to the flexor muscle collected at any time are preprocessed to generate the corresponding first preprocessed surface electromyographic signal and the second preprocessed surface electromyographic signal. The first preprocessed surface electromyographic signal and the second preprocessed surface electromyographic signal are corrected respectively to obtain the corresponding first corrected surface electromyographic signal and the second corrected surface electromyographic signal. The first corrected surface electromyographic signal and the second corrected surface electromyographic signal are integrated respectively to extract the corresponding multiple first envelope signals and multiple second envelope signals.

[0074] For example, for a fist-clenching gesture, the first surface electromyographic signal corresponding to the extensor muscle of the forearm and the second surface electromyographic signal corresponding to the flexor muscle collected for the first time are preprocessed to generate a first preprocessed surface electromyographic signal and a second preprocessed surface electromyographic signal. The first preprocessed surface electromyographic signal and the second preprocessed surface electromyographic signal are respectively corrected to obtain a first corrected surface electromyographic signal and a second corrected surface electromyographic signal. This is essentially to calibrate the baseline and reduce the impact of individual differences on the overall situation.

[0075] Then, integration operations are performed on the first corrected surface electromyography signal and the second corrected surface electromyography signal, respectively, to extract corresponding first envelope signals and second envelope signals, as shown below:

[0076] Assuming that the signal range of the first corrected surface electromyographic signal is 1-1000ms, an integration operation is performed on the interval 1-100ms to extract the first first envelope signal, an integration operation is performed on the interval 2-101ms to extract the second first envelope signal, an integration operation is performed on the interval 3-102ms to extract the third first envelope signal, and so on;

[0077] Assuming that the signal range of the second corrected surface electromyography signal is 1-1000ms, integration operation is performed on the interval 1-100ms to extract the first second envelope signal, integration operation is performed on the interval 2-101ms to extract the second second envelope signal, integration operation is performed on the interval 3-102ms to extract the third second envelope signal, and so on.

[0078] S204, determining a first effective potential segment in the first corrected surface electromyography signal based on the plurality of first envelope signals;

[0079] S205: Determine a second effective potential segment in the second corrected surface electromyography signal based on the plurality of second envelope signals.

[0080] In an embodiment of the present invention, a first effective potential segment in a first corrected surface electromyographic signal may be determined based on multiple first envelope signals, and a second effective potential segment in a second corrected surface electromyographic signal may be determined based on multiple second envelope signals.

[0081] Specifically, the embodiment of the present invention can determine the first effective potential segment in the first corrected surface electromyography signal and the second effective potential segment in the second corrected surface electromyography signal in the following manner:

[0082] Based on multiple first envelope signals, determine the first starting position and the first ending position of the first effective potential segment in the first corrected surface electromyographic signal; based on the first starting position and the first ending position, determine the first effective potential segment in the first corrected surface electromyographic signal; based on multiple second envelope signals, determine the second starting position and the second ending position in the second corrected surface electromyographic signal; based on the second starting position and the second ending position, determine the second effective potential segment in the second corrected surface electromyographic signal.

[0083] For example, for a fist-clenching gesture, the first surface electromyographic signal corresponding to the extensor muscle of the forearm and the second surface electromyographic signal corresponding to the flexor muscle collected for the first time are preprocessed to generate the corresponding first preprocessed surface electromyographic signal and the second preprocessed surface electromyographic signal. The first preprocessed surface electromyographic signal and the second preprocessed surface electromyographic signal are corrected respectively to obtain the corresponding first corrected surface electromyographic signal A and the second corrected surface electromyographic signal B. The first corrected surface electromyographic signal A and the second corrected surface electromyographic signal B are integrated respectively to extract the corresponding multiple first envelope signals and multiple second envelope signals, as shown in Table 1 below.

[0084]

[0085] Table 1

[0086] Determine the first effective potential segment in the first corrected surface electromyography signal A based on the multiple first envelope signals:

[0087] 1. First, determine whether the amplitude of the first envelope signal is greater than the first threshold. If the amplitude of the first envelope signal is not greater than the first threshold, continue to determine whether the amplitude of the second envelope signal is greater than the first threshold. If the amplitude of the second envelope signal is greater than the first threshold (i.e., this is the first time that the amplitude of the first envelope signal is greater than the first threshold), then determine the first starting position of the action potential segment, i.e., the position corresponding to the 101st electromyographic data from 2 to 101 ms participating in the integration operation of the second envelope signal.

