CN110151175A - Surface electromyogram signal noise-eliminating method based on CEEMD and improvement wavelet threshold - Google Patents
Surface electromyogram signal noise-eliminating method based on CEEMD and improvement wavelet threshold Download PDFInfo
- Publication number
- CN110151175A CN110151175A CN201910285085.5A CN201910285085A CN110151175A CN 110151175 A CN110151175 A CN 110151175A CN 201910285085 A CN201910285085 A CN 201910285085A CN 110151175 A CN110151175 A CN 110151175A
- Authority
- CN
- China
- Prior art keywords
- component
- signal
- noise
- wavelet
- coefficient
- 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.)
- Pending
Links
Classifications
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/24—Detecting, measuring or recording bioelectric or biomagnetic signals of the body or parts thereof
- A61B5/316—Modalities, i.e. specific diagnostic methods
- A61B5/389—Electromyography [EMG]
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
- A61B5/7203—Signal processing specially adapted for physiological signals or for diagnostic purposes for noise prevention, reduction or removal
Abstract
The present invention relates to a kind of based on CEEMD and improves the surface electromyogram signal noise-eliminating method of wavelet threshold.First surface electromyogram signal is decomposed to obtain intrinsic mode function component with complementation set empirical mode decomposition.Then suitable intrinsic mode function component is selected by component correlation analysis, wavelet threshold processing is improved to each intrinsic mode function component selected.Finally, signal treated by being modified wavelet threshold intrinsic mode function component and be not modified the intrinsic mode function component of wavelet threshold processing and carry out signal reconstruction, the signal after being denoised.The present invention has adaptivity in terms of signal processing, it is suitable for the analysis of non-linear non-stationary surface electromyogram signal, it can reduce since modal overlap bring adversely affects, and it is as much as possible to remain information useful in signal, reduce the influence of noise bring, it is demonstrated experimentally that electromyography signal denoising method proposed by the present invention has better effect than other denoising methods.
Description
Technical field
The invention belongs to signal noise silencing fields, are related to a kind of surface electromyogram signal based on CEEMD and improvement wavelet threshold
Noise-eliminating method.
Background technique
Surface electromyogram signal (Surface electromyography, sEMG) is faint by one kind of acquisition electrode acquisition
Bioelectrical signals, this bioelectrical signals are able to reflect the relevant information of muscle and the behavior act of people, have been widely used
In fields such as sports medical science, rehabilitation training and Mechanical courses.Surface electromyogram signal is mainly distributed between 10Hz-500Hz, width
Degree is only 1 μ V, while also having non-linear, non-stationary property.Therefore, electromyography signal is easy the pollution by noise.Noise
Main source includes three kinds: Interference from the power supply wire, white Gaussian noise and baseline drift.Therefore, keep signal it is pure be analysis and
The prerequisite of application surface electromyography signal.
The method of currently used surface electromyogram signal denoising mainly has Fourier transformation, wavelet transformation and empirical modal point
Solution.
1, wavelet transformation
Wavelet transformation is the data analysing method that pure mathematics and applied mathematics combine, and is widely used in signal processing, figure
As analysis various aspects.The method of Wavelet Denoising Method is totally divided into three kinds: the first is the modulus maxima denoising based on wavelet transformation,
On different scale, signal is different from the propagation of the modulus maxima of noise to utilize this characteristic, removes in its whole modulus maximum
The wavelet modulus maxima of noise, and the modulus maximum of useful signal is selected, last reconstruction signal is become with remaining small echo
Mold changing maximum obtains；Second is the denoising method based on correlation between wavelet transform dimension, and principle is wavelet transformation
Each scale between, noise does not have an apparent correlation, and signal has stronger correlation；, the wavelet transformation of noise mainly concentrates
In each level of small scale, however signal still has very strong correlation in edge.Signal to be treated is carried out small
After wave conversion, the wavelet coefficient under different scale is obtained, calculates the correlation between adjacent scale, then the size of correlation
Wavelet coefficient is selected, finally reconstruct obtains signal；The third wavelet threshold denoising method is to be proposed earliest by Donoho,
Its theoretical foundation is that the coefficient amplitude of the noise after wavelet decomposition is smaller than the wavelet coefficient amplitude of signal, and a threshold value is arranged,
When wavelet coefficient absolute value is greater than threshold value, retains (hard threshold method) or shrink (Soft thresholding) wavelet coefficient, work as wavelet systems
Number is then all set to zero when being less than threshold value, finally utilizes the wavelet coefficient reconstruction signal after threshold process.
