Surface myoelectric signals decoding using the continuous wavelet transform singularity detection

Electromyographic (EMG) signals are the resultant of electrical activity of muscle fibers during a muscle contraction, whose pattern can provide a significant reference of a motor rehabilitation system. The EMG decoding method using “refractory period” and “threshold” is appropriate for real-time processing system due to its low algorithm complexity and the good fidelity of time domain information. In this paper, the distribution of intervals between continuous wavelet transform modulus maxima was analyzed to provide a reasonable determination of the “refractory period”. In addition, the source signals were decoded according to the “refractory period”. Promising results are demonstrated.