[0088] 2. After determining the first starting position of the action potential segment, it is possible to determine whether the amplitude of the third first envelope signal is less than the first threshold (i.e., no longer determine whether the amplitude of the third first envelope signal is greater than the first threshold). If the amplitude of the third first envelope signal is not less than the first threshold, then continue to determine whether the amplitude of the fourth first envelope signal is less than the first threshold. If the amplitude of the fourth first envelope signal is not less than the first threshold, then continue to determine whether the amplitude of the fifth first envelope signal is less than the first threshold, and so on, until the amplitude of the Nth first envelope signal is less than the first threshold (i.e., after the second first envelope signal, the amplitude of the first envelope signal is less than the first threshold for the first time). The first ending position of the action potential segment can be determined, i.e., the corresponding position of the Nth electromyographic data from N-99 to N that participates in the integration operation of the Nth first envelope signal.

[0089] 3. Based on the first starting position and the first ending position, determine the first effective potential segment in the first corrected surface electromyographic signal A, such as Figure 3 Shown is the range from 101 to N.

[0090] The process of determining the second effective potential segment in the second corrected surface electromyography signal B based on multiple second envelope signals is similar to the process of determining the first effective potential segment in the second corrected surface electromyography signal A based on the above-mentioned multiple first envelope signals, and the embodiments of the present invention will not be described one by one here.

[0091] S102, extracting a plurality of third features from the first effective potential segment and a plurality of fourth features from the second effective potential segment based on a preset fixed time window;

[0092] Through the above steps, a first effective potential segment and a second effective potential segment can be determined. Based on a preset fixed time window, multiple third features can be extracted from the first effective potential segment, and multiple fourth features can be extracted from the second effective potential segment. The features include wavelength, number of zero crossings, number of slope sign changes, and average absolute value.

[0093] Specifically, based on the preset fixed time window, the first effective potential segment is split to obtain multiple first effective potential sub-segments, and for each first effective potential sub-segment, the corresponding third feature is extracted. Based on the preset fixed time window, the second effective potential segment is split to obtain multiple second effective potential sub-segments, and for each second effective potential sub-segment, the corresponding fourth feature is extracted.

[0094] For example, the first effective potential segment in the first corrected surface electromyographic signal A is used as an example. The signal interval includes 1-1000ms, and the preset fixed time window is 100. This can be divided into 10 first effective potential sub-segments: 1-100ms, 101-200ms, 201-300ms, etc. For each first effective potential sub-segment, the corresponding first feature is extracted, which specifically includes the following features:

[0095] A. Wavelength

[0096]

[0097] The waveform length WL is the simple accumulation of the signal lengths at point K, reflecting the complexity of the EMG signal waveform and the result of the combined effects of the EMG signal amplitude, frequency, and duration. The value range of K can be 1-100ms, 101-200ms, 201-300ms, etc., the interval of each first action potential subsegment.

[0098] B. Zero crossing number

[0099] x i x i+1 ≤0,|x i -x i+1 |≥ε;

[0100] A simple frequency statistic specialization that counts the number of times a signal waveform passes through the time axis (that is, zero) over a period of time. Given two adjacent samples x i x i+1 , when the above conditions are met, the number of zero-crossing points is increased by 1, and the corresponding statistical interval of the number of zero-crossing points can be the interval of each first action potential sub-segment such as 1-100ms, 101-200ms, 201-300ms...

[0101] C. Number of slope sign changes

[0102] (x i+1 -x i )*(x i -x i-1 )≤0,|x i -x i+1 |≥ε,|x i -x i-1 |≥ε;

[0103] This statistical feature is another feature that describes the frequency information of the signal. Given three consecutive sample values, x i+1 ,x i ,x i-1If the above conditions are met, the value of the change number is increased by 1, and the corresponding statistical interval of the slope sign change number can be 1-100ms, 101-200ms, 201-300ms..., etc., the interval of each first action potential sub-segment.