Since wavelet transformation has the characteristics that time-frequency, multiresolution, can be applied in electromyography signal analysis.However it is small
The effect of wave denoising is directly related with the wavelet function of selection, once while wavelet function it is chosen after, analyzed to signal
During cannot be changed, therefore wavelet transformation adaptivity is poor.
2, empirical mode decomposition (Empirical Mode Decomposition, EMD) method, is mentioned by Huang earliest
Out, it is multiple intrinsic mode functions (IMF) superposition that this algorithm principle, which is by signal decomposition, and each IMF representation signal is not
With the feature under scale, there is good adaptivity.Due to EMD algorithm modal overlap the shortcomings that, Wu and Huang are proposed
Gather empirical mode decomposition (EEMD).EEMD is the parser of data-driven, is added to white noise before signal decomposition
In signals and associated noises.But synchronous signal also will receive the pollution of residual noise.In order to solve this problem, Yeh proposes one kind
New noise aided analysis method, complementation set empirical mode decomposition (CEEMD), wherein aid in noise is with a pair of positive and negative opposite
Complementary form is added in the signal of Noise after multiple averaging carries out EMD decomposition again.Complementation set empirical mode decomposition energy
Noise is enough efficiently reduced, the loss of useful information and the appearance of modal overlap are avoided.
Summary of the invention
The purpose of the present invention is to the deficiencies in the prior art, propose it is a kind of based on complementary set empirical mode decomposition with
Improve the surface electromyogram signal noise-eliminating method of wavelet threshold.The essence of complementary overall experience mode decomposition is a pair by addition
The signal of complementary noise reduces the signal analysis method of remaining aid in noise.The noise number being added is enabled to subtract in this way
Few, the calculating time greatly reduces.Improving wavelet threshold can be improved the effect of signal denoising simultaneously.
The new denoising method of one kind based on complementary set empirical mode decomposition, by component correlation analysis and improved small echo
Threshold application is in surface electromyogram signal denoising.Surface electromyogram signal is carried out with complementation set empirical mode decomposition first
Decomposition obtains intrinsic mode function component.Then suitable intrinsic mode function component is selected by component correlation analysis, therefore
Each selected intrinsic mode function passes through improvement wavelet threshold and is handled.Finally, signal is by treated natural mode of vibration
Function component and the intrinsic mode function component for not being modified wavelet threshold processing carry out signal reconstruction.With empirical modal before
Decomposition is compared with the method for the denoising of empirical mode decomposition interval threshold method, and the method proposed has surface electromyogram signal
Better denoising performance.
