Dear James

After the data collection, the acquired samples were converted into features that could be used for classification. Matlab software was used to extract and classify the features using the recorded EMG signals. Features were extracted from the samples by segmenting the signal into 250 ms intervals corresponding to 256 samples in each segment. A single feature was calculated from each segment and the segment window was incremented by 125 ms (128 samples) for the next feature. This scheme ensured that a control command could be generated within 250 ms from the instant the user's intention was given. Three kinds of features were extracted from each segment namely EMG rms value, AR model coefficients and WL.

You can use Matlab Library and reach it by this link:

http://www.sce.carleton.ca/faculty/chan/matlab/matlab_library.htm

Hope it's help you

Kinds Regards

Article Surface EMG pattern recognition for real-time control of a w...

Book Digital Signal Processing using MATLAB

Dear Saif Aldeen Saad Alkadhim,

may i know how to segment the signal into 256 sample? for example i had 2560 samples per signal, is that i am using reshape function in matlab to segment it? 

y = reshape(signal, 256);