Hand-motion Pattern Recognition of SEMG Based on Hilbert-Huang Transformation and AR-model

In order to recognize the hand-motions based on the Surface Electromyography(SEMG), a feature extracting algorithm is presented in the paper, which is built by combining Hilbert-Huang Transformation(HHT) with AR-model. According to the frequency-credit of each Intrinsic Mode Function (IMF) after HHT, six frequency-effective IMFs are selected. At the same time, the rectangle window is built based on the motion-start time and the motion-end time to extract motion signal of the six IMFs. The motion-start and motion-end points are decided by the instantaneous amplitude of the IMF with the largest frequency-credit. AR-model is built to extract the SEMG features for each IMF which are selected by frequency-credit and processed by the rectangle window. At last, two SEMG makes up the motion-feature vector. At last, Principal Components of the motion-feature vector are extracted and then put them into the Support Vector Machine (SVM) classifier to do the multi-patterns classification. The experimental results indicate that above method can have a good performance of discriminating four hand-movement patterns namely, palmar dorsiflexion and flexion, hand opening and closing.