In order to achieve the goal above, the method for the present invention mainly comprises the steps that
1. a kind of surface electromyogram signal noise-eliminating method based on complementary set empirical mode decomposition and improvement wavelet threshold,
It is characterized in that this method comprises the following steps:
The surface electromyogram signal x (t) of step (1), acquisition can contain noise n (t):
X (t)=s (t)+δ n (t)
In formula, s (t) is the surface electromyogram signal of not Noise, and x (t) is noise-containing surface electromyogram signal, and n (t) is
White noise, δ are the scale factor of noise；
Step (2) carries out complementary set empirical mode decomposition to the resulting signals and associated noises x (t) of step (1), will contain first
A pair of complementary positive and negative opposite white noise of noise cancellation signal x (t) addition, obtains the signal x that positive aid in noise is added+Auxiliary is born with being added
The signal x of noise-, decomposition two groups of set intrinsic mode function components of generation, which are carried out, using empirical mode decomposition algorithm is denoted asWithI-th of IMF component that i is；Average statistical using uncorrelated random sequence is 0；
Step (3), repeats step (2) M times, finally obtains the mean value C of IMFi(t)；
Step (4) finds intrinsic mode function IMF points for capableing of representation signal main feature using component correlation analysis
Amount handles the improvement wavelet threshold that the component of selection carries out next step, the specific steps are as follows:
1) x (t) and each component C are calculatedi(t) variance between:
C in formulaxCIt (i) is x (t) and Ci(t) variance, uxFor the mean value of x (t), uCiFor Ci(t) mean value；
2) related coefficient is calculated:
ρ in formulaxCIt (i) is x (t) and Ci(t) related coefficient, σxAnd σCiRespectively represent x (t) and Ci(t) standard deviation；
3) it calculates and refers to related coefficient, J is to calculate as follows with reference to related coefficient:
N is the number of plies for decomposing obtained IMF in formula；Find correlation coefficient ρxC(i) it is greater than the component with reference to related coefficient J,
The ρ of componentxC(i) when being greater than with reference to related coefficient J, the property of the component energy representation signal at place, i.e. component Ci(t) it is chosen
It selects；
Step (5) is handled the component selected using wavelet thresholding methods are improved, to each selected component
Ci(t) following processing is done:
1) it is changed commanders C using discrete wavelet transformeri(t) it is decomposed into multilayer, is defined " peak and the ratio " of detail coefficients, specific as follows:
Wherein wjFor the wavelet coefficient of jth layer, wherein wj,iIt is the wavelet coefficient of j layers of i point；If Sj≤ 0.2 < Sj+1, then
Select j thus IMF carry out wavelet transformation Decomposition order；
2) the bound λ of threshold value is selectedLAnd λHHandle each layer of wavelet coefficient,
λj,L=μj-κj,Lσj
λj,H=μj+κj,Hσj
Wherein λj,LAnd λj,HIt is the threshold value bound of jth layer；κj,LAnd κj,HIt is adjustable parameter；μjAnd σjIt is jth layer wavelet systems
Several mean values and variance；
Then, κ is calculatedj,LAnd κj,HMinimum value obtain the exact value of κ；
If Sj≤ 0.01, then it is arrangedWithOtherwise,
Sr,LAnd Sr,HIt is defined asWithSj,LAnd Sj,HIt is wavelet coefficient respectively
The peak and value of positive negative part, wherein L < k < H；
After each layer of threshold value has been determined, wavelet coefficient is handled as follows:
3) with treated wavelet coefficient reconstructs the IMF that is exactly that treated, i.e.,
Step (6) obtains processed part component with step (5) and is denoted asWith the C being not handled byiCarry out signal weight
Structure, wherein the C being not handled byiIt does not include one-component C1；Since the noise proportional in IMF1 is very big, give up first
A component IMF1, that is, C1, it reconstructs as follows:
In formulaTo reconstruct the signal after obtained denoising, q is to be modified the processed IMF component of wavelet threshold algorithmsQuantity, p is the remaining IMF component C not handlediQuantity.
The present invention has a characteristic that compared with the denoising algorithm of existing many surface electromyogram signals
Since surface electromyogram signal is non-linear, non-stationary signal, energy focuses primarily upon low frequency part, utilizes the present invention
Denoising is carried out, can be reduced since modal overlap bring adversely affects, as much as possible remain has in signal
Information reduces the influence of noise bring, while improving discrimination.
Detailed description of the invention
Fig. 1 is implementation flow chart of the invention；
Fig. 2 is the electromyography signal figure of the original Noise of the embodiment of the present invention；
Fig. 3 is the result figure that the embodiment of the present invention passes through that CEEMD is decomposed；
Fig. 4 is component of the related coefficient of the selection of the embodiment of the present invention greater than threshold value T by improving wavelet threshold processing
The component obtained afterwards；
Fig. 5 is the denoising result figure of the embodiment of the present invention；
Specific embodiment
Elaborate with reference to the accompanying drawing to the embodiment of the present invention: the present embodiment before being with technical solution of the present invention
It puts and is implemented, the detailed implementation method and specific operation process are given, but protection scope of the present invention is not limited to down
The embodiment stated.
