Am trying to extract features from EMG signal and attaching the code herewith... plz help me whether i have provided the proper initialization...

x = load('emg.mat');

function feat = getfTDDfeat_Online(x,steps,winsize,wininc)
NFPC       = 6;% we extract 6 features/channel
steps = 4;
datasize   = size(x,1);
Nsignals   = size(x,2);
winsize    = 150;
wininc     = 50;
numwin     = floor((datasize - winsize)/wininc)+1;
NumSigFeat = Nsignals*NFPC;
% allocate memory
feat       = zeros(numwin,Nsignals*NFPC);
% sliding windows increments
st         = 1;
en         = winsize;
for i = 1:numwin
    
    curwin = x(st:en,:);
    
    % steps1: Extract features from original signal and a nonlinear version of it
    ebp = KSM1(curwin);
    efp = KSM1(log(curwin.^2+eps).^2);
    
    % steps2: Correlation analysis
    num = -2.*ebp.*efp;num = sum(num'*num)./NumSigFeat;
    den = efp.*efp+ebp.*ebp;den = sum(den'*den)./NumSigFeat;
    
   

Answer 1

[arushi]

Hi Emimal,

I understand that you want to extract the features from EMG signals file. In the provided code, there are a few potential areas of improvement:

1. Function Input Parameters: It seems that the input parameters steps, winsize, and wininc are defined as input arguments to the function getfTDDfeat_Online. However, within the function, these parameters are reassigned with fixed values. This might cause confusion and unexpected behavior. If these values are meant to be fixed, it's better to remove them from the function signature and define them directly within the function.
2. Data Loading: The code starts with loading the EMG data from a file named 'emg.mat'. It's assumed that this file contains the EMG data in a suitable format for further processing.
3. Sliding Window: The code sets up a sliding window loop to process the EMG signal in overlapping windows. The current implementation appears to extract features from each window separately.

Hope this helps.