Real-time Pattern Recognition for Hand Gesture Based on ANN and Surface EMG

Hand gesture recognition has many applications in engineering and medical fields. This paper proposes a real-time hand gesture recognition method using the surface electromyographic (sEMG) signal on the forearm. we apply a sliding window which allow us to observe a segment of the signals. Features are extracted from the signals of each sliding window, and then are inputted into a feed-forward artificial neural network (ANN) classifier which is trained at first. When the number of one hand gesture type of identified features reaches the threshold, it would be considered that the hand gesture is identified. Experiments show that the classification accuracy of real-time systems reaches 96%, and hand gestures can be recognized before they are completed.