Research on Gesture Recognition Algorithms Based on Electromyographic Signals

With the advancement of myoelectric control technology, surface electromyography (sEMG) signals have been widely applied in prosthesis control. However, effectively decoding movement intentions remains a challenge, making the construction of intention recognition models a critical issue. This study proposes combining the transient trigonometric Harris Hawks optimizer (TTHHO) with a radial basis function (RBF) network to build a gesture recognition model. Five hand gestures were collected from ten volunteers using a self-developed acquisition board, and the performance of the proposed model was compared with linear discriminant analysis (LDA), support vector machine (SVM), and K-nearest neighbor (KNN) models. The results show that the proposed model achieved an average recognition accuracy of 87.54%, which is 14% to 16% higher than the other three models, with a maximum accuracy of 92.47%.