Surface EMG Based Basic Hand Movement Detection using Singular Value Decomposition

Surface electromyography (sEMG) signals play a vital role in action detection. Researchers have been trying to classify different types of hand movements using different methodologies from image, video and physiological signals like sEMG. Hand movements from sEMG signals have been classified using statistical features, AR parameters and reflection coefficients which need vigorous work to extract feature space. In this paper, an efficient and effective method to analyze the sEMG signal is proposed based on singular value decomposition. In the proposed method sEMG signals are decomposed into intrinsic mode functions (IMFs) using empirical mode decomposition (EMD) algorithm. Then singular value decomposition is used on these IMFs to extract singular values that are used as features. A publicly available dataset is used to verify our assumption and K-nearest neighbor (KNN) is adopted to classify six basic hand movements. Experimental results prove that our proposed method can achieve considerably high accuracy with a lower computational complexity which outperforms other methods and can be used in analyzing sEMG signals,