Algorithm for compression of EMG signals

This paper presents an algorithm for the compression of electromyography (EMG) signals. Its main application is in the efficient construction banks of EMG signals. The developed algorithm uses an orthogonal transform followed by a thresholding process for choosing significant coefficients. Many orthogonal transforms were studied and the Daubechies-4 wavelet transform has shown a better performance for data compression. The proposed algorithm was implemented, and the experimental results have shown that a compression gain greater than 10:1 can be achieved without perceptible errors in the decoded signal.