[0104] D. Average absolute value

[0105]

[0106] The statistical interval may be 1-100ms, 101-200ms, 201-300ms, etc., the interval of each first action potential sub-segment.

[0107] In this way, the above four features A, B, C, and D are extracted for each first effective potential sub-segment, and the same is true for each second effective potential sub-segment. The above four features A, B, C, and D are extracted for each second effective potential sub-segment.

[0108] S103, splicing the plurality of third features and the plurality of fourth features based on a preset splicing rule to generate a plurality of target feature vectors;

[0109] For each first effective potential sub-segment, the above four features A, B, C, and D are extracted; for each second effective potential sub-segment, the above four features A, B, C, and D are extracted. Thus, each third feature includes the above four features A, B, C, and D; and each fourth feature includes the above four features A, B, C, and D. Multiple third features and multiple fourth features can be spliced ​​together based on preset splicing rules to generate multiple target feature vectors.

[0110] Specifically, the plurality of third features and the plurality of fourth features are spliced ​​one-to-one based on a preset splicing rule to generate a plurality of target feature vectors.

[0111] For example, for the first third feature and the first fourth feature, direct concatenation can generate the first target feature vector, for the second third feature and the second fourth feature, direct concatenation can generate the second target feature vector, for the third third feature and the third fourth feature, direct concatenation can generate the third target feature vector, and so on, multiple target feature vectors can be obtained, and you can use x1, x2, x3, ... x N Indicates that each target feature vector includes eight features.

[0112] S104: using a preset classifier to classify and identify each of the target feature vectors.

[0113] For multiple target feature vectors, you can use x1,x2,x3,…x NIndicates that each target feature vector is classified and identified using a preset classifier. Specifically, each target feature vector can be classified and identified using a K-nearest neighbor classifier.

[0114] like Figure 4 FIG. 1 is a schematic diagram of an implementation flow of a surface electromyography signal post-processing method provided by an embodiment of the present invention. The method may specifically include the following steps:

[0115] S401, acquiring a first surface electromyographic signal and a second surface electromyographic signal, and determining a first effective potential segment in the first surface electromyographic signal and a second effective potential segment in the second surface electromyographic signal;

[0116] In the embodiment of the present invention, this step is similar to the above-mentioned step S101, and the embodiment of the present invention will not be described in detail here.

[0117] S402, extracting a plurality of third features from the first effective potential segment and a plurality of fourth features from the second effective potential segment based on a preset fixed time window;

[0118] In the embodiment of the present invention, this step is similar to the above-mentioned step S102, and the embodiment of the present invention will not be described in detail here.

[0119] S403, splicing the plurality of third features and the plurality of fourth features based on a preset splicing rule to generate a plurality of target feature vectors;

[0120] In the embodiment of the present invention, this step is similar to the above-mentioned step S103, and the embodiment of the present invention will not be described in detail here.

[0121] S404, using a preset classifier to classify and identify each of the target feature vectors;

[0122] In the embodiment of the present invention, this step is similar to the above-mentioned step S104, and the embodiment of the present invention will not be described in detail here.

[0123] S405, extracting a plurality of first features from the first effective potential segment and a plurality of second features from the second effective potential segment based on any one of a plurality of time windows of different lengths;

[0124] In an embodiment of the present invention, in order to improve the accuracy of gesture recognition, multiple first features can be extracted from the first effective potential segment and multiple second features can be extracted from the second effective potential segment based on any time window of multiple time windows of different lengths.

[0125] Specifically, based on any one of a plurality of time windows of different lengths, the first effective potential segment is split to obtain a plurality of first effective potential sub-segments; for each of the first effective potential sub-segments, the corresponding first feature is extracted; based on any one of a plurality of time windows of different lengths, the second effective potential segment is split to obtain a plurality of second effective potential sub-segments; for each of the second effective potential sub-segments, the corresponding second feature is extracted; wherein, the features include wavelength, number of zero crossing points, number of slope sign changes, and average absolute value.

