Classification of EMG signals

The EMG signal is the process of measuring the electrical activation that occurs as a result of muscular contraction. For this reason, EMG signals from the muscles provide information about the muscles. This information is currently used in the diagnosis of muscular diseases, prosthetic arm and motion detection studies. In this thesis, motion detection is aimed by EMG signals using Artificial Neural Networks. Primarily, time-frequency representations of signals are obtained by applying Short Time Fourier Transform (STFT) to the received EMG signals. From the obtained time-frequency properties, the attributes of the EMG signal were extracted with the statistical methods, Gray-Level Co-Occurrence Matrix (GLCM) and Local Binary Pattern (LBP) methods. These extracted attributes are given as input data to Artificial Neural Network (ANN) and the system performance is calculated. When the experimental results were examined, it was observed that the designed systsem had a successful result on the used EMG data. KEYWORDS: EMG Signal Processing, Short Time Fourier Transform, Gray Level Co-Occurrence Matrix, Local Binary Patterns, Artificial Neural Networks