As shown in Figure 1 and Figure 2, the present embodiment includes the following steps:
The surface electromyogram signal x (t) of step 1, acquisition can contain noise n (t):
X (t)=s (t)+δ n (t)
In formula, s (t) is the surface electromyogram signal of not Noise, and x (t) is noise-containing surface electromyogram signal, and n (t) is
White noise, δ are the scale factor of noise.
Step 2 carries out complementary set empirical mode decomposition to the resulting signals and associated noises x (t) of step 1, and specific steps are such as
Under:
1. a pair of complementary positive and negative opposite white noise of signals and associated noises x (t) addition, to generate two groups of set IMF.
X (t) is signals and associated noises in formula；NjIt (t) is the aid in noise being added,For the noisy letter that positive aid in noise is added
Number,For the signal of the negative aid in noise of addition, M is the number that aid in noise is added, and j is the iteration of jth time in the present embodiment
In, M=20.
②WithIt is decomposed by EMD, respectively obtains IMF componentWithThe layer of IMF is obtained in embodiment
Number is 12.
3. calculating the average value of multiple IMF components, obtain.
IMF component is denoted as C in formulai.The different scale component obtained by CEEMD complementation set empirical mode decomposition is as schemed
Shown in 3.
Step 3 finds intrinsic mode function IMF points for capableing of representation signal main feature using component correlation analysis
Amount handles the improvement wavelet threshold that the component of selection carries out next step, the specific steps are as follows:
1. calculating x (t) and each component Ci(t) variance between:
C in formulaxCIt (i) is x (t) and Ci(t) variance, uxFor the mean value of x (t), uCiFor Ci(t) mean value.
2. calculating related coefficient:
ρ in formulaxCIt (i) is x (t) and Ci(t) related coefficient, σxAnd σCiRespectively represent x (t) and Ci(t) standard deviation.
Related coefficient is referred to 3. calculating, T is to calculate as follows with reference to related coefficient:
N is the number of plies for decomposing obtained IMF in formula；Find correlation coefficient ρxC(i) it is greater than the component with reference to related coefficient T,
The ρ of componentxC(i) when being greater than with reference to related coefficient T, the property of the component energy representation signal at place, i.e. component Ci(t) it is chosen
It selects.
N=12 in the present embodiment, T=0.2161.Find correlation coefficient ρxC(i) it is greater than the component with reference to related coefficient T.
And these components are improved into wavelet threshold processing.
The related coefficient of table 1 each component and signals and associated noises
Find correlation coefficient ρxC(i) it is as shown in table 1 to be greater than the related coefficient being calculated with reference to the component of related coefficient T.
Showing IMF2, IMF3, IMF4 according to the related coefficient of table 1, IMF5, IMF6, the value of IMF7, which is higher than, refers to related coefficient T, so
Component higher than reference related coefficient T can indicate the main feature of sEMG signal.The high frequency being expressed as due to IMF1 in signal
Part, the noise proportional for including among these is very big, we give up one-component IMF1 i.e. C1。
Step 4 is handled the component selected using wavelet thresholding methods are improved, to each selected component
Ci(t) it is handled, the IMF that obtains that treated, i.e.,
In the present embodiment, IMF2 step 3 obtained, IMF3, IMF4, IMF5, IMF6, IMF7 are by improving section
Threshold value is handled to have obtained the later each component of processing, as shown in Figure 4.
Step 5 obtains processed part component with step 4 and is denoted asWith the C being not handled byi(remove the first point
Amount) signal reconstruction is carried out, since the noise proportional in IMF1 is very big, we have given up one-component IMF1 i.e. Ci, reconstruct
It is as follows:
In formulaTo reconstruct obtained denoised signal, q is to be modified the processed IMF component of wavelet threshold algorithms's
Quantity, p are remaining without processed IMF component CiQuantity.
It is as shown in Figure 5 to reconstruct obtained denoised signal.N is the quantity by improving the processed component of wavelet threshold, and m is
It is not modified the component of wavelet threshold processing.Since the noise proportional in IMF1 is very big, we have given up one-component
IMF1, that is, C1.By by improving wavelet threshold, treated in step 4The C obtained with step 28,
C9,C10,C11,C12Addition obtains reconstruction signal.
