how to choose windwo size in the EMG signal ?

Question

Seemab 2024-6-1

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/2124581-how-to-choose-windwo-size-in-the-emg-signal

评论： Star Strider 2024-6-3

Ciao,

If I have an emg dataset of length 260001 almost 4.3min and fs=1000. I want to analyse the data in small windows,but I am unable to decide how can I choose a window size. I need a scientific calculation to choose a window size. I read somewhere that the The duration of the window must be five time longer than the period of the signal, that is :T(Window) = 5* T(Signal) but i need a proper reasoning.

Thank you

0 个评论
显示 -2更早的评论隐藏 -2更早的评论

请先登录，再进行评论。

请先登录，再回答此问题。

Answer 1

Star Strider 2024-6-1

0
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/2124581-how-to-choose-windwo-size-in-the-emg-signal#answer_1466241

I am not certain where you read it either, or the rationale for that particular window length calculation.

I am not aware of that particular value for the window length, since it usually in terms of a specific time segment (usually 100-150 ms) rather than a fraction of the total signal length, with varying overlap percentages.

A PubMed search just now turned up 83 citations. Choose an approach that works for your particular application, then use the stft function to calculate the Fourier transforms (that I assume is what you want to do). It should be able to do everything you want, including setting the window length, window function, and percent overlap between the samples.

.

4 个评论
显示 2更早的评论隐藏 2更早的评论

Star Strider 2024-6-2

That document refers to sound files in that software version. MATLAB can do everything you need to in that regard, and at least as I read it, everything that software does if you have the Signal Processing Toolbox licensed and installed.

You asked about EMG signal processing, and that requires a different approach because the EMG signals are not sound files (although it may be possible to listen to them, however I have never thought of doing that).

With respect to the window (and window function) size, review the documents I discovered in my PubMed search. The window size may change depending on what you want to do with the information. Also, the stft function has its own way of windowing the data it is presented with. I suggest that you review the stft documentation that I linked to in my original Answer. The purpose of window functions — except the rectangular window, since it is essentially no windoow — is to correct for the discrete Fourier transform being finite rather than infinite. Windowing using the commonly-used hann window produces a significantly improved spectrum, as does zero-padding the fft (uising the second argument to fft) to improve the frequency resolution of the transformed signal.

(I have not done any sort of classification with respect to EMG signals, although I have — and published on — classification of EEG signals.)

Seemab 2024-6-3

I read the paper you suggested. thanks for that.

Is it possible to find the minimum centeroid value for multiple combinations of the window size and overlap and then choose the combination with the minimum centeroid value ?

Star Strider 2024-6-3

I do not understand what you want to do.

I would experiment with the window size (and if necessary, overlap fraction) that works best in whatever you are doing. I do not believe there is any one specific window size that would work best in every situation, since that may depend on the spectral features you want to work with, and the overall length of your signal. (When yoiu decide on that, use it throughout your analysis.) Longer windows will have greater frequency resolution.

请先登录，再进行评论。

how to choose windwo size in the EMG signal ?

0 个评论
显示 -2更早的评论隐藏 -2更早的评论

采纳的回答

4 个评论
显示 2更早的评论隐藏 2更早的评论

更多回答（0 个）

另请参阅

类别

标签

产品

Community Treasure Hunt

how to choose windwo size in the EMG signal ?

0 个评论 显示 -2更早的评论隐藏 -2更早的评论

采纳的回答

4 个评论 显示 2更早的评论隐藏 2更早的评论

更多回答（0 个）

另请参阅

类别

标签

产品

Community Treasure Hunt

0 个评论
显示 -2更早的评论隐藏 -2更早的评论

4 个评论
显示 2更早的评论隐藏 2更早的评论