[0126] For example, taking the first effective potential segment in the first corrected surface electromyography signal A as an example, its signal interval includes 1-1000ms, and multiple time windows WS of different lengths can be pre-set: 100, 150, 200, 250. Based on any time window of the multiple time windows of different lengths, the first effective potential segment is split, as shown in Table 2 below.

[0127]

[0128]

[0129] Table 2

[0130] Based on the time window 100, the first effective potential segment is split to obtain multiple first effective potential sub-segments, as shown in Table 2 above. For each first effective potential sub-segment, the corresponding first feature is extracted, specifically including the wavelength, the number of zero crossing points, the number of slope sign changes, and the average absolute value;

[0131] Based on the time window 150, the first effective potential segment is split to obtain multiple first effective potential sub-segments, as shown in Table 2 above. For each first effective potential sub-segment, the corresponding first feature is extracted, specifically including the wavelength, the number of zero crossing points, the number of slope sign changes, and the average absolute value;

[0132] Based on the time window 200, the first effective potential segment is split to obtain multiple first effective potential sub-segments, as shown in Table 2 above. For each first effective potential sub-segment, the corresponding first feature is extracted, specifically including the wavelength, the number of zero crossing points, the number of slope sign changes, and the average absolute value;

[0133] Based on the time window 250, the first effective potential segment is split to obtain multiple first effective potential sub-segments, as shown in Table 2 above. For each first effective potential sub-segment, the corresponding first feature is extracted, specifically including the wavelength, the number of zero crossing points, the number of slope sign changes, and the average absolute value.

[0134] Based on any one of multiple time windows of different lengths, the second effective potential segment is split to obtain multiple second effective potential sub-segments. For each of the second effective potential sub-segments, the corresponding second feature is extracted. The processing flow is similar to the above, and the embodiments of the present invention will not be repeated here.

[0135] S406, combining the first features and the second features based on a preset combination rule to generate a plurality of feature vectors, and classifying and identifying each of the feature vectors using a preset classifier;

[0136] In an embodiment of the present invention, a plurality of first features and a plurality of second features are spliced ​​together based on a preset splicing rule to generate a plurality of feature vectors, wherein the plurality of feature vectors can be spliced ​​together in a one-to-one correspondence.

[0137] It should be noted that, in the process of splicing and generating multiple feature vectors, multiple first features and multiple second features are spliced ​​together based on preset splicing rules to generate multiple feature vectors within the same time window.

[0138] For example, within a time window 100, the first effective potential segment is split to obtain multiple first effective potential sub-segments 100. For each first effective potential sub-segment 100, a corresponding first feature 100 is extracted. The second effective potential segment is split to obtain multiple second effective potential sub-segments 100. For each second effective potential sub-segment 100, a corresponding second feature 100 is extracted. The first features 100 and the second features 100 are concatenated in a one-to-one correspondence to generate a feature vector 100.

[0139] In a time window 150, the first effective potential segment is split to obtain a plurality of first effective potential sub-segments 150. For each first effective potential sub-segment 150, a corresponding first feature 150 is extracted. The second effective potential segment is split to obtain a plurality of second effective potential sub-segments 150. For each second effective potential sub-segment 150, a corresponding second feature 150 is extracted. The first features 150 and the second features 150 are concatenated in a one-to-one correspondence to generate a feature vector 150.

[0140] The processing flows within the time windows 200 and 250 are similar, and the embodiments of the present invention will not be described in detail here.

[0141] After the above steps, multiple feature vectors can be obtained. For the multiple feature vectors, a classifier can be preset to classify and identify each feature vector. Specifically, a K-nearest neighbor classifier can be used to classify and identify each feature vector.

[0142] S407: Determine target classification recognition results corresponding to the first surface electromyography signal and the second surface electromyography signal from the classification recognition results.

[0143] In an embodiment of the present invention, a target classification recognition result corresponding to the first surface electromyographic signal and the second surface electromyographic signal is determined from the classification recognition result, wherein the classification recognition result is a result of classifying and recognizing each of the feature vectors using a preset classifier.

