Savitzky-Golay filtering on emg signal

Question

Zainab 2024-6-26，3:15

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/2132006-savitzky-golay-filtering-on-emg-signal

回答： Star Strider 2024-6-26，11:05

Hello everyone,

I am working on a project that uses SEMG signals to move a prosthetic arm. I am trying to figure out how to use the Savitzky Golay filter to smooth the "sound" on the graph. I am trying to only have the full "contractions" and the smooth the "rest" parts of the graph

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

Diego Caro 2024-6-26，5:55

0
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/2132006-savitzky-golay-filtering-on-emg-signal#answer_1477141

在 MATLAB Online 中打开

What works best for me is just filtering the EMG signal with a lowpass butterworth filter with a very low cutoff frequency.

For example:

% Take a look at the raw signal
t = time_range;
emg = raw_emg;
figure
plot(t,emg)

Now lets add the lowpass filter and compare its output.

fc = 3;     % Cutoff frequency in Hz.
n = 4;      % Filter order.
fs = 100;   % Sample rate. 
[b,a] = butter(n,fc/(fs/2),"low"); 
emg_filt = filter(b,a,emg);
figure
hold on
plot(t,emg)
plot(t,emg_filt,'LineWidth',1.5,'Color','r') 
legend('Raw EMG','Filtered EMG')

Here I used a cutoff frequency of 3 Hz, which is very low, but you can go as low as you want if you desire a more agressive smoothing. There are several other methods you can use to smooth an EMG signal, but as I said, this is my personal preference.

Regards,

Diego.

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。

Answer 2

Umar 2024-6-26，7:08

0
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/2132006-savitzky-golay-filtering-on-emg-signal#answer_1477241

Hi Zainab,

To achieve this in Matlab, you can utilize the sgolayfilt function. This function applies the Savitzky-Golay filter to your signal. You can specify the frame length and polynomial order to control the smoothing effect. Here's a simple example of how to use the Savitzky-Golay filter in Matlab:

% Sample SEMG signal data signal = your_signal_data;

% Define parameters for the filter frame_length = 15; % Adjust based on your signal characteristics poly_order = 3; % Adjust for desired smoothing

% Apply Savitzky-Golay filter smoothed_signal = sgolayfilt(signal, poly_order, frame_length);

% Plot the original and smoothed signals figure; plot(signal, 'b', 'LineWidth', 1.5); hold on; plot(smoothed_signal, 'r', 'LineWidth', 1.5); legend('Original Signal', 'Smoothed Signal'); xlabel('Time'); ylabel('Amplitude'); title('SEM Signal Smoothing using Savitzky-Golay Filter');

By adjusting the frame_length and poly_order parameters, you can tailor the smoothing effect to enhance the "contractions" while effectively smoothing the "rest" parts of your SEMG signal graph.

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。

Answer 3

Star Strider 2024-6-26，11:05

0
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/2132006-savitzky-golay-filtering-on-emg-signal#answer_1477406

Calcualte the Fourier transform of the data (fft) to determine tthe frequency content of the signal, and then use a frequency-selective filter (either lowpass or bandpass) to remove the high-frequency noise and, if necessary, baseline offset and baseline drift. The Savitzky-Golay filter is essentially a multiband FIR filter that while appropriate in some applications, is simply too slow for any sort of real-time application. Experiment with the discrete (digital) filters with your recorded signals, and then implement them in hardware for your intended real-time continuous application.