Surface electromyography (EMG) is a noninvasive signal acquisition method that plays a central role in too many applications, including clinical diagnostics, prosthetic device monitoring, biomedicine, and human-machine interactions. Processing commonly begins with feature extraction, that is, by surface electrodes to amputees, then follows with the application of a dimensionality reduction technique (a time window is used to segment the data). The reduced features obtained are inputs to a machine learning classifier. The built machine learning model can classify the new registered movements. Features extracted from EMG signals typically capture information from both the time and frequency domains. To then apply the spectral analysis using the reduced time Fourier transform (STFT) in order to produce the time-frequency graph of the EMG signal in the recognition of hand grip postures.