[0144] In this embodiment of the present invention, the target classification recognition result corresponding to the first surface electromyography signal and the second surface electromyography signal can be determined from the classification recognition results based on a voting strategy:

[0145] Count the number of classification recognition results, and for any classification recognition result, count the number of times the classification result appears, and determine from the classification recognition results that the classification recognition result with the largest number of occurrences is the target classification recognition result corresponding to the first surface electromyography signal and the second surface electromyography signal.

[0146] For example, the following formula is used to count the number of classification recognition results:

[0147]

[0148] in, DS is the first effective potential segment or the second effective potential segment (1000ms), K is the number of preset time windows of different lengths, ws j , wi j are the length and step of the j-th time window respectively.

[0149] For any classification recognition result, the number of times the classification result appears is counted, and a majority vote is performed to determine the classification recognition result with the largest number of occurrences from the classification recognition results as the target classification recognition result corresponding to the first surface electromyography signal and the second surface electromyography signal.

[0150] Through the above description of the technical solution provided by the embodiment of the present invention, by obtaining a first surface electromyography signal and a second surface electromyography signal, and determining the first effective potential segment in the first surface electromyography signal and the second effective potential segment in the second surface electromyography signal, based on any time window of multiple time windows of different lengths, extracting multiple first features from the first effective potential segment, and extracting multiple second features from the second effective potential segment, the multiple first features and the multiple second features are spliced ​​together based on a preset splicing rule to generate multiple feature vectors, and each feature vector is classified and identified using a preset classifier, and the target classification recognition result corresponding to the first surface electromyography signal and the second surface electromyography signal is determined from the classification recognition result. In this way, the embodiment of the present invention extracts features from the effective potential segment through any time window of multiple time windows of different lengths, and uses a classifier to identify the splicing features, and determines the gesture action corresponding to the surface electromyography signal from the classification recognition result, which can achieve the beneficial effect of identifying the gesture action corresponding to the surface electromyography signal and improving the accuracy of gesture action recognition.

[0151] Corresponding to the above method embodiment, the embodiment of the present invention also provides a surface electromyography signal post-processing device, such as Figure 5 As shown, the device may include: a signal acquisition module 510, a potential segment determination module 520, a feature extraction module 530, a feature splicing module 540, a separate identification module 550, and a classification determination module 560.

[0152] A signal acquisition module 510 is configured to acquire a first surface electromyography signal and a second surface electromyography signal;

[0153] A potential segment determining module 520 is configured to determine a first effective potential segment in the first surface electromyographic signal and a second effective potential segment in the second surface electromyographic signal;

[0154] a feature extraction module 530 for extracting a plurality of first features from the first effective potential segment and a plurality of second features from the second effective potential segment based on any one of a plurality of time windows of different lengths;

[0155] A feature splicing module 540 is configured to splice a plurality of the first features and a plurality of the second features based on a preset splicing rule to generate a plurality of feature vectors;

[0156] A separate identification module 550 is used to classify and identify each of the feature vectors using a preset classifier;

[0157] The classification determination module 560 is configured to determine, from the classification recognition results, target classification recognition results corresponding to the first surface electromyography signal and the second surface electromyography signal.

[0158] The embodiment of the present invention also provides a smart wearable device, such as Figure 6 As shown, it includes a processor 61, a communication interface 62, a memory 63 and a communication bus 64, wherein the processor 61, the communication interface 62, and the memory 63 communicate with each other through the communication bus 64.

[0159] Memory 63, for storing computer programs;

[0160] The processor 61 is configured to execute the program stored in the memory 63 by performing the following steps:

[0161] Acquire a first surface electromyographic signal and a second surface electromyographic signal, and determine a first effective potential segment in the first surface electromyographic signal and a second effective potential segment in the second surface electromyographic signal; extract a plurality of first features from the first effective potential segment and a plurality of second features from the second effective potential segment based on any time window of a plurality of time windows of different lengths; splice a plurality of the first features and a plurality of the second features based on a preset splicing rule to generate a plurality of feature vectors, and use a preset classifier to classify and identify each of the feature vectors; determine a target classification recognition result corresponding to the first surface electromyographic signal and the second surface electromyographic signal from the classification recognition result.

[0162] The communication bus mentioned in the smart wearable device above can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus. This communication bus can be divided into an address bus, a data bus, a control bus, etc. For ease of illustration, only one thick line is used in the figure, but this does not mean that there is only one bus or only one type of bus.