Using EMD, EMD-IT and this method carry out the comparison of denoising effect, obtain that the results are shown in Table 2.SNRinIt is defeated
Enter the signal-to-noise ratio of signal, SNRoutFor the signal-to-noise ratio obtained after denoising, RSME is root mean square error.
The Comparative result of 2 three kinds of denoising methods of table denoising
Claims (1)
1. the surface electromyogram signal noise-eliminating method based on CEEMD and improvement wavelet threshold, it is characterised in that this method includes as follows
Step:
The surface electromyogram signal x (t) of step (1), acquisition can contain noise n (t):
X (t)=s (t)+δ n (t)
In formula, s (t) is the surface electromyogram signal of not Noise, and x (t) is noise-containing surface electromyogram signal, and n (t) is white noise
Sound, δ are the scale factor of noise；
Step (2) carries out complementary set empirical mode decomposition to the resulting signals and associated noises x (t) of step (1), first by noisy letter
A pair of complementary positive and negative opposite white noise of number x (t) addition, obtains the signal x that positive aid in noise is added+With the negative aid in noise of addition
Signal x-, decomposition two groups of set intrinsic mode function components of generation, which are carried out, using empirical mode decomposition algorithm is denoted asWithI-th of IMF component that i is；Average statistical using uncorrelated random sequence is 0；
Step (3), repeats step (2) M times, finally obtains the mean value C of IMFi(t)；
Step (4) finds the intrinsic mode function IMF component for capableing of representation signal main feature using component correlation analysis, will
The component of selection carries out the improvement wavelet threshold processing of next step, the specific steps are as follows:
1) x (t) and each component C are calculatedi(t) variance between:
C in formulaxCIt (i) is x (t) and Ci(t) variance, uxFor the mean value of x (t), uCiFor Ci(t) mean value；
2) related coefficient is calculated:
ρ in formulaxCIt (i) is x (t) and Ci(t) related coefficient, σxAnd σCiRespectively represent x (t) and Ci(t) standard deviation；
3) it calculates and refers to related coefficient, J is to calculate as follows with reference to related coefficient:
N is the number of plies for decomposing obtained IMF in formula；Find correlation coefficient ρxC(i) it is greater than the component with reference to related coefficient J, component
ρxC(i) when being greater than with reference to related coefficient J, the property of the component energy representation signal at place, i.e. component Ci(t) it is selected；
Step (5) is handled the component selected using wavelet thresholding methods are improved, to each selected component Ci(t)
Do following processing:
1) it is changed commanders C using discrete wavelet transformeri(t) it is decomposed into multilayer, is defined " peak and the ratio " of detail coefficients, specific as follows:
Wherein wjFor the wavelet coefficient of jth layer, wherein wj,iIt is the wavelet coefficient of j layers of i point；If Sj≤ 0.2 < Sj+1, then j is selected
IMF carries out the Decomposition order of wavelet transformation thus；
2) the bound λ of threshold value is selectedLAnd λHHandle each layer of wavelet coefficient,
λj,L=μj-κj,Lσj
λj,H=μj+κj,Hσj
Wherein λj,LAnd λj,HIt is the threshold value bound of jth layer；κj,LAnd κj,HIt is adjustable parameter；μjAnd σjIt is jth layer wavelet coefficient
Mean value and variance；
Then, κ is calculatedj,LAnd κj,HMinimum value obtain the exact value of κ；
If Sj≤ 0.01, then it is arrangedWithOtherwise,
Sr,LAnd Sr,HIt is defined asWithSj,LAnd Sj,HIt is that wavelet coefficient is positive and negative respectively
Partial peak and value, wherein L < k < H；
After each layer of threshold value has been determined, wavelet coefficient is handled as follows:
3) with treated wavelet coefficient reconstructs the IMF that is exactly that treated, i.e.,
Step (6) obtains processed part component with step (5) and is denoted asWith the C being not handled byiSignal reconstruction is carried out,