[0163] The communication interface is used for communication between the above-mentioned smart wearable device and other devices.

[0164] The memory may include random access memory (RAM) or non-volatile memory, such as at least one disk storage. Alternatively, the memory may be at least one storage device located away from the processor.

[0165] The above-mentioned processor can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, and discrete hardware components.

[0166] In another embodiment of the present invention, a storage medium is provided. The storage medium stores instructions that, when executed on a computer, enable the computer to execute the surface electromyography signal post-processing method described in any one of the above embodiments.

[0167] In another embodiment of the present invention, a computer program product including instructions is provided. When the computer program product is run on a computer, the computer executes the surface electromyography signal post-processing method described in any one of the above embodiments.

[0168] In the above embodiments, all or part of the embodiments can be implemented by software, hardware, firmware, or any combination thereof. When implemented using software, all or part of the embodiments can be implemented in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the processes or functions described in the embodiments of the present invention are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a storage medium or transmitted from one storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via a wired (e.g., coaxial cable, optical fiber, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) method. The storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server or data center that includes one or more available media. The available medium can be a magnetic medium (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium (e.g., a solid-state drive (SSD)).

[0169] It should be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply the existence of any such actual relationship or order between these entities or operations. Moreover, the terms "comprises," "comprising," or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article, or device comprising a series of elements includes not only those elements, but also other elements not explicitly listed, or elements inherent to such process, method, article, or device. In the absence of further limitations, an element defined by the phrase "comprising a ..." does not exclude the presence of other identical elements in the process, method, article, or device comprising the element.

[0170] Each embodiment in this specification is described in a related manner. Similar parts between the various embodiments can be referred to in conjunction with each other. Each embodiment focuses on the differences between the other embodiments. In particular, the system embodiment is generally similar to the method embodiment, so the description is relatively simple. For related parts, refer to the description of the method embodiment.

[0171] The above description is only a preferred embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present invention are included in the scope of protection of the present invention.

Claims

1. A surface electromyography signal post-processing method, characterized in that: The method comprises: Obtain a first surface electromyographic signal corresponding to an extensor position and a second surface electromyographic signal corresponding to a flexor position, and determine a first effective potential segment in the first surface electromyographic signal and a second effective potential segment in the second surface electromyographic signal; determining the first effective potential segment in the first surface electromyographic signal comprises: performing preprocessing, correction, and integration operations on the first surface electromyographic signal, extracting a plurality of corresponding first envelope signals, taking the first appearance of a first envelope signal greater than a first threshold as a first starting position, taking the first appearance of a first envelope signal less than the first threshold as a first ending position, and determining the first effective potential segment based on the first starting position and the first ending position; extracting a plurality of first features from the first effective potential segment and a plurality of second features from the second effective potential segment based on any one of a plurality of time windows of different lengths; splicing the plurality of first features and the plurality of second features based on a preset splicing rule to generate a plurality of feature vectors, and classifying and identifying each of the feature vectors using a preset classifier; Determining the target classification recognition result corresponding to the first surface electromyographic signal and the second surface electromyographic signal from the classification recognition results, including: counting the number of occurrences for any classification recognition result; and determining the classification recognition result with the largest number of occurrences from the classification recognition results as the target classification recognition result corresponding to the first surface electromyographic signal and the second surface electromyographic signal.

2. The method according to claim 1, characterized in that The extracting of a plurality of first features from the first effective potential segment and a plurality of second features from the second effective potential segment based on any one of a plurality of time windows of different lengths comprises: splitting the first effective potential segment based on any one of a plurality of time windows of different lengths to obtain a plurality of first effective potential sub-segments; For each of the first effective potential sub-segments, extracting a corresponding first feature; splitting the second effective potential segment based on any one of a plurality of time windows of different lengths to obtain a plurality of second effective potential sub-segments; For each of the second effective potential sub-segments, extracting a corresponding second feature; The features include wavelength, number of zero crossings, number of slope sign changes, and average absolute value.

3. The method according to claim 1, characterized in that Determining a first effective potential segment in the first surface electromyographic signal and a second effective potential segment in the second surface electromyographic signal includes: Preprocessing the first surface electromyographic signal and the second surface electromyographic signal respectively to generate a corresponding first preprocessed surface electromyographic signal and a second preprocessed surface electromyographic signal, wherein the preprocessing includes removing power interference and bandpass filtering; A first effective potential segment in the first preprocessed surface electromyography signal and a second effective potential segment in the second preprocessed surface electromyography signal are determined.

4. The method according to claim 3, characterized in that The determining of a first effective potential segment in the first preprocessed surface electromyography signal and a second effective potential segment in the second preprocessed surface electromyography signal comprises: Correcting the first preprocessed surface electromyography signal and the second preprocessed surface electromyography signal respectively to obtain a corresponding first corrected surface electromyography signal and a second corrected surface electromyography signal; Performing integration operations on the first corrected surface electromyography signal and the second corrected surface electromyography signal, respectively, to extract corresponding multiple first envelope signals and multiple second envelope signals; determining a first effective potential segment in the first corrected surface electromyography signal based on a plurality of the first envelope signals; A second effective potential segment in the second corrected surface electromyographic signal is determined based on a plurality of the second envelope signals.

5. The method according to claim 4, characterized in that The determining of a first effective potential segment in the first corrected surface electromyography signal based on the plurality of first envelope signals comprises: Determining a first starting position and a first ending position of a first effective potential segment in the first corrected surface electromyography signal based on the plurality of first envelope signals; Determining a first effective potential segment in the first corrected surface electromyography signal based on the first starting position and the first ending position; The determining of a second effective potential segment in the second corrected surface electromyography signal based on the plurality of second envelope signals comprises: determining a second starting position and a second ending position in the second corrected surface electromyography signal based on a plurality of the second envelope signals; Based on the second starting position and the second ending position, a second effective potential segment in the second corrected surface electromyography signal is determined.

6. The method according to any one of claims 1 to 5, characterized in that The method further comprises: extracting a plurality of third features from the first effective potential segment and a plurality of fourth features from the second effective potential segment based on a preset fixed time window; Splicing the plurality of third features and the plurality of fourth features based on a preset splicing rule to generate a plurality of target feature vectors; A preset classifier is used to classify and identify each target feature vector.

7. A surface electromyography signal post-processing device, characterized in that: The device comprises: A signal acquisition module, configured to acquire a first surface electromyographic signal corresponding to an extensor muscle position and a second surface electromyographic signal corresponding to a flexor muscle position; A potential segment determination module is used to determine a first effective potential segment in the first surface electromyographic signal and a second effective potential segment in the second surface electromyographic signal; determining the first effective potential segment in the first surface electromyographic signal includes: preprocessing, correcting, and integrating the first surface electromyographic signal, extracting multiple corresponding first envelope signals, taking the first appearance of the first envelope signal greater than a first threshold as a first starting position, taking the first appearance of the first envelope signal less than the first threshold as a first ending position, and determining the first effective potential segment based on the first starting position and the first ending position; a feature extraction module, configured to extract a plurality of first features from the first effective potential segment and a plurality of second features from the second effective potential segment based on any one of a plurality of time windows of different lengths; a feature splicing module, configured to splice a plurality of the first features and a plurality of the second features based on a preset splicing rule to generate a plurality of feature vectors; A separate identification module, configured to classify and identify each of the feature vectors using a preset classifier; A classification determination module is used to determine the target classification recognition result corresponding to the first surface electromyographic signal and the second surface electromyographic signal from the classification recognition results, including: counting the number of occurrences of any classification recognition result; and determining the classification recognition result with the largest number of occurrences from the classification recognition results as the target classification recognition result corresponding to the first surface electromyographic signal and the second surface electromyographic signal.

8. A smart wearable device, characterized in that: It includes a processor, a communication interface, a memory and a communication bus, wherein the processor, the communication interface and the memory communicate with each other via the communication bus; Memory for storing computer programs; A processor, configured to implement the method steps described in any one of claims 1 to 6 when executing a program stored in a memory.

9. A storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the method according to any one of claims 1 to 6 is implemented.

Images