In the C that is not handled byiIt does not include one-component C1；Since the noise proportional in IMF1 is very big, one-component is given up
IMF1, that is, C1, it reconstructs as follows:
In formulaTo reconstruct the signal after obtained denoising, q is to be modified the processed IMF component of wavelet threshold algorithms's
Quantity, p are the remaining IMF component C not handlediQuantity.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910285085.5A CN110151175A (en) | 2019-04-10 | 2019-04-10 | Surface electromyogram signal noise-eliminating method based on CEEMD and improvement wavelet threshold |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910285085.5A CN110151175A (en) | 2019-04-10 | 2019-04-10 | Surface electromyogram signal noise-eliminating method based on CEEMD and improvement wavelet threshold |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110151175A true CN110151175A (en) | 2019-08-23 |
Family
ID=67638544
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910285085.5A Pending CN110151175A (en) | 2019-04-10 | 2019-04-10 | Surface electromyogram signal noise-eliminating method based on CEEMD and improvement wavelet threshold |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110151175A (en) |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110646682A (en) * | 2019-09-23 | 2020-01-03 | 辽宁工程技术大学 | System for monitoring interference potential of metal buried pipeline in real time |
| CN111553308A (en) * | 2020-05-11 | 2020-08-18 | 成都亿科康德电气有限公司 | Reconstruction method of partial discharge signal of power transformer |
| CN111582132A (en) * | 2020-04-30 | 2020-08-25 | 南京信息工程大学 | Improved EEMD and PCNN-based gas leakage signal noise reduction method |
| CN112923847A (en) * | 2021-01-21 | 2021-06-08 | 广东工业大学 | Local sine auxiliary grating ruler measurement error adaptive compensation method |
| CN113925482A (en) * | 2021-10-26 | 2022-01-14 | 西安易朴通讯技术有限公司 | Heart rate calculation method, wearable electronic device and storage medium |
| CN114027847A (en) * | 2021-11-17 | 2022-02-11 | 湖南万脉医疗科技有限公司 | Electrocardiosignal analysis method based on time-frequency analysis |
| CN114088077A (en) * | 2021-12-10 | 2022-02-25 | 哈尔滨工业大学 | Improved hemispherical resonator gyro signal denoising method |
| CN116318437A (en) * | 2023-03-16 | 2023-06-23 | 中国科学院空天信息创新研究院 | Cross-medium communication interference suppression method and system |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5175710A (en) * | 1990-12-14 | 1992-12-29 | Hutson William H | Multi-dimensional data processing and display |
| CN102631195A (en) * | 2012-04-18 | 2012-08-15 | 太原科技大学 | Single-channel blind source separation method of surface electromyogram signals of human body |
| CN109145729A (en) * | 2018-07-13 | 2019-01-04 | 杭州电子科技大学 | Based on the electromyography signal denoising method for improving wavelet threshold and EEMD |
| CN109589114A (en) * | 2018-12-26 | 2019-04-09 | 杭州电子科技大学 | Myoelectricity noise-eliminating method based on CEEMD and interval threshold |
-
2019
- 2019-04-10 CN CN201910285085.5A patent/CN110151175A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5175710A (en) * | 1990-12-14 | 1992-12-29 | Hutson William H | Multi-dimensional data processing and display |
| CN102631195A (en) * | 2012-04-18 | 2012-08-15 | 太原科技大学 | Single-channel blind source separation method of surface electromyogram signals of human body |
| CN109145729A (en) * | 2018-07-13 | 2019-01-04 | 杭州电子科技大学 | Based on the electromyography signal denoising method for improving wavelet threshold and EEMD |
| CN109589114A (en) * | 2018-12-26 | 2019-04-09 | 杭州电子科技大学 | Myoelectricity noise-eliminating method based on CEEMD and interval threshold |
Cited By (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110646682A (en) * | 2019-09-23 | 2020-01-03 | 辽宁工程技术大学 | System for monitoring interference potential of metal buried pipeline in real time |
| CN110646682B (en) * | 2019-09-23 | 2021-06-22 | 辽宁工程技术大学 | System for monitoring interference potential of metal buried pipeline in real time |
| CN111582132A (en) * | 2020-04-30 | 2020-08-25 | 南京信息工程大学 | Improved EEMD and PCNN-based gas leakage signal noise reduction method |
| CN111582132B (en) * | 2020-04-30 | 2023-04-18 | 南京信息工程大学 | Improved EEMD and PCNN-based gas leakage signal noise reduction method |
| CN111553308A (en) * | 2020-05-11 | 2020-08-18 | 成都亿科康德电气有限公司 | Reconstruction method of partial discharge signal of power transformer |
| CN112923847A (en) * | 2021-01-21 | 2021-06-08 | 广东工业大学 | Local sine auxiliary grating ruler measurement error adaptive compensation method |
| CN112923847B (en) * | 2021-01-21 | 2022-03-18 | 广东工业大学 | Local sine auxiliary grating ruler measurement error adaptive compensation method |
| CN113925482A (en) * | 2021-10-26 | 2022-01-14 | 西安易朴通讯技术有限公司 | Heart rate calculation method, wearable electronic device and storage medium |
| CN114027847A (en) * | 2021-11-17 | 2022-02-11 | 湖南万脉医疗科技有限公司 | Electrocardiosignal analysis method based on time-frequency analysis |
| CN114088077A (en) * | 2021-12-10 | 2022-02-25 | 哈尔滨工业大学 | Improved hemispherical resonator gyro signal denoising method |
| CN116318437A (en) * | 2023-03-16 | 2023-06-23 | 中国科学院空天信息创新研究院 | Cross-medium communication interference suppression method and system |
| CN116318437B (en) * | 2023-03-16 | 2023-12-01 | 中国科学院空天信息创新研究院 | Cross-medium communication interference suppression method and system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110151175A (en) | Surface electromyogram signal noise-eliminating method based on CEEMD and improvement wavelet threshold | |
| CN102697495B (en) | Second-generation wavelet electromyographic signal noise eliminating method based on ensemble empirical mode decomposition | |
| Alfaouri et al. | ECG signal denoising by wavelet transform thresholding | |
| CN109589114A (en) | Myoelectricity noise-eliminating method based on CEEMD and interval threshold | |
| CN103961092B (en) | EEG Noise Cancellation based on adaptive thresholding | |
| CN109145729A (en) | Based on the electromyography signal denoising method for improving wavelet threshold and EEMD | |
| CN102928517A (en) | Method for denoising acoustic testing data of porcelain insulator vibration based on wavelet decomposition threshold denoising | |
| CN110680308B (en) | Electrocardiosignal denoising method based on fusion of improved EMD and threshold method | |
| Reddy et al. | ECG De-Noising using improved thresholding based on Wavelet transforms | |
| CN103932687B (en) | Method and device for preprocessing pulse condition signal | |
| CN110059639A (en) | A kind of Frequency Hopping Signal detection method based on fractional wavelet transform and Hilbert-Huang transform | |
| CN112084845B (en) | Low-frequency 1/f noise elimination method based on multi-scale wavelet coefficient autocorrelation | |
| CN104732498B (en) | A kind of thresholded image denoising method based on non-downsampling Contourlet conversion | |
| Zhang et al. | ECG signals denoising method based on improved wavelet threshold algorithm | |
| CN104280776B (en) | Self-adaption wavelet threshold solving method | |
| Wang et al. | Research on denoising algorithm for ECG signals | |
| Ayat et al. | ECG denoising using modulus maxima of wavelet transform | |
| Li | Speech denoising based on improved discrete wavelet packet decomposition | |
| CN103915102B (en) | Method for noise abatement of LFM underwater sound multi-path signals | |
| Huang et al. | ECG baseline wander correction based on ensemble empirical mode decomposition with complementary adaptive noise | |
| CN116584960A (en) | Diaphragmatic electromyographic signal noise reduction method | |
| CN113204051B (en) | Low-rank tensor seismic data denoising method based on variational modal decomposition | |
| CN115376540A (en) | Biological radar voice enhancement method and system based on variational modal decomposition | |
| Jaffery et al. | Performance Comparision of Wavelet Threshold Estimators for ECG Signal Denoising | |
| CN107203972A (en) | A kind of Ultrasonic Image Denoising and Enhancement Method |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190823 |
|
| RJ01 | Rejection of invention patent application after publication |